{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.999883054613495,
  "eval_steps": 100,
  "global_step": 42750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002338907730090048,
      "grad_norm": 3.421875,
      "learning_rate": 6.000000000000001e-07,
      "loss": 1.8651,
      "step": 1
    },
    {
      "epoch": 0.0004677815460180096,
      "grad_norm": 3.25,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 1.9817,
      "step": 2
    },
    {
      "epoch": 0.0007016723190270144,
      "grad_norm": 3.71875,
      "learning_rate": 1.8e-06,
      "loss": 2.1048,
      "step": 3
    },
    {
      "epoch": 0.0009355630920360191,
      "grad_norm": 6.15625,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 2.0897,
      "step": 4
    },
    {
      "epoch": 0.0011694538650450239,
      "grad_norm": 3.890625,
      "learning_rate": 3e-06,
      "loss": 1.9743,
      "step": 5
    },
    {
      "epoch": 0.0014033446380540288,
      "grad_norm": 4.0625,
      "learning_rate": 3.6e-06,
      "loss": 2.2394,
      "step": 6
    },
    {
      "epoch": 0.0016372354110630336,
      "grad_norm": 3.71875,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 1.9162,
      "step": 7
    },
    {
      "epoch": 0.0018711261840720383,
      "grad_norm": 3.984375,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.903,
      "step": 8
    },
    {
      "epoch": 0.0021050169570810432,
      "grad_norm": 4.0,
      "learning_rate": 5.4e-06,
      "loss": 2.119,
      "step": 9
    },
    {
      "epoch": 0.0023389077300900478,
      "grad_norm": 3.90625,
      "learning_rate": 6e-06,
      "loss": 1.8696,
      "step": 10
    },
    {
      "epoch": 0.0025727985030990527,
      "grad_norm": 3.21875,
      "learning_rate": 6.6e-06,
      "loss": 1.8993,
      "step": 11
    },
    {
      "epoch": 0.0028066892761080577,
      "grad_norm": 4.28125,
      "learning_rate": 7.2e-06,
      "loss": 2.1159,
      "step": 12
    },
    {
      "epoch": 0.003040580049117062,
      "grad_norm": 3.15625,
      "learning_rate": 7.8e-06,
      "loss": 2.1535,
      "step": 13
    },
    {
      "epoch": 0.003274470822126067,
      "grad_norm": 4.21875,
      "learning_rate": 8.400000000000001e-06,
      "loss": 2.2426,
      "step": 14
    },
    {
      "epoch": 0.003508361595135072,
      "grad_norm": 3.796875,
      "learning_rate": 9e-06,
      "loss": 2.0264,
      "step": 15
    },
    {
      "epoch": 0.0037422523681440766,
      "grad_norm": 3.921875,
      "learning_rate": 9.600000000000001e-06,
      "loss": 2.2371,
      "step": 16
    },
    {
      "epoch": 0.003976143141153081,
      "grad_norm": 5.8125,
      "learning_rate": 1.02e-05,
      "loss": 2.1816,
      "step": 17
    },
    {
      "epoch": 0.0042100339141620865,
      "grad_norm": 3.8125,
      "learning_rate": 1.08e-05,
      "loss": 2.2146,
      "step": 18
    },
    {
      "epoch": 0.004443924687171091,
      "grad_norm": 3.78125,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 1.7436,
      "step": 19
    },
    {
      "epoch": 0.0046778154601800955,
      "grad_norm": 3.53125,
      "learning_rate": 1.2e-05,
      "loss": 2.0947,
      "step": 20
    },
    {
      "epoch": 0.004911706233189101,
      "grad_norm": 3.0,
      "learning_rate": 1.26e-05,
      "loss": 2.0194,
      "step": 21
    },
    {
      "epoch": 0.005145597006198105,
      "grad_norm": 3.4375,
      "learning_rate": 1.32e-05,
      "loss": 2.0893,
      "step": 22
    },
    {
      "epoch": 0.00537948777920711,
      "grad_norm": 3.015625,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 1.8847,
      "step": 23
    },
    {
      "epoch": 0.005613378552216115,
      "grad_norm": 3.5,
      "learning_rate": 1.44e-05,
      "loss": 2.0642,
      "step": 24
    },
    {
      "epoch": 0.00584726932522512,
      "grad_norm": 3.1875,
      "learning_rate": 1.5e-05,
      "loss": 1.8056,
      "step": 25
    },
    {
      "epoch": 0.006081160098234124,
      "grad_norm": 2.859375,
      "learning_rate": 1.56e-05,
      "loss": 1.8361,
      "step": 26
    },
    {
      "epoch": 0.00631505087124313,
      "grad_norm": 2.953125,
      "learning_rate": 1.62e-05,
      "loss": 1.9673,
      "step": 27
    },
    {
      "epoch": 0.006548941644252134,
      "grad_norm": 3.75,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 2.0262,
      "step": 28
    },
    {
      "epoch": 0.006782832417261139,
      "grad_norm": 4.1875,
      "learning_rate": 1.74e-05,
      "loss": 2.2111,
      "step": 29
    },
    {
      "epoch": 0.007016723190270144,
      "grad_norm": 3.015625,
      "learning_rate": 1.8e-05,
      "loss": 2.0545,
      "step": 30
    },
    {
      "epoch": 0.007250613963279149,
      "grad_norm": 3.609375,
      "learning_rate": 1.86e-05,
      "loss": 2.2164,
      "step": 31
    },
    {
      "epoch": 0.007484504736288153,
      "grad_norm": 3.03125,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.8286,
      "step": 32
    },
    {
      "epoch": 0.0077183955092971586,
      "grad_norm": 3.171875,
      "learning_rate": 1.98e-05,
      "loss": 1.8854,
      "step": 33
    },
    {
      "epoch": 0.007952286282306162,
      "grad_norm": 3.71875,
      "learning_rate": 2.04e-05,
      "loss": 2.0545,
      "step": 34
    },
    {
      "epoch": 0.008186177055315168,
      "grad_norm": 3.484375,
      "learning_rate": 2.1e-05,
      "loss": 2.0008,
      "step": 35
    },
    {
      "epoch": 0.008420067828324173,
      "grad_norm": 4.125,
      "learning_rate": 2.16e-05,
      "loss": 2.171,
      "step": 36
    },
    {
      "epoch": 0.008653958601333177,
      "grad_norm": 3.453125,
      "learning_rate": 2.22e-05,
      "loss": 1.8945,
      "step": 37
    },
    {
      "epoch": 0.008887849374342182,
      "grad_norm": 3.46875,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 2.2038,
      "step": 38
    },
    {
      "epoch": 0.009121740147351187,
      "grad_norm": 2.953125,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 1.751,
      "step": 39
    },
    {
      "epoch": 0.009355630920360191,
      "grad_norm": 3.203125,
      "learning_rate": 2.4e-05,
      "loss": 2.0843,
      "step": 40
    },
    {
      "epoch": 0.009589521693369197,
      "grad_norm": 3.609375,
      "learning_rate": 2.4599999999999998e-05,
      "loss": 2.4521,
      "step": 41
    },
    {
      "epoch": 0.009823412466378202,
      "grad_norm": 3.3125,
      "learning_rate": 2.52e-05,
      "loss": 1.9771,
      "step": 42
    },
    {
      "epoch": 0.010057303239387206,
      "grad_norm": 3.484375,
      "learning_rate": 2.58e-05,
      "loss": 2.0536,
      "step": 43
    },
    {
      "epoch": 0.01029119401239621,
      "grad_norm": 3.515625,
      "learning_rate": 2.64e-05,
      "loss": 2.1146,
      "step": 44
    },
    {
      "epoch": 0.010525084785405215,
      "grad_norm": 3.90625,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 2.2792,
      "step": 45
    },
    {
      "epoch": 0.01075897555841422,
      "grad_norm": 2.875,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 1.6505,
      "step": 46
    },
    {
      "epoch": 0.010992866331423226,
      "grad_norm": 3.515625,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 2.2707,
      "step": 47
    },
    {
      "epoch": 0.01122675710443223,
      "grad_norm": 3.0,
      "learning_rate": 2.88e-05,
      "loss": 1.9885,
      "step": 48
    },
    {
      "epoch": 0.011460647877441235,
      "grad_norm": 4.03125,
      "learning_rate": 2.94e-05,
      "loss": 2.1422,
      "step": 49
    },
    {
      "epoch": 0.01169453865045024,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 50
    },
    {
      "epoch": 0.011928429423459244,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 51
    },
    {
      "epoch": 0.012162320196468249,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2971,
      "step": 52
    },
    {
      "epoch": 0.012396210969477255,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 53
    },
    {
      "epoch": 0.01263010174248626,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6241,
      "step": 54
    },
    {
      "epoch": 0.012863992515495264,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 55
    },
    {
      "epoch": 0.013097883288504268,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 56
    },
    {
      "epoch": 0.013331774061513273,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 57
    },
    {
      "epoch": 0.013565664834522278,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2295,
      "step": 58
    },
    {
      "epoch": 0.013799555607531282,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 59
    },
    {
      "epoch": 0.014033446380540288,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 60
    },
    {
      "epoch": 0.014267337153549293,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 61
    },
    {
      "epoch": 0.014501227926558297,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 62
    },
    {
      "epoch": 0.014735118699567302,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7501,
      "step": 63
    },
    {
      "epoch": 0.014969009472576306,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 64
    },
    {
      "epoch": 0.01520290024558531,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 65
    },
    {
      "epoch": 0.015436791018594317,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7687,
      "step": 66
    },
    {
      "epoch": 0.01567068179160332,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 67
    },
    {
      "epoch": 0.015904572564612324,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 68
    },
    {
      "epoch": 0.016138463337621332,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 69
    },
    {
      "epoch": 0.016372354110630337,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 70
    },
    {
      "epoch": 0.01660624488363934,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 71
    },
    {
      "epoch": 0.016840135656648346,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 72
    },
    {
      "epoch": 0.01707402642965735,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 73
    },
    {
      "epoch": 0.017307917202666355,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 74
    },
    {
      "epoch": 0.01754180797567536,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 75
    },
    {
      "epoch": 0.017775698748684364,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5842,
      "step": 76
    },
    {
      "epoch": 0.01800958952169337,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 77
    },
    {
      "epoch": 0.018243480294702373,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6144,
      "step": 78
    },
    {
      "epoch": 0.018477371067711378,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 79
    },
    {
      "epoch": 0.018711261840720382,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.885,
      "step": 80
    },
    {
      "epoch": 0.01894515261372939,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 81
    },
    {
      "epoch": 0.019179043386738395,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 82
    },
    {
      "epoch": 0.0194129341597474,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 83
    },
    {
      "epoch": 0.019646824932756404,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 84
    },
    {
      "epoch": 0.019880715705765408,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 85
    },
    {
      "epoch": 0.020114606478774413,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9811,
      "step": 86
    },
    {
      "epoch": 0.020348497251783417,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 87
    },
    {
      "epoch": 0.02058238802479242,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6319,
      "step": 88
    },
    {
      "epoch": 0.020816278797801426,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1914,
      "step": 89
    },
    {
      "epoch": 0.02105016957081043,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7219,
      "step": 90
    },
    {
      "epoch": 0.021284060343819435,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 91
    },
    {
      "epoch": 0.02151795111682844,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 92
    },
    {
      "epoch": 0.021751841889837444,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7365,
      "step": 93
    },
    {
      "epoch": 0.021985732662846452,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 94
    },
    {
      "epoch": 0.022219623435855457,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 95
    },
    {
      "epoch": 0.02245351420886446,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 96
    },
    {
      "epoch": 0.022687404981873466,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 97
    },
    {
      "epoch": 0.02292129575488247,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 98
    },
    {
      "epoch": 0.023155186527891475,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 99
    },
    {
      "epoch": 0.02338907730090048,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 100
    },
    {
      "epoch": 0.02338907730090048,
      "eval_runtime": 4.6305,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 100
    },
    {
      "epoch": 0.023622968073909484,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 101
    },
    {
      "epoch": 0.02385685884691849,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4909,
      "step": 102
    },
    {
      "epoch": 0.024090749619927493,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 103
    },
    {
      "epoch": 0.024324640392936497,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 104
    },
    {
      "epoch": 0.024558531165945502,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 105
    },
    {
      "epoch": 0.02479242193895451,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 106
    },
    {
      "epoch": 0.025026312711963514,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 107
    },
    {
      "epoch": 0.02526020348497252,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 108
    },
    {
      "epoch": 0.025494094257981523,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 109
    },
    {
      "epoch": 0.025727985030990528,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 110
    },
    {
      "epoch": 0.025961875803999532,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8166,
      "step": 111
    },
    {
      "epoch": 0.026195766577008537,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7855,
      "step": 112
    },
    {
      "epoch": 0.02642965735001754,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 113
    },
    {
      "epoch": 0.026663548123026546,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 114
    },
    {
      "epoch": 0.02689743889603555,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 115
    },
    {
      "epoch": 0.027131329669044555,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 116
    },
    {
      "epoch": 0.02736522044205356,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 117
    },
    {
      "epoch": 0.027599111215062564,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 118
    },
    {
      "epoch": 0.027833001988071572,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 119
    },
    {
      "epoch": 0.028066892761080577,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8354,
      "step": 120
    },
    {
      "epoch": 0.02830078353408958,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1028,
      "step": 121
    },
    {
      "epoch": 0.028534674307098586,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 122
    },
    {
      "epoch": 0.02876856508010759,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 123
    },
    {
      "epoch": 0.029002455853116595,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 124
    },
    {
      "epoch": 0.0292363466261256,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7526,
      "step": 125
    },
    {
      "epoch": 0.029470237399134604,
      "grad_norm": 7.6875,
      "learning_rate": 3e-05,
      "loss": 2.6772,
      "step": 126
    },
    {
      "epoch": 0.029704128172143608,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 127
    },
    {
      "epoch": 0.029938018945152613,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 128
    },
    {
      "epoch": 0.030171909718161617,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 129
    },
    {
      "epoch": 0.03040580049117062,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 130
    },
    {
      "epoch": 0.03063969126417963,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2192,
      "step": 131
    },
    {
      "epoch": 0.030873582037188634,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 132
    },
    {
      "epoch": 0.03110747281019764,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 133
    },
    {
      "epoch": 0.03134136358320664,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 134
    },
    {
      "epoch": 0.031575254356215644,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 135
    },
    {
      "epoch": 0.03180914512922465,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2295,
      "step": 136
    },
    {
      "epoch": 0.03204303590223366,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 137
    },
    {
      "epoch": 0.032276926675242665,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 138
    },
    {
      "epoch": 0.03251081744825167,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 139
    },
    {
      "epoch": 0.032744708221260674,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 140
    },
    {
      "epoch": 0.03297859899426968,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 141
    },
    {
      "epoch": 0.03321248976727868,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 142
    },
    {
      "epoch": 0.03344638054028769,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8156,
      "step": 143
    },
    {
      "epoch": 0.03368027131329669,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 144
    },
    {
      "epoch": 0.033914162086305696,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8719,
      "step": 145
    },
    {
      "epoch": 0.0341480528593147,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3186,
      "step": 146
    },
    {
      "epoch": 0.034381943632323705,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2829,
      "step": 147
    },
    {
      "epoch": 0.03461583440533271,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 148
    },
    {
      "epoch": 0.034849725178341714,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 149
    },
    {
      "epoch": 0.03508361595135072,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2706,
      "step": 150
    },
    {
      "epoch": 0.035317506724359723,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 151
    },
    {
      "epoch": 0.03555139749736873,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 152
    },
    {
      "epoch": 0.03578528827037773,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 153
    },
    {
      "epoch": 0.03601917904338674,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 154
    },
    {
      "epoch": 0.03625306981639574,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 155
    },
    {
      "epoch": 0.036486960589404746,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 156
    },
    {
      "epoch": 0.03672085136241375,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 157
    },
    {
      "epoch": 0.036954742135422755,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 158
    },
    {
      "epoch": 0.03718863290843176,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 159
    },
    {
      "epoch": 0.037422523681440764,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 160
    },
    {
      "epoch": 0.03765641445444977,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 161
    },
    {
      "epoch": 0.03789030522745878,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 162
    },
    {
      "epoch": 0.038124196000467785,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 163
    },
    {
      "epoch": 0.03835808677347679,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 164
    },
    {
      "epoch": 0.038591977546485794,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 165
    },
    {
      "epoch": 0.0388258683194948,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 166
    },
    {
      "epoch": 0.0390597590925038,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 167
    },
    {
      "epoch": 0.03929364986551281,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 168
    },
    {
      "epoch": 0.03952754063852181,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 169
    },
    {
      "epoch": 0.039761431411530816,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7644,
      "step": 170
    },
    {
      "epoch": 0.03999532218453982,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 171
    },
    {
      "epoch": 0.040229212957548825,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 172
    },
    {
      "epoch": 0.04046310373055783,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 173
    },
    {
      "epoch": 0.040696994503566834,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7273,
      "step": 174
    },
    {
      "epoch": 0.04093088527657584,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 175
    },
    {
      "epoch": 0.04116477604958484,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 176
    },
    {
      "epoch": 0.04139866682259385,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6956,
      "step": 177
    },
    {
      "epoch": 0.04163255759560285,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 178
    },
    {
      "epoch": 0.04186644836861186,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 179
    },
    {
      "epoch": 0.04210033914162086,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.712,
      "step": 180
    },
    {
      "epoch": 0.042334229914629866,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 181
    },
    {
      "epoch": 0.04256812068763887,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 182
    },
    {
      "epoch": 0.042802011460647875,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 183
    },
    {
      "epoch": 0.04303590223365688,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.5033,
      "step": 184
    },
    {
      "epoch": 0.043269793006665884,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5621,
      "step": 185
    },
    {
      "epoch": 0.04350368377967489,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 186
    },
    {
      "epoch": 0.0437375745526839,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 187
    },
    {
      "epoch": 0.043971465325692904,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3485,
      "step": 188
    },
    {
      "epoch": 0.04420535609870191,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 189
    },
    {
      "epoch": 0.044439246871710913,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8366,
      "step": 190
    },
    {
      "epoch": 0.04467313764471992,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 191
    },
    {
      "epoch": 0.04490702841772892,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 192
    },
    {
      "epoch": 0.04514091919073793,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 193
    },
    {
      "epoch": 0.04537480996374693,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 194
    },
    {
      "epoch": 0.045608700736755936,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 195
    },
    {
      "epoch": 0.04584259150976494,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 196
    },
    {
      "epoch": 0.046076482282773945,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6682,
      "step": 197
    },
    {
      "epoch": 0.04631037305578295,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 198
    },
    {
      "epoch": 0.046544263828791954,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 199
    },
    {
      "epoch": 0.04677815460180096,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7442,
      "step": 200
    },
    {
      "epoch": 0.04677815460180096,
      "eval_runtime": 4.6281,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 200
    },
    {
      "epoch": 0.04701204537480996,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 201
    },
    {
      "epoch": 0.04724593614781897,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 202
    },
    {
      "epoch": 0.04747982692082797,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 203
    },
    {
      "epoch": 0.04771371769383698,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 204
    },
    {
      "epoch": 0.04794760846684598,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 205
    },
    {
      "epoch": 0.048181499239854986,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 206
    },
    {
      "epoch": 0.04841539001286399,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 207
    },
    {
      "epoch": 0.048649280785872995,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 208
    },
    {
      "epoch": 0.048883171558882,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 209
    },
    {
      "epoch": 0.049117062331891004,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 210
    },
    {
      "epoch": 0.04935095310490001,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 211
    },
    {
      "epoch": 0.04958484387790902,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 212
    },
    {
      "epoch": 0.049818734650918024,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 213
    },
    {
      "epoch": 0.05005262542392703,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 214
    },
    {
      "epoch": 0.05028651619693603,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3411,
      "step": 215
    },
    {
      "epoch": 0.05052040696994504,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 216
    },
    {
      "epoch": 0.05075429774295404,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 217
    },
    {
      "epoch": 0.05098818851596305,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 218
    },
    {
      "epoch": 0.05122207928897205,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 219
    },
    {
      "epoch": 0.051455970061981056,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 220
    },
    {
      "epoch": 0.05168986083499006,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 221
    },
    {
      "epoch": 0.051923751607999065,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7443,
      "step": 222
    },
    {
      "epoch": 0.05215764238100807,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 223
    },
    {
      "epoch": 0.052391533154017074,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 224
    },
    {
      "epoch": 0.05262542392702608,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7694,
      "step": 225
    },
    {
      "epoch": 0.05285931470003508,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 226
    },
    {
      "epoch": 0.05309320547304409,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.467,
      "step": 227
    },
    {
      "epoch": 0.05332709624605309,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 228
    },
    {
      "epoch": 0.053560987019062097,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3283,
      "step": 229
    },
    {
      "epoch": 0.0537948777920711,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 230
    },
    {
      "epoch": 0.054028768565080106,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 231
    },
    {
      "epoch": 0.05426265933808911,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 232
    },
    {
      "epoch": 0.054496550111098115,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 233
    },
    {
      "epoch": 0.05473044088410712,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 234
    },
    {
      "epoch": 0.054964331657116124,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 235
    },
    {
      "epoch": 0.05519822243012513,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 236
    },
    {
      "epoch": 0.05543211320313414,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 237
    },
    {
      "epoch": 0.055666003976143144,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 238
    },
    {
      "epoch": 0.05589989474915215,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 239
    },
    {
      "epoch": 0.05613378552216115,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 240
    },
    {
      "epoch": 0.05636767629517016,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5982,
      "step": 241
    },
    {
      "epoch": 0.05660156706817916,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 242
    },
    {
      "epoch": 0.05683545784118817,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9924,
      "step": 243
    },
    {
      "epoch": 0.05706934861419717,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 244
    },
    {
      "epoch": 0.057303239387206176,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 245
    },
    {
      "epoch": 0.05753713016021518,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 246
    },
    {
      "epoch": 0.057771020933224185,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 247
    },
    {
      "epoch": 0.05800491170623319,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3657,
      "step": 248
    },
    {
      "epoch": 0.058238802479242194,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2889,
      "step": 249
    },
    {
      "epoch": 0.0584726932522512,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 250
    },
    {
      "epoch": 0.0587065840252602,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 251
    },
    {
      "epoch": 0.05894047479826921,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 252
    },
    {
      "epoch": 0.05917436557127821,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 253
    },
    {
      "epoch": 0.059408256344287216,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 254
    },
    {
      "epoch": 0.05964214711729622,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9241,
      "step": 255
    },
    {
      "epoch": 0.059876037890305225,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 256
    },
    {
      "epoch": 0.06010992866331423,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 257
    },
    {
      "epoch": 0.060343819436323234,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 258
    },
    {
      "epoch": 0.06057771020933224,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 259
    },
    {
      "epoch": 0.06081160098234124,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.554,
      "step": 260
    },
    {
      "epoch": 0.06104549175535025,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 261
    },
    {
      "epoch": 0.06127938252835926,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 262
    },
    {
      "epoch": 0.061513273301368264,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 263
    },
    {
      "epoch": 0.06174716407437727,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 264
    },
    {
      "epoch": 0.06198105484738627,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 265
    },
    {
      "epoch": 0.06221494562039528,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 266
    },
    {
      "epoch": 0.06244883639340428,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 267
    },
    {
      "epoch": 0.06268272716641328,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3962,
      "step": 268
    },
    {
      "epoch": 0.06291661793942228,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3128,
      "step": 269
    },
    {
      "epoch": 0.06315050871243129,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 270
    },
    {
      "epoch": 0.0633843994854403,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 271
    },
    {
      "epoch": 0.0636182902584493,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 272
    },
    {
      "epoch": 0.0638521810314583,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 273
    },
    {
      "epoch": 0.06408607180446732,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 274
    },
    {
      "epoch": 0.06431996257747633,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 275
    },
    {
      "epoch": 0.06455385335048533,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 276
    },
    {
      "epoch": 0.06478774412349433,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 277
    },
    {
      "epoch": 0.06502163489650334,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1778,
      "step": 278
    },
    {
      "epoch": 0.06525552566951234,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4982,
      "step": 279
    },
    {
      "epoch": 0.06548941644252135,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 280
    },
    {
      "epoch": 0.06572330721553035,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 281
    },
    {
      "epoch": 0.06595719798853936,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 282
    },
    {
      "epoch": 0.06619108876154836,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6541,
      "step": 283
    },
    {
      "epoch": 0.06642497953455737,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 284
    },
    {
      "epoch": 0.06665887030756637,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6966,
      "step": 285
    },
    {
      "epoch": 0.06689276108057537,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 286
    },
    {
      "epoch": 0.06712665185358438,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 287
    },
    {
      "epoch": 0.06736054262659338,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 288
    },
    {
      "epoch": 0.06759443339960239,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 289
    },
    {
      "epoch": 0.06782832417261139,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 290
    },
    {
      "epoch": 0.0680622149456204,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 291
    },
    {
      "epoch": 0.0682961057186294,
      "grad_norm": 2.546875,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 292
    },
    {
      "epoch": 0.0685299964916384,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2638,
      "step": 293
    },
    {
      "epoch": 0.06876388726464741,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 294
    },
    {
      "epoch": 0.06899777803765642,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 295
    },
    {
      "epoch": 0.06923166881066542,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 296
    },
    {
      "epoch": 0.06946555958367442,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 297
    },
    {
      "epoch": 0.06969945035668343,
      "grad_norm": 12.125,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 298
    },
    {
      "epoch": 0.06993334112969243,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 299
    },
    {
      "epoch": 0.07016723190270144,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 300
    },
    {
      "epoch": 0.07016723190270144,
      "eval_runtime": 4.6435,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 300
    },
    {
      "epoch": 0.07040112267571044,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 301
    },
    {
      "epoch": 0.07063501344871945,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 302
    },
    {
      "epoch": 0.07086890422172845,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 303
    },
    {
      "epoch": 0.07110279499473746,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 304
    },
    {
      "epoch": 0.07133668576774646,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 305
    },
    {
      "epoch": 0.07157057654075547,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 306
    },
    {
      "epoch": 0.07180446731376447,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 307
    },
    {
      "epoch": 0.07203835808677347,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 308
    },
    {
      "epoch": 0.07227224885978248,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 309
    },
    {
      "epoch": 0.07250613963279148,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 310
    },
    {
      "epoch": 0.07274003040580049,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 311
    },
    {
      "epoch": 0.07297392117880949,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 312
    },
    {
      "epoch": 0.0732078119518185,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.951,
      "step": 313
    },
    {
      "epoch": 0.0734417027248275,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 314
    },
    {
      "epoch": 0.0736755934978365,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 315
    },
    {
      "epoch": 0.07390948427084551,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 316
    },
    {
      "epoch": 0.07414337504385451,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 317
    },
    {
      "epoch": 0.07437726581686352,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 318
    },
    {
      "epoch": 0.07461115658987252,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 319
    },
    {
      "epoch": 0.07484504736288153,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9556,
      "step": 320
    },
    {
      "epoch": 0.07507893813589053,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 321
    },
    {
      "epoch": 0.07531282890889954,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6288,
      "step": 322
    },
    {
      "epoch": 0.07554671968190854,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 323
    },
    {
      "epoch": 0.07578061045491756,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.32,
      "step": 324
    },
    {
      "epoch": 0.07601450122792656,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2342,
      "step": 325
    },
    {
      "epoch": 0.07624839200093557,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.393,
      "step": 326
    },
    {
      "epoch": 0.07648228277394457,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 327
    },
    {
      "epoch": 0.07671617354695358,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 328
    },
    {
      "epoch": 0.07695006431996258,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2067,
      "step": 329
    },
    {
      "epoch": 0.07718395509297159,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 330
    },
    {
      "epoch": 0.07741784586598059,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 331
    },
    {
      "epoch": 0.0776517366389896,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 332
    },
    {
      "epoch": 0.0778856274119986,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 333
    },
    {
      "epoch": 0.0781195181850076,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 334
    },
    {
      "epoch": 0.07835340895801661,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7802,
      "step": 335
    },
    {
      "epoch": 0.07858729973102561,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 336
    },
    {
      "epoch": 0.07882119050403462,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5946,
      "step": 337
    },
    {
      "epoch": 0.07905508127704362,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 338
    },
    {
      "epoch": 0.07928897205005263,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 339
    },
    {
      "epoch": 0.07952286282306163,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 340
    },
    {
      "epoch": 0.07975675359607064,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2952,
      "step": 341
    },
    {
      "epoch": 0.07999064436907964,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 342
    },
    {
      "epoch": 0.08022453514208865,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 343
    },
    {
      "epoch": 0.08045842591509765,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6948,
      "step": 344
    },
    {
      "epoch": 0.08069231668810666,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 345
    },
    {
      "epoch": 0.08092620746111566,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7845,
      "step": 346
    },
    {
      "epoch": 0.08116009823412466,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 347
    },
    {
      "epoch": 0.08139398900713367,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7056,
      "step": 348
    },
    {
      "epoch": 0.08162787978014267,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 349
    },
    {
      "epoch": 0.08186177055315168,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 350
    },
    {
      "epoch": 0.08209566132616068,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 351
    },
    {
      "epoch": 0.08232955209916969,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.346,
      "step": 352
    },
    {
      "epoch": 0.08256344287217869,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 353
    },
    {
      "epoch": 0.0827973336451877,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 354
    },
    {
      "epoch": 0.0830312244181967,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 355
    },
    {
      "epoch": 0.0832651151912057,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 356
    },
    {
      "epoch": 0.08349900596421471,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 357
    },
    {
      "epoch": 0.08373289673722371,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 358
    },
    {
      "epoch": 0.08396678751023272,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 359
    },
    {
      "epoch": 0.08420067828324172,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 360
    },
    {
      "epoch": 0.08443456905625073,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 361
    },
    {
      "epoch": 0.08466845982925973,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 362
    },
    {
      "epoch": 0.08490235060226874,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 363
    },
    {
      "epoch": 0.08513624137527774,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 364
    },
    {
      "epoch": 0.08537013214828675,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 365
    },
    {
      "epoch": 0.08560402292129575,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 366
    },
    {
      "epoch": 0.08583791369430475,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2668,
      "step": 367
    },
    {
      "epoch": 0.08607180446731376,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 368
    },
    {
      "epoch": 0.08630569524032276,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 369
    },
    {
      "epoch": 0.08653958601333177,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 370
    },
    {
      "epoch": 0.08677347678634077,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 371
    },
    {
      "epoch": 0.08700736755934978,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5816,
      "step": 372
    },
    {
      "epoch": 0.0872412583323588,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 373
    },
    {
      "epoch": 0.0874751491053678,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 374
    },
    {
      "epoch": 0.0877090398783768,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 375
    },
    {
      "epoch": 0.08794293065138581,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.779,
      "step": 376
    },
    {
      "epoch": 0.08817682142439481,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 377
    },
    {
      "epoch": 0.08841071219740382,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 378
    },
    {
      "epoch": 0.08864460297041282,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 379
    },
    {
      "epoch": 0.08887849374342183,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 380
    },
    {
      "epoch": 0.08911238451643083,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.592,
      "step": 381
    },
    {
      "epoch": 0.08934627528943984,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 382
    },
    {
      "epoch": 0.08958016606244884,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 383
    },
    {
      "epoch": 0.08981405683545785,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8549,
      "step": 384
    },
    {
      "epoch": 0.09004794760846685,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 385
    },
    {
      "epoch": 0.09028183838147585,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 386
    },
    {
      "epoch": 0.09051572915448486,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 387
    },
    {
      "epoch": 0.09074961992749386,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 388
    },
    {
      "epoch": 0.09098351070050287,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 389
    },
    {
      "epoch": 0.09121740147351187,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 390
    },
    {
      "epoch": 0.09145129224652088,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 391
    },
    {
      "epoch": 0.09168518301952988,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 392
    },
    {
      "epoch": 0.09191907379253889,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8473,
      "step": 393
    },
    {
      "epoch": 0.09215296456554789,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 394
    },
    {
      "epoch": 0.0923868553385569,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 395
    },
    {
      "epoch": 0.0926207461115659,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 396
    },
    {
      "epoch": 0.0928546368845749,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 397
    },
    {
      "epoch": 0.09308852765758391,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 398
    },
    {
      "epoch": 0.09332241843059291,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 399
    },
    {
      "epoch": 0.09355630920360192,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 400
    },
    {
      "epoch": 0.09355630920360192,
      "eval_runtime": 4.6545,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 400
    },
    {
      "epoch": 0.09379019997661092,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8547,
      "step": 401
    },
    {
      "epoch": 0.09402409074961993,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 402
    },
    {
      "epoch": 0.09425798152262893,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 403
    },
    {
      "epoch": 0.09449187229563794,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 404
    },
    {
      "epoch": 0.09472576306864694,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 405
    },
    {
      "epoch": 0.09495965384165594,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 406
    },
    {
      "epoch": 0.09519354461466495,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 407
    },
    {
      "epoch": 0.09542743538767395,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 408
    },
    {
      "epoch": 0.09566132616068296,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 409
    },
    {
      "epoch": 0.09589521693369196,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 410
    },
    {
      "epoch": 0.09612910770670097,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 411
    },
    {
      "epoch": 0.09636299847970997,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 412
    },
    {
      "epoch": 0.09659688925271898,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 413
    },
    {
      "epoch": 0.09683078002572798,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 414
    },
    {
      "epoch": 0.09706467079873699,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 415
    },
    {
      "epoch": 0.09729856157174599,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 416
    },
    {
      "epoch": 0.097532452344755,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 417
    },
    {
      "epoch": 0.097766343117764,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 418
    },
    {
      "epoch": 0.098000233890773,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 419
    },
    {
      "epoch": 0.09823412466378201,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 420
    },
    {
      "epoch": 0.09846801543679101,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 421
    },
    {
      "epoch": 0.09870190620980002,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2917,
      "step": 422
    },
    {
      "epoch": 0.09893579698280904,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 423
    },
    {
      "epoch": 0.09916968775581804,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 424
    },
    {
      "epoch": 0.09940357852882704,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 425
    },
    {
      "epoch": 0.09963746930183605,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 426
    },
    {
      "epoch": 0.09987136007484505,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 427
    },
    {
      "epoch": 0.10010525084785406,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 428
    },
    {
      "epoch": 0.10033914162086306,
      "grad_norm": 7.5,
      "learning_rate": 3e-05,
      "loss": 2.5819,
      "step": 429
    },
    {
      "epoch": 0.10057303239387207,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 430
    },
    {
      "epoch": 0.10080692316688107,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 431
    },
    {
      "epoch": 0.10104081393989008,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 432
    },
    {
      "epoch": 0.10127470471289908,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 433
    },
    {
      "epoch": 0.10150859548590808,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 434
    },
    {
      "epoch": 0.10174248625891709,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 435
    },
    {
      "epoch": 0.1019763770319261,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 436
    },
    {
      "epoch": 0.1022102678049351,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.5849,
      "step": 437
    },
    {
      "epoch": 0.1024441585779441,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7876,
      "step": 438
    },
    {
      "epoch": 0.10267804935095311,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2125,
      "step": 439
    },
    {
      "epoch": 0.10291194012396211,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 440
    },
    {
      "epoch": 0.10314583089697112,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 441
    },
    {
      "epoch": 0.10337972166998012,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 442
    },
    {
      "epoch": 0.10361361244298913,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 443
    },
    {
      "epoch": 0.10384750321599813,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 444
    },
    {
      "epoch": 0.10408139398900713,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 445
    },
    {
      "epoch": 0.10431528476201614,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 446
    },
    {
      "epoch": 0.10454917553502514,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 447
    },
    {
      "epoch": 0.10478306630803415,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 448
    },
    {
      "epoch": 0.10501695708104315,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 449
    },
    {
      "epoch": 0.10525084785405216,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.77,
      "step": 450
    },
    {
      "epoch": 0.10548473862706116,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 451
    },
    {
      "epoch": 0.10571862940007017,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 452
    },
    {
      "epoch": 0.10595252017307917,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 453
    },
    {
      "epoch": 0.10618641094608817,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7463,
      "step": 454
    },
    {
      "epoch": 0.10642030171909718,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 455
    },
    {
      "epoch": 0.10665419249210618,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 456
    },
    {
      "epoch": 0.10688808326511519,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 457
    },
    {
      "epoch": 0.10712197403812419,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 458
    },
    {
      "epoch": 0.1073558648111332,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 459
    },
    {
      "epoch": 0.1075897555841422,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2729,
      "step": 460
    },
    {
      "epoch": 0.1078236463571512,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 461
    },
    {
      "epoch": 0.10805753713016021,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 462
    },
    {
      "epoch": 0.10829142790316922,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 463
    },
    {
      "epoch": 0.10852531867617822,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 464
    },
    {
      "epoch": 0.10875920944918722,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 465
    },
    {
      "epoch": 0.10899310022219623,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1878,
      "step": 466
    },
    {
      "epoch": 0.10922699099520523,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 467
    },
    {
      "epoch": 0.10946088176821424,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 468
    },
    {
      "epoch": 0.10969477254122324,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 469
    },
    {
      "epoch": 0.10992866331423225,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 470
    },
    {
      "epoch": 0.11016255408724125,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 471
    },
    {
      "epoch": 0.11039644486025026,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 472
    },
    {
      "epoch": 0.11063033563325927,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 473
    },
    {
      "epoch": 0.11086422640626828,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 474
    },
    {
      "epoch": 0.11109811717927728,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 475
    },
    {
      "epoch": 0.11133200795228629,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 476
    },
    {
      "epoch": 0.11156589872529529,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3432,
      "step": 477
    },
    {
      "epoch": 0.1117997894983043,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9811,
      "step": 478
    },
    {
      "epoch": 0.1120336802713133,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 479
    },
    {
      "epoch": 0.1122675710443223,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 480
    },
    {
      "epoch": 0.11250146181733131,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 481
    },
    {
      "epoch": 0.11273535259034032,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 482
    },
    {
      "epoch": 0.11296924336334932,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 483
    },
    {
      "epoch": 0.11320313413635832,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3864,
      "step": 484
    },
    {
      "epoch": 0.11343702490936733,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 485
    },
    {
      "epoch": 0.11367091568237633,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 486
    },
    {
      "epoch": 0.11390480645538534,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 487
    },
    {
      "epoch": 0.11413869722839434,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 488
    },
    {
      "epoch": 0.11437258800140335,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 489
    },
    {
      "epoch": 0.11460647877441235,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 490
    },
    {
      "epoch": 0.11484036954742136,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 491
    },
    {
      "epoch": 0.11507426032043036,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 492
    },
    {
      "epoch": 0.11530815109343936,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 493
    },
    {
      "epoch": 0.11554204186644837,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 494
    },
    {
      "epoch": 0.11577593263945737,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 495
    },
    {
      "epoch": 0.11600982341246638,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 496
    },
    {
      "epoch": 0.11624371418547538,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 497
    },
    {
      "epoch": 0.11647760495848439,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 498
    },
    {
      "epoch": 0.11671149573149339,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 499
    },
    {
      "epoch": 0.1169453865045024,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7946,
      "step": 500
    },
    {
      "epoch": 0.1169453865045024,
      "eval_runtime": 4.5882,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 500
    },
    {
      "epoch": 0.1171792772775114,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 501
    },
    {
      "epoch": 0.1174131680505204,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 502
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 503
    },
    {
      "epoch": 0.11788094959653841,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 504
    },
    {
      "epoch": 0.11811484036954742,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 505
    },
    {
      "epoch": 0.11834873114255642,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 506
    },
    {
      "epoch": 0.11858262191556543,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 507
    },
    {
      "epoch": 0.11881651268857443,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 508
    },
    {
      "epoch": 0.11905040346158344,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 509
    },
    {
      "epoch": 0.11928429423459244,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 510
    },
    {
      "epoch": 0.11951818500760145,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 511
    },
    {
      "epoch": 0.11975207578061045,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 512
    },
    {
      "epoch": 0.11998596655361946,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 513
    },
    {
      "epoch": 0.12021985732662846,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 514
    },
    {
      "epoch": 0.12045374809963746,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 515
    },
    {
      "epoch": 0.12068763887264647,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.329,
      "step": 516
    },
    {
      "epoch": 0.12092152964565547,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7426,
      "step": 517
    },
    {
      "epoch": 0.12115542041866448,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 518
    },
    {
      "epoch": 0.12138931119167348,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6988,
      "step": 519
    },
    {
      "epoch": 0.12162320196468249,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 520
    },
    {
      "epoch": 0.12185709273769149,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8721,
      "step": 521
    },
    {
      "epoch": 0.1220909835107005,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 522
    },
    {
      "epoch": 0.12232487428370951,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 523
    },
    {
      "epoch": 0.12255876505671852,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 524
    },
    {
      "epoch": 0.12279265582972752,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 525
    },
    {
      "epoch": 0.12302654660273653,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 526
    },
    {
      "epoch": 0.12326043737574553,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 527
    },
    {
      "epoch": 0.12349432814875454,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 528
    },
    {
      "epoch": 0.12372821892176354,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 529
    },
    {
      "epoch": 0.12396210969477255,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 530
    },
    {
      "epoch": 0.12419600046778155,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6194,
      "step": 531
    },
    {
      "epoch": 0.12442989124079055,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 532
    },
    {
      "epoch": 0.12466378201379956,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 533
    },
    {
      "epoch": 0.12489767278680856,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0806,
      "step": 534
    },
    {
      "epoch": 0.12513156355981755,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 535
    },
    {
      "epoch": 0.12536545433282656,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 536
    },
    {
      "epoch": 0.12559934510583556,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 537
    },
    {
      "epoch": 0.12583323587884457,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 538
    },
    {
      "epoch": 0.12606712665185357,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3572,
      "step": 539
    },
    {
      "epoch": 0.12630101742486258,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 540
    },
    {
      "epoch": 0.12653490819787158,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 541
    },
    {
      "epoch": 0.1267687989708806,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2439,
      "step": 542
    },
    {
      "epoch": 0.1270026897438896,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6197,
      "step": 543
    },
    {
      "epoch": 0.1272365805168986,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5535,
      "step": 544
    },
    {
      "epoch": 0.1274704712899076,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.901,
      "step": 545
    },
    {
      "epoch": 0.1277043620629166,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 546
    },
    {
      "epoch": 0.12793825283592564,
      "grad_norm": 10.1875,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 547
    },
    {
      "epoch": 0.12817214360893464,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 548
    },
    {
      "epoch": 0.12840603438194365,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 549
    },
    {
      "epoch": 0.12863992515495265,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 550
    },
    {
      "epoch": 0.12887381592796165,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 551
    },
    {
      "epoch": 0.12910770670097066,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 552
    },
    {
      "epoch": 0.12934159747397966,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 553
    },
    {
      "epoch": 0.12957548824698867,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6746,
      "step": 554
    },
    {
      "epoch": 0.12980937901999767,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 555
    },
    {
      "epoch": 0.13004326979300668,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1779,
      "step": 556
    },
    {
      "epoch": 0.13027716056601568,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 557
    },
    {
      "epoch": 0.1305110513390247,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 558
    },
    {
      "epoch": 0.1307449421120337,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 559
    },
    {
      "epoch": 0.1309788328850427,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6156,
      "step": 560
    },
    {
      "epoch": 0.1312127236580517,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6768,
      "step": 561
    },
    {
      "epoch": 0.1314466144310607,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6804,
      "step": 562
    },
    {
      "epoch": 0.1316805052040697,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 563
    },
    {
      "epoch": 0.1319143959770787,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 564
    },
    {
      "epoch": 0.13214828675008772,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 565
    },
    {
      "epoch": 0.13238217752309672,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 566
    },
    {
      "epoch": 0.13261606829610573,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.301,
      "step": 567
    },
    {
      "epoch": 0.13284995906911473,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 568
    },
    {
      "epoch": 0.13308384984212374,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7988,
      "step": 569
    },
    {
      "epoch": 0.13331774061513274,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 570
    },
    {
      "epoch": 0.13355163138814174,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 571
    },
    {
      "epoch": 0.13378552216115075,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 572
    },
    {
      "epoch": 0.13401941293415975,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 573
    },
    {
      "epoch": 0.13425330370716876,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 574
    },
    {
      "epoch": 0.13448719448017776,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 575
    },
    {
      "epoch": 0.13472108525318677,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 576
    },
    {
      "epoch": 0.13495497602619577,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.5458,
      "step": 577
    },
    {
      "epoch": 0.13518886679920478,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9751,
      "step": 578
    },
    {
      "epoch": 0.13542275757221378,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7443,
      "step": 579
    },
    {
      "epoch": 0.13565664834522279,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 580
    },
    {
      "epoch": 0.1358905391182318,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 581
    },
    {
      "epoch": 0.1361244298912408,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 582
    },
    {
      "epoch": 0.1363583206642498,
      "grad_norm": 9.375,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 583
    },
    {
      "epoch": 0.1365922114372588,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6651,
      "step": 584
    },
    {
      "epoch": 0.1368261022102678,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 585
    },
    {
      "epoch": 0.1370599929832768,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 586
    },
    {
      "epoch": 0.13729388375628582,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0676,
      "step": 587
    },
    {
      "epoch": 0.13752777452929482,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 588
    },
    {
      "epoch": 0.13776166530230383,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 589
    },
    {
      "epoch": 0.13799555607531283,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 590
    },
    {
      "epoch": 0.13822944684832184,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.634,
      "step": 591
    },
    {
      "epoch": 0.13846333762133084,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 592
    },
    {
      "epoch": 0.13869722839433984,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 593
    },
    {
      "epoch": 0.13893111916734885,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 594
    },
    {
      "epoch": 0.13916500994035785,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.255,
      "step": 595
    },
    {
      "epoch": 0.13939890071336686,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 596
    },
    {
      "epoch": 0.13963279148637586,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.782,
      "step": 597
    },
    {
      "epoch": 0.13986668225938487,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 598
    },
    {
      "epoch": 0.14010057303239387,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 599
    },
    {
      "epoch": 0.14033446380540288,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 600
    },
    {
      "epoch": 0.14033446380540288,
      "eval_runtime": 4.6426,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 600
    },
    {
      "epoch": 0.14056835457841188,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 601
    },
    {
      "epoch": 0.14080224535142088,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.6545,
      "step": 602
    },
    {
      "epoch": 0.1410361361244299,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2753,
      "step": 603
    },
    {
      "epoch": 0.1412700268974389,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 604
    },
    {
      "epoch": 0.1415039176704479,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 605
    },
    {
      "epoch": 0.1417378084434569,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 606
    },
    {
      "epoch": 0.1419716992164659,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7543,
      "step": 607
    },
    {
      "epoch": 0.1422055899894749,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 608
    },
    {
      "epoch": 0.14243948076248392,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 609
    },
    {
      "epoch": 0.14267337153549292,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 610
    },
    {
      "epoch": 0.14290726230850193,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2328,
      "step": 611
    },
    {
      "epoch": 0.14314115308151093,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 612
    },
    {
      "epoch": 0.14337504385451993,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 613
    },
    {
      "epoch": 0.14360893462752894,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 614
    },
    {
      "epoch": 0.14384282540053794,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 615
    },
    {
      "epoch": 0.14407671617354695,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 616
    },
    {
      "epoch": 0.14431060694655595,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 617
    },
    {
      "epoch": 0.14454449771956496,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 618
    },
    {
      "epoch": 0.14477838849257396,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6541,
      "step": 619
    },
    {
      "epoch": 0.14501227926558297,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 620
    },
    {
      "epoch": 0.14524617003859197,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 621
    },
    {
      "epoch": 0.14548006081160098,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 622
    },
    {
      "epoch": 0.14571395158460998,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 623
    },
    {
      "epoch": 0.14594784235761898,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2033,
      "step": 624
    },
    {
      "epoch": 0.146181733130628,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 625
    },
    {
      "epoch": 0.146415623903637,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 626
    },
    {
      "epoch": 0.146649514676646,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 627
    },
    {
      "epoch": 0.146883405449655,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 628
    },
    {
      "epoch": 0.147117296222664,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3582,
      "step": 629
    },
    {
      "epoch": 0.147351186995673,
      "grad_norm": 9.125,
      "learning_rate": 3e-05,
      "loss": 2.2052,
      "step": 630
    },
    {
      "epoch": 0.14758507776868202,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 631
    },
    {
      "epoch": 0.14781896854169102,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 632
    },
    {
      "epoch": 0.14805285931470002,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 633
    },
    {
      "epoch": 0.14828675008770903,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 634
    },
    {
      "epoch": 0.14852064086071803,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 635
    },
    {
      "epoch": 0.14875453163372704,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 636
    },
    {
      "epoch": 0.14898842240673604,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6393,
      "step": 637
    },
    {
      "epoch": 0.14922231317974505,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 638
    },
    {
      "epoch": 0.14945620395275405,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 639
    },
    {
      "epoch": 0.14969009472576306,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8528,
      "step": 640
    },
    {
      "epoch": 0.14992398549877206,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 641
    },
    {
      "epoch": 0.15015787627178107,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 642
    },
    {
      "epoch": 0.15039176704479007,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 643
    },
    {
      "epoch": 0.15062565781779907,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7081,
      "step": 644
    },
    {
      "epoch": 0.15085954859080808,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 645
    },
    {
      "epoch": 0.15109343936381708,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 646
    },
    {
      "epoch": 0.15132733013682612,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 647
    },
    {
      "epoch": 0.15156122090983512,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 648
    },
    {
      "epoch": 0.15179511168284412,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 649
    },
    {
      "epoch": 0.15202900245585313,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 650
    },
    {
      "epoch": 0.15226289322886213,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 651
    },
    {
      "epoch": 0.15249678400187114,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 652
    },
    {
      "epoch": 0.15273067477488014,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 653
    },
    {
      "epoch": 0.15296456554788915,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 654
    },
    {
      "epoch": 0.15319845632089815,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5994,
      "step": 655
    },
    {
      "epoch": 0.15343234709390716,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 656
    },
    {
      "epoch": 0.15366623786691616,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 657
    },
    {
      "epoch": 0.15390012863992517,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4731,
      "step": 658
    },
    {
      "epoch": 0.15413401941293417,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 659
    },
    {
      "epoch": 0.15436791018594317,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2089,
      "step": 660
    },
    {
      "epoch": 0.15460180095895218,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 661
    },
    {
      "epoch": 0.15483569173196118,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 662
    },
    {
      "epoch": 0.1550695825049702,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 663
    },
    {
      "epoch": 0.1553034732779792,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 664
    },
    {
      "epoch": 0.1555373640509882,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7614,
      "step": 665
    },
    {
      "epoch": 0.1557712548239972,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 666
    },
    {
      "epoch": 0.1560051455970062,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 667
    },
    {
      "epoch": 0.1562390363700152,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 668
    },
    {
      "epoch": 0.15647292714302422,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.2953,
      "step": 669
    },
    {
      "epoch": 0.15670681791603322,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 670
    },
    {
      "epoch": 0.15694070868904222,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 671
    },
    {
      "epoch": 0.15717459946205123,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 672
    },
    {
      "epoch": 0.15740849023506023,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 673
    },
    {
      "epoch": 0.15764238100806924,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 674
    },
    {
      "epoch": 0.15787627178107824,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 675
    },
    {
      "epoch": 0.15811016255408725,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 676
    },
    {
      "epoch": 0.15834405332709625,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7261,
      "step": 677
    },
    {
      "epoch": 0.15857794410010526,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 678
    },
    {
      "epoch": 0.15881183487311426,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 679
    },
    {
      "epoch": 0.15904572564612326,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 680
    },
    {
      "epoch": 0.15927961641913227,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 681
    },
    {
      "epoch": 0.15951350719214127,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1401,
      "step": 682
    },
    {
      "epoch": 0.15974739796515028,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.784,
      "step": 683
    },
    {
      "epoch": 0.15998128873815928,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 684
    },
    {
      "epoch": 0.1602151795111683,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2886,
      "step": 685
    },
    {
      "epoch": 0.1604490702841773,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 686
    },
    {
      "epoch": 0.1606829610571863,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 687
    },
    {
      "epoch": 0.1609168518301953,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 688
    },
    {
      "epoch": 0.1611507426032043,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 689
    },
    {
      "epoch": 0.1613846333762133,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 690
    },
    {
      "epoch": 0.16161852414922231,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 691
    },
    {
      "epoch": 0.16185241492223132,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 692
    },
    {
      "epoch": 0.16208630569524032,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 693
    },
    {
      "epoch": 0.16232019646824933,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 694
    },
    {
      "epoch": 0.16255408724125833,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 695
    },
    {
      "epoch": 0.16278797801426734,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 696
    },
    {
      "epoch": 0.16302186878727634,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 697
    },
    {
      "epoch": 0.16325575956028535,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7855,
      "step": 698
    },
    {
      "epoch": 0.16348965033329435,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 699
    },
    {
      "epoch": 0.16372354110630336,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 700
    },
    {
      "epoch": 0.16372354110630336,
      "eval_runtime": 4.6319,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 700
    },
    {
      "epoch": 0.16395743187931236,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7372,
      "step": 701
    },
    {
      "epoch": 0.16419132265232136,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 702
    },
    {
      "epoch": 0.16442521342533037,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 703
    },
    {
      "epoch": 0.16465910419833937,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 704
    },
    {
      "epoch": 0.16489299497134838,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 705
    },
    {
      "epoch": 0.16512688574435738,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 706
    },
    {
      "epoch": 0.1653607765173664,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2206,
      "step": 707
    },
    {
      "epoch": 0.1655946672903754,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 708
    },
    {
      "epoch": 0.1658285580633844,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 709
    },
    {
      "epoch": 0.1660624488363934,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 710
    },
    {
      "epoch": 0.1662963396094024,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 711
    },
    {
      "epoch": 0.1665302303824114,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 712
    },
    {
      "epoch": 0.16676412115542041,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 713
    },
    {
      "epoch": 0.16699801192842942,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 714
    },
    {
      "epoch": 0.16723190270143842,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6403,
      "step": 715
    },
    {
      "epoch": 0.16746579347444743,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 716
    },
    {
      "epoch": 0.16769968424745643,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 717
    },
    {
      "epoch": 0.16793357502046544,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2347,
      "step": 718
    },
    {
      "epoch": 0.16816746579347444,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 719
    },
    {
      "epoch": 0.16840135656648345,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 720
    },
    {
      "epoch": 0.16863524733949245,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 721
    },
    {
      "epoch": 0.16886913811250145,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 722
    },
    {
      "epoch": 0.16910302888551046,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 723
    },
    {
      "epoch": 0.16933691965851946,
      "grad_norm": 7.28125,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 724
    },
    {
      "epoch": 0.16957081043152847,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 725
    },
    {
      "epoch": 0.16980470120453747,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 726
    },
    {
      "epoch": 0.17003859197754648,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 727
    },
    {
      "epoch": 0.17027248275055548,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 728
    },
    {
      "epoch": 0.1705063735235645,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 729
    },
    {
      "epoch": 0.1707402642965735,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 730
    },
    {
      "epoch": 0.1709741550695825,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 731
    },
    {
      "epoch": 0.1712080458425915,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 732
    },
    {
      "epoch": 0.1714419366156005,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 733
    },
    {
      "epoch": 0.1716758273886095,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6946,
      "step": 734
    },
    {
      "epoch": 0.1719097181616185,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.813,
      "step": 735
    },
    {
      "epoch": 0.17214360893462752,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 736
    },
    {
      "epoch": 0.17237749970763652,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 737
    },
    {
      "epoch": 0.17261139048064553,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 738
    },
    {
      "epoch": 0.17284528125365453,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 739
    },
    {
      "epoch": 0.17307917202666354,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 740
    },
    {
      "epoch": 0.17331306279967254,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 741
    },
    {
      "epoch": 0.17354695357268154,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 742
    },
    {
      "epoch": 0.17378084434569055,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 743
    },
    {
      "epoch": 0.17401473511869955,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 744
    },
    {
      "epoch": 0.17424862589170856,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 745
    },
    {
      "epoch": 0.1744825166647176,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 746
    },
    {
      "epoch": 0.1747164074377266,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 747
    },
    {
      "epoch": 0.1749502982107356,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 748
    },
    {
      "epoch": 0.1751841889837446,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 749
    },
    {
      "epoch": 0.1754180797567536,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 750
    },
    {
      "epoch": 0.1756519705297626,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8719,
      "step": 751
    },
    {
      "epoch": 0.17588586130277162,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 752
    },
    {
      "epoch": 0.17611975207578062,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 753
    },
    {
      "epoch": 0.17635364284878963,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 754
    },
    {
      "epoch": 0.17658753362179863,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 755
    },
    {
      "epoch": 0.17682142439480764,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 756
    },
    {
      "epoch": 0.17705531516781664,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 757
    },
    {
      "epoch": 0.17728920594082564,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 758
    },
    {
      "epoch": 0.17752309671383465,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 759
    },
    {
      "epoch": 0.17775698748684365,
      "grad_norm": 8.9375,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 760
    },
    {
      "epoch": 0.17799087825985266,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 761
    },
    {
      "epoch": 0.17822476903286166,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 762
    },
    {
      "epoch": 0.17845865980587067,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 763
    },
    {
      "epoch": 0.17869255057887967,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3133,
      "step": 764
    },
    {
      "epoch": 0.17892644135188868,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 765
    },
    {
      "epoch": 0.17916033212489768,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 766
    },
    {
      "epoch": 0.17939422289790669,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 767
    },
    {
      "epoch": 0.1796281136709157,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 768
    },
    {
      "epoch": 0.1798620044439247,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 769
    },
    {
      "epoch": 0.1800958952169337,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 770
    },
    {
      "epoch": 0.1803297859899427,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 771
    },
    {
      "epoch": 0.1805636767629517,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6999,
      "step": 772
    },
    {
      "epoch": 0.1807975675359607,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 773
    },
    {
      "epoch": 0.18103145830896972,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 774
    },
    {
      "epoch": 0.18126534908197872,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 775
    },
    {
      "epoch": 0.18149923985498773,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 776
    },
    {
      "epoch": 0.18173313062799673,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 777
    },
    {
      "epoch": 0.18196702140100574,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 778
    },
    {
      "epoch": 0.18220091217401474,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 779
    },
    {
      "epoch": 0.18243480294702374,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 780
    },
    {
      "epoch": 0.18266869372003275,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 781
    },
    {
      "epoch": 0.18290258449304175,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5935,
      "step": 782
    },
    {
      "epoch": 0.18313647526605076,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 783
    },
    {
      "epoch": 0.18337036603905976,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 784
    },
    {
      "epoch": 0.18360425681206877,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 785
    },
    {
      "epoch": 0.18383814758507777,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 786
    },
    {
      "epoch": 0.18407203835808678,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8124,
      "step": 787
    },
    {
      "epoch": 0.18430592913109578,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7803,
      "step": 788
    },
    {
      "epoch": 0.18453981990410478,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5033,
      "step": 789
    },
    {
      "epoch": 0.1847737106771138,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 790
    },
    {
      "epoch": 0.1850076014501228,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 791
    },
    {
      "epoch": 0.1852414922231318,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 792
    },
    {
      "epoch": 0.1854753829961408,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 793
    },
    {
      "epoch": 0.1857092737691498,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 794
    },
    {
      "epoch": 0.1859431645421588,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 795
    },
    {
      "epoch": 0.18617705531516782,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 796
    },
    {
      "epoch": 0.18641094608817682,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2145,
      "step": 797
    },
    {
      "epoch": 0.18664483686118583,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 798
    },
    {
      "epoch": 0.18687872763419483,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 799
    },
    {
      "epoch": 0.18711261840720383,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 800
    },
    {
      "epoch": 0.18711261840720383,
      "eval_runtime": 4.6548,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 800
    },
    {
      "epoch": 0.18734650918021284,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 801
    },
    {
      "epoch": 0.18758039995322184,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 802
    },
    {
      "epoch": 0.18781429072623085,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 803
    },
    {
      "epoch": 0.18804818149923985,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 804
    },
    {
      "epoch": 0.18828207227224886,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 805
    },
    {
      "epoch": 0.18851596304525786,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 806
    },
    {
      "epoch": 0.18874985381826687,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 807
    },
    {
      "epoch": 0.18898374459127587,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 808
    },
    {
      "epoch": 0.18921763536428488,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 809
    },
    {
      "epoch": 0.18945152613729388,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.063,
      "step": 810
    },
    {
      "epoch": 0.18968541691030288,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.6009,
      "step": 811
    },
    {
      "epoch": 0.1899193076833119,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 812
    },
    {
      "epoch": 0.1901531984563209,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 813
    },
    {
      "epoch": 0.1903870892293299,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 814
    },
    {
      "epoch": 0.1906209800023389,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 815
    },
    {
      "epoch": 0.1908548707753479,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 816
    },
    {
      "epoch": 0.1910887615483569,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 817
    },
    {
      "epoch": 0.19132265232136592,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7662,
      "step": 818
    },
    {
      "epoch": 0.19155654309437492,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 819
    },
    {
      "epoch": 0.19179043386738392,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0836,
      "step": 820
    },
    {
      "epoch": 0.19202432464039293,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 821
    },
    {
      "epoch": 0.19225821541340193,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 822
    },
    {
      "epoch": 0.19249210618641094,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 823
    },
    {
      "epoch": 0.19272599695941994,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3857,
      "step": 824
    },
    {
      "epoch": 0.19295988773242895,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 825
    },
    {
      "epoch": 0.19319377850543795,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 826
    },
    {
      "epoch": 0.19342766927844696,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 827
    },
    {
      "epoch": 0.19366156005145596,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 828
    },
    {
      "epoch": 0.19389545082446497,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2472,
      "step": 829
    },
    {
      "epoch": 0.19412934159747397,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4121,
      "step": 830
    },
    {
      "epoch": 0.19436323237048297,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 831
    },
    {
      "epoch": 0.19459712314349198,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 832
    },
    {
      "epoch": 0.19483101391650098,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3995,
      "step": 833
    },
    {
      "epoch": 0.19506490468951,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 834
    },
    {
      "epoch": 0.195298795462519,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 835
    },
    {
      "epoch": 0.195532686235528,
      "grad_norm": 7.40625,
      "learning_rate": 3e-05,
      "loss": 2.6438,
      "step": 836
    },
    {
      "epoch": 0.195766577008537,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4571,
      "step": 837
    },
    {
      "epoch": 0.196000467781546,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 838
    },
    {
      "epoch": 0.196234358554555,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 839
    },
    {
      "epoch": 0.19646824932756402,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 840
    },
    {
      "epoch": 0.19670214010057302,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 841
    },
    {
      "epoch": 0.19693603087358202,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 842
    },
    {
      "epoch": 0.19716992164659103,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 843
    },
    {
      "epoch": 0.19740381241960003,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 844
    },
    {
      "epoch": 0.19763770319260904,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9461,
      "step": 845
    },
    {
      "epoch": 0.19787159396561807,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 846
    },
    {
      "epoch": 0.19810548473862707,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 847
    },
    {
      "epoch": 0.19833937551163608,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 848
    },
    {
      "epoch": 0.19857326628464508,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 849
    },
    {
      "epoch": 0.1988071570576541,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 850
    },
    {
      "epoch": 0.1990410478306631,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 851
    },
    {
      "epoch": 0.1992749386036721,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 852
    },
    {
      "epoch": 0.1995088293766811,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 853
    },
    {
      "epoch": 0.1997427201496901,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9255,
      "step": 854
    },
    {
      "epoch": 0.1999766109226991,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 855
    },
    {
      "epoch": 0.20021050169570812,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 856
    },
    {
      "epoch": 0.20044439246871712,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 857
    },
    {
      "epoch": 0.20067828324172612,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 858
    },
    {
      "epoch": 0.20091217401473513,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 859
    },
    {
      "epoch": 0.20114606478774413,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 860
    },
    {
      "epoch": 0.20137995556075314,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 861
    },
    {
      "epoch": 0.20161384633376214,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 862
    },
    {
      "epoch": 0.20184773710677115,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3815,
      "step": 863
    },
    {
      "epoch": 0.20208162787978015,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 864
    },
    {
      "epoch": 0.20231551865278916,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 865
    },
    {
      "epoch": 0.20254940942579816,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 866
    },
    {
      "epoch": 0.20278330019880716,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 867
    },
    {
      "epoch": 0.20301719097181617,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 868
    },
    {
      "epoch": 0.20325108174482517,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 869
    },
    {
      "epoch": 0.20348497251783418,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 870
    },
    {
      "epoch": 0.20371886329084318,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6109,
      "step": 871
    },
    {
      "epoch": 0.2039527540638522,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 872
    },
    {
      "epoch": 0.2041866448368612,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 873
    },
    {
      "epoch": 0.2044205356098702,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.3228,
      "step": 874
    },
    {
      "epoch": 0.2046544263828792,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 875
    },
    {
      "epoch": 0.2048883171558882,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3872,
      "step": 876
    },
    {
      "epoch": 0.2051222079288972,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3506,
      "step": 877
    },
    {
      "epoch": 0.20535609870190621,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 878
    },
    {
      "epoch": 0.20558998947491522,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 879
    },
    {
      "epoch": 0.20582388024792422,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 880
    },
    {
      "epoch": 0.20605777102093323,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5778,
      "step": 881
    },
    {
      "epoch": 0.20629166179394223,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 882
    },
    {
      "epoch": 0.20652555256695124,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6936,
      "step": 883
    },
    {
      "epoch": 0.20675944333996024,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 884
    },
    {
      "epoch": 0.20699333411296925,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.302,
      "step": 885
    },
    {
      "epoch": 0.20722722488597825,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2746,
      "step": 886
    },
    {
      "epoch": 0.20746111565898726,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 887
    },
    {
      "epoch": 0.20769500643199626,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 888
    },
    {
      "epoch": 0.20792889720500526,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 889
    },
    {
      "epoch": 0.20816278797801427,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 890
    },
    {
      "epoch": 0.20839667875102327,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 891
    },
    {
      "epoch": 0.20863056952403228,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 892
    },
    {
      "epoch": 0.20886446029704128,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 893
    },
    {
      "epoch": 0.2090983510700503,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 894
    },
    {
      "epoch": 0.2093322418430593,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.334,
      "step": 895
    },
    {
      "epoch": 0.2095661326160683,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 896
    },
    {
      "epoch": 0.2098000233890773,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 897
    },
    {
      "epoch": 0.2100339141620863,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 898
    },
    {
      "epoch": 0.2102678049350953,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 899
    },
    {
      "epoch": 0.2105016957081043,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 900
    },
    {
      "epoch": 0.2105016957081043,
      "eval_runtime": 4.5705,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 900
    },
    {
      "epoch": 0.21073558648111332,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 901
    },
    {
      "epoch": 0.21096947725412232,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 902
    },
    {
      "epoch": 0.21120336802713133,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 903
    },
    {
      "epoch": 0.21143725880014033,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 904
    },
    {
      "epoch": 0.21167114957314934,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 905
    },
    {
      "epoch": 0.21190504034615834,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 906
    },
    {
      "epoch": 0.21213893111916735,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 907
    },
    {
      "epoch": 0.21237282189217635,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 908
    },
    {
      "epoch": 0.21260671266518535,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 909
    },
    {
      "epoch": 0.21284060343819436,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 910
    },
    {
      "epoch": 0.21307449421120336,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 911
    },
    {
      "epoch": 0.21330838498421237,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 912
    },
    {
      "epoch": 0.21354227575722137,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 913
    },
    {
      "epoch": 0.21377616653023038,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 914
    },
    {
      "epoch": 0.21401005730323938,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 915
    },
    {
      "epoch": 0.21424394807624839,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 916
    },
    {
      "epoch": 0.2144778388492574,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 917
    },
    {
      "epoch": 0.2147117296222664,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 918
    },
    {
      "epoch": 0.2149456203952754,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 919
    },
    {
      "epoch": 0.2151795111682844,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 920
    },
    {
      "epoch": 0.2154134019412934,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 921
    },
    {
      "epoch": 0.2156472927143024,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 922
    },
    {
      "epoch": 0.21588118348731142,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 923
    },
    {
      "epoch": 0.21611507426032042,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 924
    },
    {
      "epoch": 0.21634896503332943,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 925
    },
    {
      "epoch": 0.21658285580633843,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 926
    },
    {
      "epoch": 0.21681674657934744,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 927
    },
    {
      "epoch": 0.21705063735235644,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3562,
      "step": 928
    },
    {
      "epoch": 0.21728452812536544,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 929
    },
    {
      "epoch": 0.21751841889837445,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.7644,
      "step": 930
    },
    {
      "epoch": 0.21775230967138345,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 931
    },
    {
      "epoch": 0.21798620044439246,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 932
    },
    {
      "epoch": 0.21822009121740146,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 933
    },
    {
      "epoch": 0.21845398199041047,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 934
    },
    {
      "epoch": 0.21868787276341947,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 935
    },
    {
      "epoch": 0.21892176353642848,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2797,
      "step": 936
    },
    {
      "epoch": 0.21915565430943748,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 937
    },
    {
      "epoch": 0.21938954508244649,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 938
    },
    {
      "epoch": 0.2196234358554555,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5462,
      "step": 939
    },
    {
      "epoch": 0.2198573266284645,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 940
    },
    {
      "epoch": 0.2200912174014735,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 941
    },
    {
      "epoch": 0.2203251081744825,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 942
    },
    {
      "epoch": 0.2205589989474915,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 943
    },
    {
      "epoch": 0.2207928897205005,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 944
    },
    {
      "epoch": 0.22102678049350954,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 945
    },
    {
      "epoch": 0.22126067126651855,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7207,
      "step": 946
    },
    {
      "epoch": 0.22149456203952755,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.72,
      "step": 947
    },
    {
      "epoch": 0.22172845281253656,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6304,
      "step": 948
    },
    {
      "epoch": 0.22196234358554556,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 949
    },
    {
      "epoch": 0.22219623435855457,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 950
    },
    {
      "epoch": 0.22243012513156357,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5565,
      "step": 951
    },
    {
      "epoch": 0.22266401590457258,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 952
    },
    {
      "epoch": 0.22289790667758158,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 953
    },
    {
      "epoch": 0.22313179745059059,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 954
    },
    {
      "epoch": 0.2233656882235996,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 955
    },
    {
      "epoch": 0.2235995789966086,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 956
    },
    {
      "epoch": 0.2238334697696176,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 957
    },
    {
      "epoch": 0.2240673605426266,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9513,
      "step": 958
    },
    {
      "epoch": 0.2243012513156356,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 959
    },
    {
      "epoch": 0.2245351420886446,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 960
    },
    {
      "epoch": 0.22476903286165362,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 961
    },
    {
      "epoch": 0.22500292363466262,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 962
    },
    {
      "epoch": 0.22523681440767163,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 963
    },
    {
      "epoch": 0.22547070518068063,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2851,
      "step": 964
    },
    {
      "epoch": 0.22570459595368964,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 965
    },
    {
      "epoch": 0.22593848672669864,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 966
    },
    {
      "epoch": 0.22617237749970764,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 967
    },
    {
      "epoch": 0.22640626827271665,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6892,
      "step": 968
    },
    {
      "epoch": 0.22664015904572565,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 969
    },
    {
      "epoch": 0.22687404981873466,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 970
    },
    {
      "epoch": 0.22710794059174366,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 971
    },
    {
      "epoch": 0.22734183136475267,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 972
    },
    {
      "epoch": 0.22757572213776167,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4189,
      "step": 973
    },
    {
      "epoch": 0.22780961291077068,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9081,
      "step": 974
    },
    {
      "epoch": 0.22804350368377968,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 975
    },
    {
      "epoch": 0.22827739445678868,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 976
    },
    {
      "epoch": 0.2285112852297977,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 977
    },
    {
      "epoch": 0.2287451760028067,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 978
    },
    {
      "epoch": 0.2289790667758157,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 979
    },
    {
      "epoch": 0.2292129575488247,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 980
    },
    {
      "epoch": 0.2294468483218337,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 981
    },
    {
      "epoch": 0.2296807390948427,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7453,
      "step": 982
    },
    {
      "epoch": 0.22991462986785172,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 983
    },
    {
      "epoch": 0.23014852064086072,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 984
    },
    {
      "epoch": 0.23038241141386973,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 985
    },
    {
      "epoch": 0.23061630218687873,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 986
    },
    {
      "epoch": 0.23085019295988773,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 987
    },
    {
      "epoch": 0.23108408373289674,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 988
    },
    {
      "epoch": 0.23131797450590574,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 989
    },
    {
      "epoch": 0.23155186527891475,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 990
    },
    {
      "epoch": 0.23178575605192375,
      "grad_norm": 16.75,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 991
    },
    {
      "epoch": 0.23201964682493276,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7327,
      "step": 992
    },
    {
      "epoch": 0.23225353759794176,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 993
    },
    {
      "epoch": 0.23248742837095077,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 994
    },
    {
      "epoch": 0.23272131914395977,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4578,
      "step": 995
    },
    {
      "epoch": 0.23295520991696878,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 996
    },
    {
      "epoch": 0.23318910068997778,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 997
    },
    {
      "epoch": 0.23342299146298678,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 998
    },
    {
      "epoch": 0.2336568822359958,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 999
    },
    {
      "epoch": 0.2338907730090048,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 1000
    },
    {
      "epoch": 0.2338907730090048,
      "eval_runtime": 4.6276,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 1000
    },
    {
      "epoch": 0.2341246637820138,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 1001
    },
    {
      "epoch": 0.2343585545550228,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6584,
      "step": 1002
    },
    {
      "epoch": 0.2345924453280318,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 1003
    },
    {
      "epoch": 0.2348263361010408,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 1004
    },
    {
      "epoch": 0.23506022687404982,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 1005
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 1006
    },
    {
      "epoch": 0.23552800842006782,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 1007
    },
    {
      "epoch": 0.23576189919307683,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.5113,
      "step": 1008
    },
    {
      "epoch": 0.23599578996608583,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.844,
      "step": 1009
    },
    {
      "epoch": 0.23622968073909484,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0649,
      "step": 1010
    },
    {
      "epoch": 0.23646357151210384,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 1011
    },
    {
      "epoch": 0.23669746228511285,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 1012
    },
    {
      "epoch": 0.23693135305812185,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 1013
    },
    {
      "epoch": 0.23716524383113086,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 1014
    },
    {
      "epoch": 0.23739913460413986,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 1015
    },
    {
      "epoch": 0.23763302537714887,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8289,
      "step": 1016
    },
    {
      "epoch": 0.23786691615015787,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.1937,
      "step": 1017
    },
    {
      "epoch": 0.23810080692316687,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 1018
    },
    {
      "epoch": 0.23833469769617588,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 1019
    },
    {
      "epoch": 0.23856858846918488,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 1020
    },
    {
      "epoch": 0.2388024792421939,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 1021
    },
    {
      "epoch": 0.2390363700152029,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6864,
      "step": 1022
    },
    {
      "epoch": 0.2392702607882119,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 1023
    },
    {
      "epoch": 0.2395041515612209,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 1024
    },
    {
      "epoch": 0.2397380423342299,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 1025
    },
    {
      "epoch": 0.2399719331072389,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 1026
    },
    {
      "epoch": 0.24020582388024792,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 1027
    },
    {
      "epoch": 0.24043971465325692,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 1028
    },
    {
      "epoch": 0.24067360542626592,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 1029
    },
    {
      "epoch": 0.24090749619927493,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 1030
    },
    {
      "epoch": 0.24114138697228393,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 1031
    },
    {
      "epoch": 0.24137527774529294,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 1032
    },
    {
      "epoch": 0.24160916851830194,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 1033
    },
    {
      "epoch": 0.24184305929131095,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7503,
      "step": 1034
    },
    {
      "epoch": 0.24207695006431995,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7923,
      "step": 1035
    },
    {
      "epoch": 0.24231084083732896,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 1036
    },
    {
      "epoch": 0.24254473161033796,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1996,
      "step": 1037
    },
    {
      "epoch": 0.24277862238334696,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 1038
    },
    {
      "epoch": 0.24301251315635597,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6945,
      "step": 1039
    },
    {
      "epoch": 0.24324640392936497,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 1040
    },
    {
      "epoch": 0.24348029470237398,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3801,
      "step": 1041
    },
    {
      "epoch": 0.24371418547538298,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 1042
    },
    {
      "epoch": 0.243948076248392,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 1043
    },
    {
      "epoch": 0.244181967021401,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2847,
      "step": 1044
    },
    {
      "epoch": 0.24441585779441002,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 1045
    },
    {
      "epoch": 0.24464974856741903,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 1046
    },
    {
      "epoch": 0.24488363934042803,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 1047
    },
    {
      "epoch": 0.24511753011343704,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 1048
    },
    {
      "epoch": 0.24535142088644604,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9937,
      "step": 1049
    },
    {
      "epoch": 0.24558531165945505,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 1050
    },
    {
      "epoch": 0.24581920243246405,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 1051
    },
    {
      "epoch": 0.24605309320547306,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 1052
    },
    {
      "epoch": 0.24628698397848206,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 1053
    },
    {
      "epoch": 0.24652087475149106,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 1054
    },
    {
      "epoch": 0.24675476552450007,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 1055
    },
    {
      "epoch": 0.24698865629750907,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 1056
    },
    {
      "epoch": 0.24722254707051808,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 1057
    },
    {
      "epoch": 0.24745643784352708,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 1058
    },
    {
      "epoch": 0.2476903286165361,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 1059
    },
    {
      "epoch": 0.2479242193895451,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 1060
    },
    {
      "epoch": 0.2481581101625541,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 1061
    },
    {
      "epoch": 0.2483920009355631,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 1062
    },
    {
      "epoch": 0.2486258917085721,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 1063
    },
    {
      "epoch": 0.2488597824815811,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6158,
      "step": 1064
    },
    {
      "epoch": 0.24909367325459011,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 1065
    },
    {
      "epoch": 0.24932756402759912,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 1066
    },
    {
      "epoch": 0.24956145480060812,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2132,
      "step": 1067
    },
    {
      "epoch": 0.24979534557361713,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 1068
    },
    {
      "epoch": 0.2500292363466261,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 1069
    },
    {
      "epoch": 0.2502631271196351,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 1070
    },
    {
      "epoch": 0.2504970178926441,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 1071
    },
    {
      "epoch": 0.2507309086656531,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 1072
    },
    {
      "epoch": 0.2509647994386621,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 1073
    },
    {
      "epoch": 0.2511986902116711,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.779,
      "step": 1074
    },
    {
      "epoch": 0.25143258098468013,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 1075
    },
    {
      "epoch": 0.25166647175768914,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.4055,
      "step": 1076
    },
    {
      "epoch": 0.25190036253069814,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 1077
    },
    {
      "epoch": 0.25213425330370715,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7572,
      "step": 1078
    },
    {
      "epoch": 0.25236814407671615,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 1079
    },
    {
      "epoch": 0.25260203484972515,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4043,
      "step": 1080
    },
    {
      "epoch": 0.25283592562273416,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 1081
    },
    {
      "epoch": 0.25306981639574316,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 1082
    },
    {
      "epoch": 0.25330370716875217,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 1083
    },
    {
      "epoch": 0.2535375979417612,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 1084
    },
    {
      "epoch": 0.2537714887147702,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 1085
    },
    {
      "epoch": 0.2540053794877792,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 1086
    },
    {
      "epoch": 0.2542392702607882,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 1087
    },
    {
      "epoch": 0.2544731610337972,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9103,
      "step": 1088
    },
    {
      "epoch": 0.2547070518068062,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 1089
    },
    {
      "epoch": 0.2549409425798152,
      "grad_norm": 10.6875,
      "learning_rate": 3e-05,
      "loss": 2.3201,
      "step": 1090
    },
    {
      "epoch": 0.2551748333528242,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 1091
    },
    {
      "epoch": 0.2554087241258332,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3069,
      "step": 1092
    },
    {
      "epoch": 0.2556426148988422,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 1093
    },
    {
      "epoch": 0.2558765056718513,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 1094
    },
    {
      "epoch": 0.2561103964448603,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6542,
      "step": 1095
    },
    {
      "epoch": 0.2563442872178693,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 1096
    },
    {
      "epoch": 0.2565781779908783,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5423,
      "step": 1097
    },
    {
      "epoch": 0.2568120687638873,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 1098
    },
    {
      "epoch": 0.2570459595368963,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 1099
    },
    {
      "epoch": 0.2572798503099053,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 1100
    },
    {
      "epoch": 0.2572798503099053,
      "eval_runtime": 4.6099,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 1100
    },
    {
      "epoch": 0.2575137410829143,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 1101
    },
    {
      "epoch": 0.2577476318559233,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8676,
      "step": 1102
    },
    {
      "epoch": 0.2579815226289323,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 1103
    },
    {
      "epoch": 0.2582154134019413,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6097,
      "step": 1104
    },
    {
      "epoch": 0.2584493041749503,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 1105
    },
    {
      "epoch": 0.2586831949479593,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 1106
    },
    {
      "epoch": 0.25891708572096833,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 1107
    },
    {
      "epoch": 0.25915097649397734,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 1108
    },
    {
      "epoch": 0.25938486726698634,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 1109
    },
    {
      "epoch": 0.25961875803999535,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 1110
    },
    {
      "epoch": 0.25985264881300435,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 1111
    },
    {
      "epoch": 0.26008653958601335,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 1112
    },
    {
      "epoch": 0.26032043035902236,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 1113
    },
    {
      "epoch": 0.26055432113203136,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 1114
    },
    {
      "epoch": 0.26078821190504037,
      "grad_norm": 7.5625,
      "learning_rate": 3e-05,
      "loss": 1.6484,
      "step": 1115
    },
    {
      "epoch": 0.2610221026780494,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 1116
    },
    {
      "epoch": 0.2612559934510584,
      "grad_norm": 8.0,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 1117
    },
    {
      "epoch": 0.2614898842240674,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 1118
    },
    {
      "epoch": 0.2617237749970764,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5536,
      "step": 1119
    },
    {
      "epoch": 0.2619576657700854,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 1120
    },
    {
      "epoch": 0.2621915565430944,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8382,
      "step": 1121
    },
    {
      "epoch": 0.2624254473161034,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 1122
    },
    {
      "epoch": 0.2626593380891124,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 1123
    },
    {
      "epoch": 0.2628932288621214,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 1124
    },
    {
      "epoch": 0.2631271196351304,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 1125
    },
    {
      "epoch": 0.2633610104081394,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 1126
    },
    {
      "epoch": 0.2635949011811484,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 1127
    },
    {
      "epoch": 0.2638287919541574,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 1128
    },
    {
      "epoch": 0.26406268272716643,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 1129
    },
    {
      "epoch": 0.26429657350017544,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 1130
    },
    {
      "epoch": 0.26453046427318444,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 1131
    },
    {
      "epoch": 0.26476435504619344,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 1132
    },
    {
      "epoch": 0.26499824581920245,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2616,
      "step": 1133
    },
    {
      "epoch": 0.26523213659221145,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 1134
    },
    {
      "epoch": 0.26546602736522046,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 1135
    },
    {
      "epoch": 0.26569991813822946,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 1136
    },
    {
      "epoch": 0.26593380891123847,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 1137
    },
    {
      "epoch": 0.26616769968424747,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 1138
    },
    {
      "epoch": 0.2664015904572565,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 1139
    },
    {
      "epoch": 0.2666354812302655,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 1140
    },
    {
      "epoch": 0.2668693720032745,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7003,
      "step": 1141
    },
    {
      "epoch": 0.2671032627762835,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 1142
    },
    {
      "epoch": 0.2673371535492925,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 1143
    },
    {
      "epoch": 0.2675710443223015,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3239,
      "step": 1144
    },
    {
      "epoch": 0.2678049350953105,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 1145
    },
    {
      "epoch": 0.2680388258683195,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 1146
    },
    {
      "epoch": 0.2682727166413285,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2714,
      "step": 1147
    },
    {
      "epoch": 0.2685066074143375,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 1148
    },
    {
      "epoch": 0.2687404981873465,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 1149
    },
    {
      "epoch": 0.2689743889603555,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 1150
    },
    {
      "epoch": 0.26920827973336453,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 1151
    },
    {
      "epoch": 0.26944217050637354,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 1152
    },
    {
      "epoch": 0.26967606127938254,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 1153
    },
    {
      "epoch": 0.26990995205239154,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 1154
    },
    {
      "epoch": 0.27014384282540055,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 1155
    },
    {
      "epoch": 0.27037773359840955,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 1156
    },
    {
      "epoch": 0.27061162437141856,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 1157
    },
    {
      "epoch": 0.27084551514442756,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7532,
      "step": 1158
    },
    {
      "epoch": 0.27107940591743657,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 1159
    },
    {
      "epoch": 0.27131329669044557,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 1160
    },
    {
      "epoch": 0.2715471874634546,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 1161
    },
    {
      "epoch": 0.2717810782364636,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8776,
      "step": 1162
    },
    {
      "epoch": 0.2720149690094726,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5589,
      "step": 1163
    },
    {
      "epoch": 0.2722488597824816,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 1164
    },
    {
      "epoch": 0.2724827505554906,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 1165
    },
    {
      "epoch": 0.2727166413284996,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 1166
    },
    {
      "epoch": 0.2729505321015086,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 1167
    },
    {
      "epoch": 0.2731844228745176,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 1168
    },
    {
      "epoch": 0.2734183136475266,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 1169
    },
    {
      "epoch": 0.2736522044205356,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 1170
    },
    {
      "epoch": 0.2738860951935446,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 1171
    },
    {
      "epoch": 0.2741199859665536,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 1172
    },
    {
      "epoch": 0.27435387673956263,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 1173
    },
    {
      "epoch": 0.27458776751257163,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 1174
    },
    {
      "epoch": 0.27482165828558064,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 1175
    },
    {
      "epoch": 0.27505554905858964,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 1176
    },
    {
      "epoch": 0.27528943983159865,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6334,
      "step": 1177
    },
    {
      "epoch": 0.27552333060460765,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 1178
    },
    {
      "epoch": 0.27575722137761666,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 1179
    },
    {
      "epoch": 0.27599111215062566,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 1180
    },
    {
      "epoch": 0.27622500292363467,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 1181
    },
    {
      "epoch": 0.27645889369664367,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4115,
      "step": 1182
    },
    {
      "epoch": 0.2766927844696527,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6085,
      "step": 1183
    },
    {
      "epoch": 0.2769266752426617,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 1184
    },
    {
      "epoch": 0.2771605660156707,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 1185
    },
    {
      "epoch": 0.2773944567886797,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 1186
    },
    {
      "epoch": 0.2776283475616887,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 1187
    },
    {
      "epoch": 0.2778622383346977,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 1188
    },
    {
      "epoch": 0.2780961291077067,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 1189
    },
    {
      "epoch": 0.2783300198807157,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5819,
      "step": 1190
    },
    {
      "epoch": 0.2785639106537247,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 1191
    },
    {
      "epoch": 0.2787978014267337,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 1192
    },
    {
      "epoch": 0.2790316921997427,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 1193
    },
    {
      "epoch": 0.2792655829727517,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 1194
    },
    {
      "epoch": 0.27949947374576073,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 1195
    },
    {
      "epoch": 0.27973336451876973,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 1196
    },
    {
      "epoch": 0.27996725529177874,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 1197
    },
    {
      "epoch": 0.28020114606478774,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 1198
    },
    {
      "epoch": 0.28043503683779675,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 1199
    },
    {
      "epoch": 0.28066892761080575,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2331,
      "step": 1200
    },
    {
      "epoch": 0.28066892761080575,
      "eval_runtime": 4.6134,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 1200
    },
    {
      "epoch": 0.28090281838381476,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 1201
    },
    {
      "epoch": 0.28113670915682376,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 1202
    },
    {
      "epoch": 0.28137059992983277,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 1203
    },
    {
      "epoch": 0.28160449070284177,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 1204
    },
    {
      "epoch": 0.2818383814758508,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 1205
    },
    {
      "epoch": 0.2820722722488598,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 1206
    },
    {
      "epoch": 0.2823061630218688,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9831,
      "step": 1207
    },
    {
      "epoch": 0.2825400537948778,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 1208
    },
    {
      "epoch": 0.2827739445678868,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 1209
    },
    {
      "epoch": 0.2830078353408958,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 1210
    },
    {
      "epoch": 0.2832417261139048,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8174,
      "step": 1211
    },
    {
      "epoch": 0.2834756168869138,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 1212
    },
    {
      "epoch": 0.2837095076599228,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 1213
    },
    {
      "epoch": 0.2839433984329318,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.694,
      "step": 1214
    },
    {
      "epoch": 0.2841772892059408,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 1215
    },
    {
      "epoch": 0.2844111799789498,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.296,
      "step": 1216
    },
    {
      "epoch": 0.28464507075195883,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 1217
    },
    {
      "epoch": 0.28487896152496783,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8177,
      "step": 1218
    },
    {
      "epoch": 0.28511285229797684,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 1219
    },
    {
      "epoch": 0.28534674307098584,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 1220
    },
    {
      "epoch": 0.28558063384399485,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9133,
      "step": 1221
    },
    {
      "epoch": 0.28581452461700385,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 1222
    },
    {
      "epoch": 0.28604841539001286,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 1223
    },
    {
      "epoch": 0.28628230616302186,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 1224
    },
    {
      "epoch": 0.28651619693603086,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 1225
    },
    {
      "epoch": 0.28675008770903987,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 1226
    },
    {
      "epoch": 0.2869839784820489,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 1227
    },
    {
      "epoch": 0.2872178692550579,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 1228
    },
    {
      "epoch": 0.2874517600280669,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 1229
    },
    {
      "epoch": 0.2876856508010759,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 1230
    },
    {
      "epoch": 0.2879195415740849,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 1231
    },
    {
      "epoch": 0.2881534323470939,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 1232
    },
    {
      "epoch": 0.2883873231201029,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 1233
    },
    {
      "epoch": 0.2886212138931119,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 1234
    },
    {
      "epoch": 0.2888551046661209,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 1235
    },
    {
      "epoch": 0.2890889954391299,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 1236
    },
    {
      "epoch": 0.2893228862121389,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 1237
    },
    {
      "epoch": 0.2895567769851479,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8034,
      "step": 1238
    },
    {
      "epoch": 0.28979066775815693,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 1239
    },
    {
      "epoch": 0.29002455853116593,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.523,
      "step": 1240
    },
    {
      "epoch": 0.29025844930417494,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 1241
    },
    {
      "epoch": 0.29049234007718394,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 1242
    },
    {
      "epoch": 0.29072623085019295,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 1243
    },
    {
      "epoch": 0.29096012162320195,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 1244
    },
    {
      "epoch": 0.29119401239621096,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.216,
      "step": 1245
    },
    {
      "epoch": 0.29142790316921996,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 1246
    },
    {
      "epoch": 0.29166179394222896,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 1247
    },
    {
      "epoch": 0.29189568471523797,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 1248
    },
    {
      "epoch": 0.292129575488247,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 1249
    },
    {
      "epoch": 0.292363466261256,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 1250
    },
    {
      "epoch": 0.292597357034265,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 1251
    },
    {
      "epoch": 0.292831247807274,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 1252
    },
    {
      "epoch": 0.293065138580283,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 1253
    },
    {
      "epoch": 0.293299029353292,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 1254
    },
    {
      "epoch": 0.293532920126301,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 1255
    },
    {
      "epoch": 0.29376681089931,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 1256
    },
    {
      "epoch": 0.294000701672319,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 1257
    },
    {
      "epoch": 0.294234592445328,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 1258
    },
    {
      "epoch": 0.294468483218337,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9787,
      "step": 1259
    },
    {
      "epoch": 0.294702373991346,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 1260
    },
    {
      "epoch": 0.294936264764355,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 1261
    },
    {
      "epoch": 0.29517015553736403,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 1262
    },
    {
      "epoch": 0.29540404631037304,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 1263
    },
    {
      "epoch": 0.29563793708338204,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 1264
    },
    {
      "epoch": 0.29587182785639105,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 1265
    },
    {
      "epoch": 0.29610571862940005,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 1266
    },
    {
      "epoch": 0.29633960940240905,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 1267
    },
    {
      "epoch": 0.29657350017541806,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 1268
    },
    {
      "epoch": 0.29680739094842706,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 1269
    },
    {
      "epoch": 0.29704128172143607,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8657,
      "step": 1270
    },
    {
      "epoch": 0.2972751724944451,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7403,
      "step": 1271
    },
    {
      "epoch": 0.2975090632674541,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7148,
      "step": 1272
    },
    {
      "epoch": 0.2977429540404631,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 1273
    },
    {
      "epoch": 0.2979768448134721,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 1274
    },
    {
      "epoch": 0.2982107355864811,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 1275
    },
    {
      "epoch": 0.2984446263594901,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 1276
    },
    {
      "epoch": 0.2986785171324991,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8305,
      "step": 1277
    },
    {
      "epoch": 0.2989124079055081,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 1278
    },
    {
      "epoch": 0.2991462986785171,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 1279
    },
    {
      "epoch": 0.2993801894515261,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 1280
    },
    {
      "epoch": 0.2996140802245351,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.5406,
      "step": 1281
    },
    {
      "epoch": 0.2998479709975441,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 1282
    },
    {
      "epoch": 0.3000818617705531,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 1283
    },
    {
      "epoch": 0.30031575254356213,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 1284
    },
    {
      "epoch": 0.30054964331657114,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 1285
    },
    {
      "epoch": 0.30078353408958014,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2177,
      "step": 1286
    },
    {
      "epoch": 0.30101742486258914,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 1287
    },
    {
      "epoch": 0.30125131563559815,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 1288
    },
    {
      "epoch": 0.30148520640860715,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 1289
    },
    {
      "epoch": 0.30171909718161616,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 1290
    },
    {
      "epoch": 0.30195298795462516,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 1291
    },
    {
      "epoch": 0.30218687872763417,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 1292
    },
    {
      "epoch": 0.3024207695006432,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 1293
    },
    {
      "epoch": 0.30265466027365223,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 1294
    },
    {
      "epoch": 0.30288855104666124,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 1295
    },
    {
      "epoch": 0.30312244181967024,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 1296
    },
    {
      "epoch": 0.30335633259267925,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 1297
    },
    {
      "epoch": 0.30359022336568825,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 1298
    },
    {
      "epoch": 0.30382411413869725,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9758,
      "step": 1299
    },
    {
      "epoch": 0.30405800491170626,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 1300
    },
    {
      "epoch": 0.30405800491170626,
      "eval_runtime": 4.593,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 1300
    },
    {
      "epoch": 0.30429189568471526,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 1301
    },
    {
      "epoch": 0.30452578645772427,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6635,
      "step": 1302
    },
    {
      "epoch": 0.3047596772307333,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 1303
    },
    {
      "epoch": 0.3049935680037423,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 1304
    },
    {
      "epoch": 0.3052274587767513,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 1305
    },
    {
      "epoch": 0.3054613495497603,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 1306
    },
    {
      "epoch": 0.3056952403227693,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8644,
      "step": 1307
    },
    {
      "epoch": 0.3059291310957783,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8201,
      "step": 1308
    },
    {
      "epoch": 0.3061630218687873,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 1309
    },
    {
      "epoch": 0.3063969126417963,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6281,
      "step": 1310
    },
    {
      "epoch": 0.3066308034148053,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 1311
    },
    {
      "epoch": 0.3068646941878143,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 1312
    },
    {
      "epoch": 0.3070985849608233,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 1313
    },
    {
      "epoch": 0.3073324757338323,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 1314
    },
    {
      "epoch": 0.3075663665068413,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6425,
      "step": 1315
    },
    {
      "epoch": 0.30780025727985033,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 1316
    },
    {
      "epoch": 0.30803414805285934,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 1317
    },
    {
      "epoch": 0.30826803882586834,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 1318
    },
    {
      "epoch": 0.30850192959887734,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1656,
      "step": 1319
    },
    {
      "epoch": 0.30873582037188635,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 1320
    },
    {
      "epoch": 0.30896971114489535,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 1321
    },
    {
      "epoch": 0.30920360191790436,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 1322
    },
    {
      "epoch": 0.30943749269091336,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 1323
    },
    {
      "epoch": 0.30967138346392237,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6628,
      "step": 1324
    },
    {
      "epoch": 0.30990527423693137,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6233,
      "step": 1325
    },
    {
      "epoch": 0.3101391650099404,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 1326
    },
    {
      "epoch": 0.3103730557829494,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 1327
    },
    {
      "epoch": 0.3106069465559584,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 1328
    },
    {
      "epoch": 0.3108408373289674,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 1329
    },
    {
      "epoch": 0.3110747281019764,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 1330
    },
    {
      "epoch": 0.3113086188749854,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.857,
      "step": 1331
    },
    {
      "epoch": 0.3115425096479944,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 1332
    },
    {
      "epoch": 0.3117764004210034,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 1333
    },
    {
      "epoch": 0.3120102911940124,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 1334
    },
    {
      "epoch": 0.3122441819670214,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 1335
    },
    {
      "epoch": 0.3124780727400304,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 1336
    },
    {
      "epoch": 0.3127119635130394,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 1337
    },
    {
      "epoch": 0.31294585428604843,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 1338
    },
    {
      "epoch": 0.31317974505905743,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 1339
    },
    {
      "epoch": 0.31341363583206644,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 1340
    },
    {
      "epoch": 0.31364752660507544,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6364,
      "step": 1341
    },
    {
      "epoch": 0.31388141737808445,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8195,
      "step": 1342
    },
    {
      "epoch": 0.31411530815109345,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 1343
    },
    {
      "epoch": 0.31434919892410246,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 1344
    },
    {
      "epoch": 0.31458308969711146,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 1345
    },
    {
      "epoch": 0.31481698047012047,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 1346
    },
    {
      "epoch": 0.31505087124312947,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 1347
    },
    {
      "epoch": 0.3152847620161385,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 1348
    },
    {
      "epoch": 0.3155186527891475,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 1349
    },
    {
      "epoch": 0.3157525435621565,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 1350
    },
    {
      "epoch": 0.3159864343351655,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 1351
    },
    {
      "epoch": 0.3162203251081745,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 1352
    },
    {
      "epoch": 0.3164542158811835,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 1353
    },
    {
      "epoch": 0.3166881066541925,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 1354
    },
    {
      "epoch": 0.3169219974272015,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 1355
    },
    {
      "epoch": 0.3171558882002105,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 1356
    },
    {
      "epoch": 0.3173897789732195,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.691,
      "step": 1357
    },
    {
      "epoch": 0.3176236697462285,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 1358
    },
    {
      "epoch": 0.3178575605192375,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 1359
    },
    {
      "epoch": 0.31809145129224653,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 1360
    },
    {
      "epoch": 0.31832534206525553,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 1361
    },
    {
      "epoch": 0.31855923283826454,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 1362
    },
    {
      "epoch": 0.31879312361127354,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 1363
    },
    {
      "epoch": 0.31902701438428255,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 1364
    },
    {
      "epoch": 0.31926090515729155,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 1365
    },
    {
      "epoch": 0.31949479593030056,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 1366
    },
    {
      "epoch": 0.31972868670330956,
      "grad_norm": 7.40625,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 1367
    },
    {
      "epoch": 0.31996257747631857,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 1368
    },
    {
      "epoch": 0.32019646824932757,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 1369
    },
    {
      "epoch": 0.3204303590223366,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 1370
    },
    {
      "epoch": 0.3206642497953456,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 1371
    },
    {
      "epoch": 0.3208981405683546,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 1372
    },
    {
      "epoch": 0.3211320313413636,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 1373
    },
    {
      "epoch": 0.3213659221143726,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7624,
      "step": 1374
    },
    {
      "epoch": 0.3215998128873816,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 1375
    },
    {
      "epoch": 0.3218337036603906,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 1376
    },
    {
      "epoch": 0.3220675944333996,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 1377
    },
    {
      "epoch": 0.3223014852064086,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 1378
    },
    {
      "epoch": 0.3225353759794176,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 1379
    },
    {
      "epoch": 0.3227692667524266,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 1380
    },
    {
      "epoch": 0.3230031575254356,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 1381
    },
    {
      "epoch": 0.32323704829844463,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 1382
    },
    {
      "epoch": 0.32347093907145363,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 1383
    },
    {
      "epoch": 0.32370482984446264,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 1384
    },
    {
      "epoch": 0.32393872061747164,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 1385
    },
    {
      "epoch": 0.32417261139048065,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 1386
    },
    {
      "epoch": 0.32440650216348965,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 1387
    },
    {
      "epoch": 0.32464039293649866,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 1388
    },
    {
      "epoch": 0.32487428370950766,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 1389
    },
    {
      "epoch": 0.32510817448251667,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 1390
    },
    {
      "epoch": 0.32534206525552567,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 1391
    },
    {
      "epoch": 0.3255759560285347,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 1392
    },
    {
      "epoch": 0.3258098468015437,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 1393
    },
    {
      "epoch": 0.3260437375745527,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 1394
    },
    {
      "epoch": 0.3262776283475617,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 1395
    },
    {
      "epoch": 0.3265115191205707,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 1396
    },
    {
      "epoch": 0.3267454098935797,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 1397
    },
    {
      "epoch": 0.3269793006665887,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 1398
    },
    {
      "epoch": 0.3272131914395977,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 1399
    },
    {
      "epoch": 0.3274470822126067,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 1400
    },
    {
      "epoch": 0.3274470822126067,
      "eval_runtime": 4.6395,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 1400
    },
    {
      "epoch": 0.3276809729856157,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 1401
    },
    {
      "epoch": 0.3279148637586247,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.5474,
      "step": 1402
    },
    {
      "epoch": 0.3281487545316337,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 1403
    },
    {
      "epoch": 0.32838264530464273,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 1404
    },
    {
      "epoch": 0.32861653607765173,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 1405
    },
    {
      "epoch": 0.32885042685066074,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 1406
    },
    {
      "epoch": 0.32908431762366974,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 1407
    },
    {
      "epoch": 0.32931820839667875,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 1408
    },
    {
      "epoch": 0.32955209916968775,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 1409
    },
    {
      "epoch": 0.32978598994269676,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 1410
    },
    {
      "epoch": 0.33001988071570576,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 1411
    },
    {
      "epoch": 0.33025377148871476,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 1412
    },
    {
      "epoch": 0.33048766226172377,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.294,
      "step": 1413
    },
    {
      "epoch": 0.3307215530347328,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 1414
    },
    {
      "epoch": 0.3309554438077418,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 1415
    },
    {
      "epoch": 0.3311893345807508,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 1416
    },
    {
      "epoch": 0.3314232253537598,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 1417
    },
    {
      "epoch": 0.3316571161267688,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 1418
    },
    {
      "epoch": 0.3318910068997778,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 1419
    },
    {
      "epoch": 0.3321248976727868,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6833,
      "step": 1420
    },
    {
      "epoch": 0.3323587884457958,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 1421
    },
    {
      "epoch": 0.3325926792188048,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 1422
    },
    {
      "epoch": 0.3328265699918138,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 1423
    },
    {
      "epoch": 0.3330604607648228,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 1424
    },
    {
      "epoch": 0.3332943515378318,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 1425
    },
    {
      "epoch": 0.33352824231084083,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 1426
    },
    {
      "epoch": 0.33376213308384983,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 1427
    },
    {
      "epoch": 0.33399602385685884,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 1428
    },
    {
      "epoch": 0.33422991462986784,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 1429
    },
    {
      "epoch": 0.33446380540287685,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 1430
    },
    {
      "epoch": 0.33469769617588585,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 1431
    },
    {
      "epoch": 0.33493158694889485,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 1432
    },
    {
      "epoch": 0.33516547772190386,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 1433
    },
    {
      "epoch": 0.33539936849491286,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 1434
    },
    {
      "epoch": 0.33563325926792187,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 1435
    },
    {
      "epoch": 0.3358671500409309,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 1436
    },
    {
      "epoch": 0.3361010408139399,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 1437
    },
    {
      "epoch": 0.3363349315869489,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 1438
    },
    {
      "epoch": 0.3365688223599579,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 1439
    },
    {
      "epoch": 0.3368027131329669,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 1440
    },
    {
      "epoch": 0.3370366039059759,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 1441
    },
    {
      "epoch": 0.3372704946789849,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 1442
    },
    {
      "epoch": 0.3375043854519939,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 1443
    },
    {
      "epoch": 0.3377382762250029,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 1444
    },
    {
      "epoch": 0.3379721669980119,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 1445
    },
    {
      "epoch": 0.3382060577710209,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 1446
    },
    {
      "epoch": 0.3384399485440299,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 1447
    },
    {
      "epoch": 0.3386738393170389,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 1448
    },
    {
      "epoch": 0.33890773009004793,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 1449
    },
    {
      "epoch": 0.33914162086305694,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 1450
    },
    {
      "epoch": 0.33937551163606594,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 1451
    },
    {
      "epoch": 0.33960940240907495,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 1452
    },
    {
      "epoch": 0.33984329318208395,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 1453
    },
    {
      "epoch": 0.34007718395509295,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 1454
    },
    {
      "epoch": 0.34031107472810196,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7051,
      "step": 1455
    },
    {
      "epoch": 0.34054496550111096,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 1456
    },
    {
      "epoch": 0.34077885627411997,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 1457
    },
    {
      "epoch": 0.341012747047129,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 1458
    },
    {
      "epoch": 0.341246637820138,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 1459
    },
    {
      "epoch": 0.341480528593147,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 1460
    },
    {
      "epoch": 0.341714419366156,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 1461
    },
    {
      "epoch": 0.341948310139165,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 1462
    },
    {
      "epoch": 0.342182200912174,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 1463
    },
    {
      "epoch": 0.342416091685183,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 1464
    },
    {
      "epoch": 0.342649982458192,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 1465
    },
    {
      "epoch": 0.342883873231201,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 1466
    },
    {
      "epoch": 0.34311776400421,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 1467
    },
    {
      "epoch": 0.343351654777219,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 1468
    },
    {
      "epoch": 0.343585545550228,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 1469
    },
    {
      "epoch": 0.343819436323237,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 1470
    },
    {
      "epoch": 0.34405332709624603,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.314,
      "step": 1471
    },
    {
      "epoch": 0.34428721786925504,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 1472
    },
    {
      "epoch": 0.34452110864226404,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2216,
      "step": 1473
    },
    {
      "epoch": 0.34475499941527304,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.4973,
      "step": 1474
    },
    {
      "epoch": 0.34498889018828205,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 1475
    },
    {
      "epoch": 0.34522278096129105,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 2.4726,
      "step": 1476
    },
    {
      "epoch": 0.34545667173430006,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 1477
    },
    {
      "epoch": 0.34569056250730906,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 1478
    },
    {
      "epoch": 0.34592445328031807,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 1479
    },
    {
      "epoch": 0.34615834405332707,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2207,
      "step": 1480
    },
    {
      "epoch": 0.3463922348263361,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 1481
    },
    {
      "epoch": 0.3466261255993451,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 1482
    },
    {
      "epoch": 0.3468600163723541,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 1483
    },
    {
      "epoch": 0.3470939071453631,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 1484
    },
    {
      "epoch": 0.3473277979183721,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 1485
    },
    {
      "epoch": 0.3475616886913811,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 1486
    },
    {
      "epoch": 0.3477955794643901,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 1487
    },
    {
      "epoch": 0.3480294702373991,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1204,
      "step": 1488
    },
    {
      "epoch": 0.3482633610104081,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 1489
    },
    {
      "epoch": 0.3484972517834171,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 1490
    },
    {
      "epoch": 0.3487311425564261,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 1491
    },
    {
      "epoch": 0.3489650333294352,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 1492
    },
    {
      "epoch": 0.3491989241024442,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 1493
    },
    {
      "epoch": 0.3494328148754532,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 1494
    },
    {
      "epoch": 0.3496667056484622,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 1495
    },
    {
      "epoch": 0.3499005964214712,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 1496
    },
    {
      "epoch": 0.3501344871944802,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 1497
    },
    {
      "epoch": 0.3503683779674892,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 1498
    },
    {
      "epoch": 0.3506022687404982,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 1499
    },
    {
      "epoch": 0.3508361595135072,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 1500
    },
    {
      "epoch": 0.3508361595135072,
      "eval_runtime": 4.6584,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 1500
    },
    {
      "epoch": 0.3510700502865162,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 1501
    },
    {
      "epoch": 0.3513039410595252,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 1502
    },
    {
      "epoch": 0.35153783183253423,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 1503
    },
    {
      "epoch": 0.35177172260554324,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 1504
    },
    {
      "epoch": 0.35200561337855224,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 1505
    },
    {
      "epoch": 0.35223950415156124,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8607,
      "step": 1506
    },
    {
      "epoch": 0.35247339492457025,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 1507
    },
    {
      "epoch": 0.35270728569757925,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 1508
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 1509
    },
    {
      "epoch": 0.35317506724359726,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3056,
      "step": 1510
    },
    {
      "epoch": 0.35340895801660627,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 1511
    },
    {
      "epoch": 0.35364284878961527,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 1512
    },
    {
      "epoch": 0.3538767395626243,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6123,
      "step": 1513
    },
    {
      "epoch": 0.3541106303356333,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 1514
    },
    {
      "epoch": 0.3543445211086423,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 1515
    },
    {
      "epoch": 0.3545784118816513,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 1516
    },
    {
      "epoch": 0.3548123026546603,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 1517
    },
    {
      "epoch": 0.3550461934276693,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 1518
    },
    {
      "epoch": 0.3552800842006783,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 1519
    },
    {
      "epoch": 0.3555139749736873,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 1520
    },
    {
      "epoch": 0.3557478657466963,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2277,
      "step": 1521
    },
    {
      "epoch": 0.3559817565197053,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 1522
    },
    {
      "epoch": 0.3562156472927143,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 1523
    },
    {
      "epoch": 0.3564495380657233,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 1524
    },
    {
      "epoch": 0.35668342883873233,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 1525
    },
    {
      "epoch": 0.35691731961174133,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 1526
    },
    {
      "epoch": 0.35715121038475034,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 1527
    },
    {
      "epoch": 0.35738510115775934,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3288,
      "step": 1528
    },
    {
      "epoch": 0.35761899193076835,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 1529
    },
    {
      "epoch": 0.35785288270377735,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.3717,
      "step": 1530
    },
    {
      "epoch": 0.35808677347678636,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 1531
    },
    {
      "epoch": 0.35832066424979536,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7682,
      "step": 1532
    },
    {
      "epoch": 0.35855455502280437,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 1533
    },
    {
      "epoch": 0.35878844579581337,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 1534
    },
    {
      "epoch": 0.3590223365688224,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 1535
    },
    {
      "epoch": 0.3592562273418314,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 1536
    },
    {
      "epoch": 0.3594901181148404,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 1537
    },
    {
      "epoch": 0.3597240088878494,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 1538
    },
    {
      "epoch": 0.3599578996608584,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6197,
      "step": 1539
    },
    {
      "epoch": 0.3601917904338674,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 1540
    },
    {
      "epoch": 0.3604256812068764,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 1541
    },
    {
      "epoch": 0.3606595719798854,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7854,
      "step": 1542
    },
    {
      "epoch": 0.3608934627528944,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 1543
    },
    {
      "epoch": 0.3611273535259034,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 1544
    },
    {
      "epoch": 0.3613612442989124,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 1545
    },
    {
      "epoch": 0.3615951350719214,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 1546
    },
    {
      "epoch": 0.36182902584493043,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 1547
    },
    {
      "epoch": 0.36206291661793943,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6826,
      "step": 1548
    },
    {
      "epoch": 0.36229680739094844,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 1549
    },
    {
      "epoch": 0.36253069816395744,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 1550
    },
    {
      "epoch": 0.36276458893696645,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 1551
    },
    {
      "epoch": 0.36299847970997545,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 1552
    },
    {
      "epoch": 0.36323237048298446,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 1553
    },
    {
      "epoch": 0.36346626125599346,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2243,
      "step": 1554
    },
    {
      "epoch": 0.36370015202900247,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 1555
    },
    {
      "epoch": 0.36393404280201147,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 1556
    },
    {
      "epoch": 0.3641679335750205,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 1557
    },
    {
      "epoch": 0.3644018243480295,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 1558
    },
    {
      "epoch": 0.3646357151210385,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 1559
    },
    {
      "epoch": 0.3648696058940475,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8719,
      "step": 1560
    },
    {
      "epoch": 0.3651034966670565,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 1561
    },
    {
      "epoch": 0.3653373874400655,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 1562
    },
    {
      "epoch": 0.3655712782130745,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 1563
    },
    {
      "epoch": 0.3658051689860835,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 1564
    },
    {
      "epoch": 0.3660390597590925,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 1565
    },
    {
      "epoch": 0.3662729505321015,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 1566
    },
    {
      "epoch": 0.3665068413051105,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 1567
    },
    {
      "epoch": 0.3667407320781195,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 1568
    },
    {
      "epoch": 0.36697462285112853,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 1569
    },
    {
      "epoch": 0.36720851362413753,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2292,
      "step": 1570
    },
    {
      "epoch": 0.36744240439714654,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 1571
    },
    {
      "epoch": 0.36767629517015554,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 1572
    },
    {
      "epoch": 0.36791018594316455,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 1573
    },
    {
      "epoch": 0.36814407671617355,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 1574
    },
    {
      "epoch": 0.36837796748918256,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6873,
      "step": 1575
    },
    {
      "epoch": 0.36861185826219156,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 1576
    },
    {
      "epoch": 0.36884574903520057,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 1577
    },
    {
      "epoch": 0.36907963980820957,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 1578
    },
    {
      "epoch": 0.3693135305812186,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 1579
    },
    {
      "epoch": 0.3695474213542276,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 1580
    },
    {
      "epoch": 0.3697813121272366,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 1581
    },
    {
      "epoch": 0.3700152029002456,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 1582
    },
    {
      "epoch": 0.3702490936732546,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 1583
    },
    {
      "epoch": 0.3704829844462636,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0252,
      "step": 1584
    },
    {
      "epoch": 0.3707168752192726,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 1585
    },
    {
      "epoch": 0.3709507659922816,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 1586
    },
    {
      "epoch": 0.3711846567652906,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7147,
      "step": 1587
    },
    {
      "epoch": 0.3714185475382996,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 1588
    },
    {
      "epoch": 0.3716524383113086,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 1589
    },
    {
      "epoch": 0.3718863290843176,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 1590
    },
    {
      "epoch": 0.37212021985732663,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 1591
    },
    {
      "epoch": 0.37235411063033563,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 1592
    },
    {
      "epoch": 0.37258800140334464,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 1593
    },
    {
      "epoch": 0.37282189217635364,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 1594
    },
    {
      "epoch": 0.37305578294936265,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 1595
    },
    {
      "epoch": 0.37328967372237165,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1813,
      "step": 1596
    },
    {
      "epoch": 0.37352356449538066,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.5603,
      "step": 1597
    },
    {
      "epoch": 0.37375745526838966,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 1598
    },
    {
      "epoch": 0.37399134604139866,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7069,
      "step": 1599
    },
    {
      "epoch": 0.37422523681440767,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 1600
    },
    {
      "epoch": 0.37422523681440767,
      "eval_runtime": 4.602,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 1600
    },
    {
      "epoch": 0.3744591275874167,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 1601
    },
    {
      "epoch": 0.3746930183604257,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.577,
      "step": 1602
    },
    {
      "epoch": 0.3749269091334347,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 1603
    },
    {
      "epoch": 0.3751607999064437,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 1604
    },
    {
      "epoch": 0.3753946906794527,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 1605
    },
    {
      "epoch": 0.3756285814524617,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 1606
    },
    {
      "epoch": 0.3758624722254707,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 1607
    },
    {
      "epoch": 0.3760963629984797,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8234,
      "step": 1608
    },
    {
      "epoch": 0.3763302537714887,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 1609
    },
    {
      "epoch": 0.3765641445444977,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 1610
    },
    {
      "epoch": 0.3767980353175067,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 1611
    },
    {
      "epoch": 0.3770319260905157,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5921,
      "step": 1612
    },
    {
      "epoch": 0.37726581686352473,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 1613
    },
    {
      "epoch": 0.37749970763653373,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 1614
    },
    {
      "epoch": 0.37773359840954274,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3212,
      "step": 1615
    },
    {
      "epoch": 0.37796748918255174,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 1616
    },
    {
      "epoch": 0.37820137995556075,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 1617
    },
    {
      "epoch": 0.37843527072856975,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 1618
    },
    {
      "epoch": 0.37866916150157875,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 1619
    },
    {
      "epoch": 0.37890305227458776,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8166,
      "step": 1620
    },
    {
      "epoch": 0.37913694304759676,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 1621
    },
    {
      "epoch": 0.37937083382060577,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 1622
    },
    {
      "epoch": 0.3796047245936148,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 1623
    },
    {
      "epoch": 0.3798386153666238,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 1624
    },
    {
      "epoch": 0.3800725061396328,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1511,
      "step": 1625
    },
    {
      "epoch": 0.3803063969126418,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 1626
    },
    {
      "epoch": 0.3805402876856508,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 1627
    },
    {
      "epoch": 0.3807741784586598,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3105,
      "step": 1628
    },
    {
      "epoch": 0.3810080692316688,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 1629
    },
    {
      "epoch": 0.3812419600046778,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3653,
      "step": 1630
    },
    {
      "epoch": 0.3814758507776868,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2269,
      "step": 1631
    },
    {
      "epoch": 0.3817097415506958,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 1632
    },
    {
      "epoch": 0.3819436323237048,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 1633
    },
    {
      "epoch": 0.3821775230967138,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 1634
    },
    {
      "epoch": 0.3824114138697228,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 1635
    },
    {
      "epoch": 0.38264530464273183,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8524,
      "step": 1636
    },
    {
      "epoch": 0.38287919541574084,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 1637
    },
    {
      "epoch": 0.38311308618874984,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 1638
    },
    {
      "epoch": 0.38334697696175885,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7518,
      "step": 1639
    },
    {
      "epoch": 0.38358086773476785,
      "grad_norm": 8.8125,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 1640
    },
    {
      "epoch": 0.38381475850777685,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 1641
    },
    {
      "epoch": 0.38404864928078586,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.39,
      "step": 1642
    },
    {
      "epoch": 0.38428254005379486,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 1643
    },
    {
      "epoch": 0.38451643082680387,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 1644
    },
    {
      "epoch": 0.3847503215998129,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 1645
    },
    {
      "epoch": 0.3849842123728219,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 1646
    },
    {
      "epoch": 0.3852181031458309,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 1647
    },
    {
      "epoch": 0.3854519939188399,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7698,
      "step": 1648
    },
    {
      "epoch": 0.3856858846918489,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 1649
    },
    {
      "epoch": 0.3859197754648579,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 1650
    },
    {
      "epoch": 0.3861536662378669,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0727,
      "step": 1651
    },
    {
      "epoch": 0.3863875570108759,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.336,
      "step": 1652
    },
    {
      "epoch": 0.3866214477838849,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 1653
    },
    {
      "epoch": 0.3868553385568939,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 1654
    },
    {
      "epoch": 0.3870892293299029,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 1655
    },
    {
      "epoch": 0.3873231201029119,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 1656
    },
    {
      "epoch": 0.3875570108759209,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2737,
      "step": 1657
    },
    {
      "epoch": 0.38779090164892993,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 1658
    },
    {
      "epoch": 0.38802479242193894,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 1659
    },
    {
      "epoch": 0.38825868319494794,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 1660
    },
    {
      "epoch": 0.38849257396795694,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 1661
    },
    {
      "epoch": 0.38872646474096595,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 1662
    },
    {
      "epoch": 0.38896035551397495,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 1663
    },
    {
      "epoch": 0.38919424628698396,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.3805,
      "step": 1664
    },
    {
      "epoch": 0.38942813705999296,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 1665
    },
    {
      "epoch": 0.38966202783300197,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 1666
    },
    {
      "epoch": 0.38989591860601097,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 1667
    },
    {
      "epoch": 0.39012980937902,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 1668
    },
    {
      "epoch": 0.390363700152029,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 1669
    },
    {
      "epoch": 0.390597590925038,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 1670
    },
    {
      "epoch": 0.390831481698047,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 1671
    },
    {
      "epoch": 0.391065372471056,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 1672
    },
    {
      "epoch": 0.391299263244065,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 1673
    },
    {
      "epoch": 0.391533154017074,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 1674
    },
    {
      "epoch": 0.391767044790083,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1317,
      "step": 1675
    },
    {
      "epoch": 0.392000935563092,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 1676
    },
    {
      "epoch": 0.392234826336101,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 1677
    },
    {
      "epoch": 0.39246871710911,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 1678
    },
    {
      "epoch": 0.392702607882119,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 1679
    },
    {
      "epoch": 0.39293649865512803,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 1680
    },
    {
      "epoch": 0.39317038942813703,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 1681
    },
    {
      "epoch": 0.39340428020114604,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 1682
    },
    {
      "epoch": 0.39363817097415504,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 1683
    },
    {
      "epoch": 0.39387206174716405,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 1684
    },
    {
      "epoch": 0.39410595252017305,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 1685
    },
    {
      "epoch": 0.39433984329318206,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 1686
    },
    {
      "epoch": 0.39457373406619106,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 1687
    },
    {
      "epoch": 0.39480762483920007,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 1688
    },
    {
      "epoch": 0.39504151561220907,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 1689
    },
    {
      "epoch": 0.3952754063852181,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 1690
    },
    {
      "epoch": 0.39550929715822714,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 1691
    },
    {
      "epoch": 0.39574318793123614,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 1692
    },
    {
      "epoch": 0.39597707870424514,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 1693
    },
    {
      "epoch": 0.39621096947725415,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.236,
      "step": 1694
    },
    {
      "epoch": 0.39644486025026315,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.3707,
      "step": 1695
    },
    {
      "epoch": 0.39667875102327216,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9556,
      "step": 1696
    },
    {
      "epoch": 0.39691264179628116,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 1697
    },
    {
      "epoch": 0.39714653256929017,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 1698
    },
    {
      "epoch": 0.39738042334229917,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 1699
    },
    {
      "epoch": 0.3976143141153082,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 1700
    },
    {
      "epoch": 0.3976143141153082,
      "eval_runtime": 4.6223,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 1700
    },
    {
      "epoch": 0.3978482048883172,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 1701
    },
    {
      "epoch": 0.3980820956613262,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 1702
    },
    {
      "epoch": 0.3983159864343352,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 1703
    },
    {
      "epoch": 0.3985498772073442,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2931,
      "step": 1704
    },
    {
      "epoch": 0.3987837679803532,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 1705
    },
    {
      "epoch": 0.3990176587533622,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 1706
    },
    {
      "epoch": 0.3992515495263712,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 1707
    },
    {
      "epoch": 0.3994854402993802,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 1708
    },
    {
      "epoch": 0.3997193310723892,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 1709
    },
    {
      "epoch": 0.3999532218453982,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 1710
    },
    {
      "epoch": 0.4001871126184072,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 1711
    },
    {
      "epoch": 0.40042100339141623,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 1712
    },
    {
      "epoch": 0.40065489416442523,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 1713
    },
    {
      "epoch": 0.40088878493743424,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 1714
    },
    {
      "epoch": 0.40112267571044324,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 1715
    },
    {
      "epoch": 0.40135656648345225,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.338,
      "step": 1716
    },
    {
      "epoch": 0.40159045725646125,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 1717
    },
    {
      "epoch": 0.40182434802947026,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 1718
    },
    {
      "epoch": 0.40205823880247926,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2205,
      "step": 1719
    },
    {
      "epoch": 0.40229212957548827,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 1720
    },
    {
      "epoch": 0.40252602034849727,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2946,
      "step": 1721
    },
    {
      "epoch": 0.4027599111215063,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 1722
    },
    {
      "epoch": 0.4029938018945153,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 1723
    },
    {
      "epoch": 0.4032276926675243,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2387,
      "step": 1724
    },
    {
      "epoch": 0.4034615834405333,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 1725
    },
    {
      "epoch": 0.4036954742135423,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 1726
    },
    {
      "epoch": 0.4039293649865513,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 1727
    },
    {
      "epoch": 0.4041632557595603,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 1728
    },
    {
      "epoch": 0.4043971465325693,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9432,
      "step": 1729
    },
    {
      "epoch": 0.4046310373055783,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2628,
      "step": 1730
    },
    {
      "epoch": 0.4048649280785873,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 1731
    },
    {
      "epoch": 0.4050988188515963,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 1732
    },
    {
      "epoch": 0.4053327096246053,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 1733
    },
    {
      "epoch": 0.40556660039761433,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7764,
      "step": 1734
    },
    {
      "epoch": 0.40580049117062333,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 1735
    },
    {
      "epoch": 0.40603438194363234,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 1736
    },
    {
      "epoch": 0.40626827271664134,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 1737
    },
    {
      "epoch": 0.40650216348965035,
      "grad_norm": 7.9375,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 1738
    },
    {
      "epoch": 0.40673605426265935,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.591,
      "step": 1739
    },
    {
      "epoch": 0.40696994503566836,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 1740
    },
    {
      "epoch": 0.40720383580867736,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 1741
    },
    {
      "epoch": 0.40743772658168637,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 1742
    },
    {
      "epoch": 0.40767161735469537,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 1743
    },
    {
      "epoch": 0.4079055081277044,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2486,
      "step": 1744
    },
    {
      "epoch": 0.4081393989007134,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 1745
    },
    {
      "epoch": 0.4083732896737224,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2926,
      "step": 1746
    },
    {
      "epoch": 0.4086071804467314,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 1747
    },
    {
      "epoch": 0.4088410712197404,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 1748
    },
    {
      "epoch": 0.4090749619927494,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 1749
    },
    {
      "epoch": 0.4093088527657584,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 1750
    },
    {
      "epoch": 0.4095427435387674,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.232,
      "step": 1751
    },
    {
      "epoch": 0.4097766343117764,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6322,
      "step": 1752
    },
    {
      "epoch": 0.4100105250847854,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 1753
    },
    {
      "epoch": 0.4102444158577944,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6199,
      "step": 1754
    },
    {
      "epoch": 0.4104783066308034,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7328,
      "step": 1755
    },
    {
      "epoch": 0.41071219740381243,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 1756
    },
    {
      "epoch": 0.41094608817682143,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 1757
    },
    {
      "epoch": 0.41117997894983044,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 1758
    },
    {
      "epoch": 0.41141386972283944,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 1759
    },
    {
      "epoch": 0.41164776049584845,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7332,
      "step": 1760
    },
    {
      "epoch": 0.41188165126885745,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 1761
    },
    {
      "epoch": 0.41211554204186646,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 1762
    },
    {
      "epoch": 0.41234943281487546,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 1763
    },
    {
      "epoch": 0.41258332358788447,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 1764
    },
    {
      "epoch": 0.41281721436089347,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 1765
    },
    {
      "epoch": 0.4130511051339025,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 1766
    },
    {
      "epoch": 0.4132849959069115,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 1767
    },
    {
      "epoch": 0.4135188866799205,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 1768
    },
    {
      "epoch": 0.4137527774529295,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 1769
    },
    {
      "epoch": 0.4139866682259385,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 1770
    },
    {
      "epoch": 0.4142205589989475,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 1771
    },
    {
      "epoch": 0.4144544497719565,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 1772
    },
    {
      "epoch": 0.4146883405449655,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 1773
    },
    {
      "epoch": 0.4149222313179745,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.26,
      "step": 1774
    },
    {
      "epoch": 0.4151561220909835,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 1775
    },
    {
      "epoch": 0.4153900128639925,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7862,
      "step": 1776
    },
    {
      "epoch": 0.4156239036370015,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 1777
    },
    {
      "epoch": 0.41585779441001053,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 1778
    },
    {
      "epoch": 0.41609168518301953,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2679,
      "step": 1779
    },
    {
      "epoch": 0.41632557595602854,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7631,
      "step": 1780
    },
    {
      "epoch": 0.41655946672903754,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2647,
      "step": 1781
    },
    {
      "epoch": 0.41679335750204655,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 1782
    },
    {
      "epoch": 0.41702724827505555,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 1783
    },
    {
      "epoch": 0.41726113904806456,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 1784
    },
    {
      "epoch": 0.41749502982107356,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 1785
    },
    {
      "epoch": 0.41772892059408256,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 1786
    },
    {
      "epoch": 0.41796281136709157,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 1787
    },
    {
      "epoch": 0.4181967021401006,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 1788
    },
    {
      "epoch": 0.4184305929131096,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 1789
    },
    {
      "epoch": 0.4186644836861186,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 1790
    },
    {
      "epoch": 0.4188983744591276,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 1791
    },
    {
      "epoch": 0.4191322652321366,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 1792
    },
    {
      "epoch": 0.4193661560051456,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 1793
    },
    {
      "epoch": 0.4196000467781546,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 1794
    },
    {
      "epoch": 0.4198339375511636,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0639,
      "step": 1795
    },
    {
      "epoch": 0.4200678283241726,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 1796
    },
    {
      "epoch": 0.4203017190971816,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 1797
    },
    {
      "epoch": 0.4205356098701906,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2623,
      "step": 1798
    },
    {
      "epoch": 0.4207695006431996,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 1799
    },
    {
      "epoch": 0.4210033914162086,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 1800
    },
    {
      "epoch": 0.4210033914162086,
      "eval_runtime": 4.6184,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 1800
    },
    {
      "epoch": 0.42123728218921763,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 1801
    },
    {
      "epoch": 0.42147117296222664,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 1802
    },
    {
      "epoch": 0.42170506373523564,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 1803
    },
    {
      "epoch": 0.42193895450824465,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6533,
      "step": 1804
    },
    {
      "epoch": 0.42217284528125365,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 1805
    },
    {
      "epoch": 0.42240673605426265,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 1806
    },
    {
      "epoch": 0.42264062682727166,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 1807
    },
    {
      "epoch": 0.42287451760028066,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 1808
    },
    {
      "epoch": 0.42310840837328967,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 1809
    },
    {
      "epoch": 0.4233422991462987,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6383,
      "step": 1810
    },
    {
      "epoch": 0.4235761899193077,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 1811
    },
    {
      "epoch": 0.4238100806923167,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 1812
    },
    {
      "epoch": 0.4240439714653257,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 1813
    },
    {
      "epoch": 0.4242778622383347,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 1814
    },
    {
      "epoch": 0.4245117530113437,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 1815
    },
    {
      "epoch": 0.4247456437843527,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 1816
    },
    {
      "epoch": 0.4249795345573617,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 1817
    },
    {
      "epoch": 0.4252134253303707,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 1818
    },
    {
      "epoch": 0.4254473161033797,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 1819
    },
    {
      "epoch": 0.4256812068763887,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3609,
      "step": 1820
    },
    {
      "epoch": 0.4259150976493977,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 1821
    },
    {
      "epoch": 0.4261489884224067,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6652,
      "step": 1822
    },
    {
      "epoch": 0.42638287919541573,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 1823
    },
    {
      "epoch": 0.42661676996842474,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6782,
      "step": 1824
    },
    {
      "epoch": 0.42685066074143374,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 1825
    },
    {
      "epoch": 0.42708455151444275,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 1826
    },
    {
      "epoch": 0.42731844228745175,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 1827
    },
    {
      "epoch": 0.42755233306046075,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 1828
    },
    {
      "epoch": 0.42778622383346976,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 1829
    },
    {
      "epoch": 0.42802011460647876,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 1830
    },
    {
      "epoch": 0.42825400537948777,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7332,
      "step": 1831
    },
    {
      "epoch": 0.42848789615249677,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 1832
    },
    {
      "epoch": 0.4287217869255058,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3849,
      "step": 1833
    },
    {
      "epoch": 0.4289556776985148,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.6035,
      "step": 1834
    },
    {
      "epoch": 0.4291895684715238,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 1835
    },
    {
      "epoch": 0.4294234592445328,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 1836
    },
    {
      "epoch": 0.4296573500175418,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 1837
    },
    {
      "epoch": 0.4298912407905508,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 1838
    },
    {
      "epoch": 0.4301251315635598,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 1839
    },
    {
      "epoch": 0.4303590223365688,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 1840
    },
    {
      "epoch": 0.4305929131095778,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 1841
    },
    {
      "epoch": 0.4308268038825868,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 1842
    },
    {
      "epoch": 0.4310606946555958,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 1843
    },
    {
      "epoch": 0.4312945854286048,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 1844
    },
    {
      "epoch": 0.43152847620161383,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8329,
      "step": 1845
    },
    {
      "epoch": 0.43176236697462284,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 1846
    },
    {
      "epoch": 0.43199625774763184,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 1847
    },
    {
      "epoch": 0.43223014852064084,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6132,
      "step": 1848
    },
    {
      "epoch": 0.43246403929364985,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 1849
    },
    {
      "epoch": 0.43269793006665885,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2908,
      "step": 1850
    },
    {
      "epoch": 0.43293182083966786,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6868,
      "step": 1851
    },
    {
      "epoch": 0.43316571161267686,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 1852
    },
    {
      "epoch": 0.43339960238568587,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 1853
    },
    {
      "epoch": 0.43363349315869487,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 1854
    },
    {
      "epoch": 0.4338673839317039,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 1855
    },
    {
      "epoch": 0.4341012747047129,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 1856
    },
    {
      "epoch": 0.4343351654777219,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 1857
    },
    {
      "epoch": 0.4345690562507309,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 1858
    },
    {
      "epoch": 0.4348029470237399,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2709,
      "step": 1859
    },
    {
      "epoch": 0.4350368377967489,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 1860
    },
    {
      "epoch": 0.4352707285697579,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 1861
    },
    {
      "epoch": 0.4355046193427669,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 1862
    },
    {
      "epoch": 0.4357385101157759,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 1863
    },
    {
      "epoch": 0.4359724008887849,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 1864
    },
    {
      "epoch": 0.4362062916617939,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 1865
    },
    {
      "epoch": 0.4364401824348029,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 1866
    },
    {
      "epoch": 0.43667407320781193,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 1867
    },
    {
      "epoch": 0.43690796398082093,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.782,
      "step": 1868
    },
    {
      "epoch": 0.43714185475382994,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 1869
    },
    {
      "epoch": 0.43737574552683894,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 1870
    },
    {
      "epoch": 0.43760963629984795,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 1871
    },
    {
      "epoch": 0.43784352707285695,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 1872
    },
    {
      "epoch": 0.43807741784586596,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 1873
    },
    {
      "epoch": 0.43831130861887496,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 1874
    },
    {
      "epoch": 0.43854519939188397,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 1875
    },
    {
      "epoch": 0.43877909016489297,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 1876
    },
    {
      "epoch": 0.439012980937902,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 1877
    },
    {
      "epoch": 0.439246871710911,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 1878
    },
    {
      "epoch": 0.43948076248392,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.9136,
      "step": 1879
    },
    {
      "epoch": 0.439714653256929,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 1880
    },
    {
      "epoch": 0.439948544029938,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 1881
    },
    {
      "epoch": 0.440182434802947,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 1882
    },
    {
      "epoch": 0.440416325575956,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 1883
    },
    {
      "epoch": 0.440650216348965,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 1884
    },
    {
      "epoch": 0.440884107121974,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7411,
      "step": 1885
    },
    {
      "epoch": 0.441117997894983,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 1886
    },
    {
      "epoch": 0.441351888667992,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 1887
    },
    {
      "epoch": 0.441585779441001,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 1888
    },
    {
      "epoch": 0.44181967021401003,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 1889
    },
    {
      "epoch": 0.4420535609870191,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 1890
    },
    {
      "epoch": 0.4422874517600281,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6512,
      "step": 1891
    },
    {
      "epoch": 0.4425213425330371,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 1892
    },
    {
      "epoch": 0.4427552333060461,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 1893
    },
    {
      "epoch": 0.4429891240790551,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 1894
    },
    {
      "epoch": 0.4432230148520641,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 1895
    },
    {
      "epoch": 0.4434569056250731,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 1896
    },
    {
      "epoch": 0.4436907963980821,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2809,
      "step": 1897
    },
    {
      "epoch": 0.4439246871710911,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 1898
    },
    {
      "epoch": 0.44415857794410013,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 1899
    },
    {
      "epoch": 0.44439246871710913,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 1900
    },
    {
      "epoch": 0.44439246871710913,
      "eval_runtime": 4.6513,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 1900
    },
    {
      "epoch": 0.44462635949011814,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 1901
    },
    {
      "epoch": 0.44486025026312714,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 1902
    },
    {
      "epoch": 0.44509414103613615,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5736,
      "step": 1903
    },
    {
      "epoch": 0.44532803180914515,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 1904
    },
    {
      "epoch": 0.44556192258215416,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 1905
    },
    {
      "epoch": 0.44579581335516316,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 1906
    },
    {
      "epoch": 0.44602970412817217,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 1907
    },
    {
      "epoch": 0.44626359490118117,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 1908
    },
    {
      "epoch": 0.4464974856741902,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 1909
    },
    {
      "epoch": 0.4467313764471992,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7229,
      "step": 1910
    },
    {
      "epoch": 0.4469652672202082,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 1911
    },
    {
      "epoch": 0.4471991579932172,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7359,
      "step": 1912
    },
    {
      "epoch": 0.4474330487662262,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 1913
    },
    {
      "epoch": 0.4476669395392352,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 1914
    },
    {
      "epoch": 0.4479008303122442,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 1915
    },
    {
      "epoch": 0.4481347210852532,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 1916
    },
    {
      "epoch": 0.4483686118582622,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.299,
      "step": 1917
    },
    {
      "epoch": 0.4486025026312712,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 1918
    },
    {
      "epoch": 0.4488363934042802,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 1919
    },
    {
      "epoch": 0.4490702841772892,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 1920
    },
    {
      "epoch": 0.44930417495029823,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7814,
      "step": 1921
    },
    {
      "epoch": 0.44953806572330723,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 1922
    },
    {
      "epoch": 0.44977195649631624,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6657,
      "step": 1923
    },
    {
      "epoch": 0.45000584726932524,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 1924
    },
    {
      "epoch": 0.45023973804233425,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 1925
    },
    {
      "epoch": 0.45047362881534325,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 1926
    },
    {
      "epoch": 0.45070751958835226,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 1927
    },
    {
      "epoch": 0.45094141036136126,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 1928
    },
    {
      "epoch": 0.45117530113437027,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 1929
    },
    {
      "epoch": 0.45140919190737927,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 1930
    },
    {
      "epoch": 0.4516430826803883,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 1931
    },
    {
      "epoch": 0.4518769734533973,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 1932
    },
    {
      "epoch": 0.4521108642264063,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3536,
      "step": 1933
    },
    {
      "epoch": 0.4523447549994153,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 1934
    },
    {
      "epoch": 0.4525786457724243,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 1935
    },
    {
      "epoch": 0.4528125365454333,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 1936
    },
    {
      "epoch": 0.4530464273184423,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 1937
    },
    {
      "epoch": 0.4532803180914513,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 1938
    },
    {
      "epoch": 0.4535142088644603,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 1939
    },
    {
      "epoch": 0.4537480996374693,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 1940
    },
    {
      "epoch": 0.4539819904104783,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7256,
      "step": 1941
    },
    {
      "epoch": 0.4542158811834873,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 1942
    },
    {
      "epoch": 0.45444977195649633,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 1943
    },
    {
      "epoch": 0.45468366272950533,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 1944
    },
    {
      "epoch": 0.45491755350251434,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 1945
    },
    {
      "epoch": 0.45515144427552334,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 1946
    },
    {
      "epoch": 0.45538533504853235,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 1947
    },
    {
      "epoch": 0.45561922582154135,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 1948
    },
    {
      "epoch": 0.45585311659455036,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 1949
    },
    {
      "epoch": 0.45608700736755936,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 1950
    },
    {
      "epoch": 0.45632089814056837,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3554,
      "step": 1951
    },
    {
      "epoch": 0.45655478891357737,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 1952
    },
    {
      "epoch": 0.4567886796865864,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 1953
    },
    {
      "epoch": 0.4570225704595954,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 1954
    },
    {
      "epoch": 0.4572564612326044,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 1955
    },
    {
      "epoch": 0.4574903520056134,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8988,
      "step": 1956
    },
    {
      "epoch": 0.4577242427786224,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 1957
    },
    {
      "epoch": 0.4579581335516314,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 1958
    },
    {
      "epoch": 0.4581920243246404,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 1959
    },
    {
      "epoch": 0.4584259150976494,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 1960
    },
    {
      "epoch": 0.4586598058706584,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7595,
      "step": 1961
    },
    {
      "epoch": 0.4588936966436674,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 1962
    },
    {
      "epoch": 0.4591275874166764,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 1963
    },
    {
      "epoch": 0.4593614781896854,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 1964
    },
    {
      "epoch": 0.45959536896269443,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 1965
    },
    {
      "epoch": 0.45982925973570343,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 1966
    },
    {
      "epoch": 0.46006315050871244,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 1967
    },
    {
      "epoch": 0.46029704128172144,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 1968
    },
    {
      "epoch": 0.46053093205473045,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 1969
    },
    {
      "epoch": 0.46076482282773945,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 1970
    },
    {
      "epoch": 0.46099871360074846,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 1971
    },
    {
      "epoch": 0.46123260437375746,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 1972
    },
    {
      "epoch": 0.46146649514676646,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 1973
    },
    {
      "epoch": 0.46170038591977547,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 1974
    },
    {
      "epoch": 0.4619342766927845,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3262,
      "step": 1975
    },
    {
      "epoch": 0.4621681674657935,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 1976
    },
    {
      "epoch": 0.4624020582388025,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 1977
    },
    {
      "epoch": 0.4626359490118115,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 1978
    },
    {
      "epoch": 0.4628698397848205,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 1979
    },
    {
      "epoch": 0.4631037305578295,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 1980
    },
    {
      "epoch": 0.4633376213308385,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 1981
    },
    {
      "epoch": 0.4635715121038475,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 1982
    },
    {
      "epoch": 0.4638054028768565,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 1983
    },
    {
      "epoch": 0.4640392936498655,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 1984
    },
    {
      "epoch": 0.4642731844228745,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 1985
    },
    {
      "epoch": 0.4645070751958835,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.6502,
      "step": 1986
    },
    {
      "epoch": 0.4647409659688925,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 1987
    },
    {
      "epoch": 0.46497485674190153,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 1988
    },
    {
      "epoch": 0.46520874751491054,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 1989
    },
    {
      "epoch": 0.46544263828791954,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 1990
    },
    {
      "epoch": 0.46567652906092855,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7234,
      "step": 1991
    },
    {
      "epoch": 0.46591041983393755,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 1992
    },
    {
      "epoch": 0.46614431060694655,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 1993
    },
    {
      "epoch": 0.46637820137995556,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 1994
    },
    {
      "epoch": 0.46661209215296456,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 1995
    },
    {
      "epoch": 0.46684598292597357,
      "grad_norm": 7.75,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 1996
    },
    {
      "epoch": 0.4670798736989826,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8424,
      "step": 1997
    },
    {
      "epoch": 0.4673137644719916,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 1998
    },
    {
      "epoch": 0.4675476552450006,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 1999
    },
    {
      "epoch": 0.4677815460180096,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5217,
      "step": 2000
    },
    {
      "epoch": 0.4677815460180096,
      "eval_runtime": 4.6182,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 2000
    },
    {
      "epoch": 0.4680154367910186,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 2001
    },
    {
      "epoch": 0.4682493275640276,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 2002
    },
    {
      "epoch": 0.4684832183370366,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 2003
    },
    {
      "epoch": 0.4687171091100456,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 2004
    },
    {
      "epoch": 0.4689509998830546,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 2005
    },
    {
      "epoch": 0.4691848906560636,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6826,
      "step": 2006
    },
    {
      "epoch": 0.4694187814290726,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 2007
    },
    {
      "epoch": 0.4696526722020816,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 2008
    },
    {
      "epoch": 0.4698865629750906,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 2009
    },
    {
      "epoch": 0.47012045374809963,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 2010
    },
    {
      "epoch": 0.47035434452110864,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 2011
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 2012
    },
    {
      "epoch": 0.47082212606712665,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1469,
      "step": 2013
    },
    {
      "epoch": 0.47105601684013565,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 2014
    },
    {
      "epoch": 0.47128990761314465,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 2015
    },
    {
      "epoch": 0.47152379838615366,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 2016
    },
    {
      "epoch": 0.47175768915916266,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 2017
    },
    {
      "epoch": 0.47199157993217167,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.6913,
      "step": 2018
    },
    {
      "epoch": 0.47222547070518067,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4438,
      "step": 2019
    },
    {
      "epoch": 0.4724593614781897,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 2020
    },
    {
      "epoch": 0.4726932522511987,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 2021
    },
    {
      "epoch": 0.4729271430242077,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 2022
    },
    {
      "epoch": 0.4731610337972167,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 2023
    },
    {
      "epoch": 0.4733949245702257,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 2024
    },
    {
      "epoch": 0.4736288153432347,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6261,
      "step": 2025
    },
    {
      "epoch": 0.4738627061162437,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6388,
      "step": 2026
    },
    {
      "epoch": 0.4740965968892527,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 2027
    },
    {
      "epoch": 0.4743304876622617,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 2028
    },
    {
      "epoch": 0.4745643784352707,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8153,
      "step": 2029
    },
    {
      "epoch": 0.4747982692082797,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6667,
      "step": 2030
    },
    {
      "epoch": 0.4750321599812887,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 2031
    },
    {
      "epoch": 0.47526605075429773,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0032,
      "step": 2032
    },
    {
      "epoch": 0.47549994152730674,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2749,
      "step": 2033
    },
    {
      "epoch": 0.47573383230031574,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 2034
    },
    {
      "epoch": 0.47596772307332474,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5188,
      "step": 2035
    },
    {
      "epoch": 0.47620161384633375,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 2036
    },
    {
      "epoch": 0.47643550461934275,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2143,
      "step": 2037
    },
    {
      "epoch": 0.47666939539235176,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 2038
    },
    {
      "epoch": 0.47690328616536076,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 2039
    },
    {
      "epoch": 0.47713717693836977,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 2040
    },
    {
      "epoch": 0.47737106771137877,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 2041
    },
    {
      "epoch": 0.4776049584843878,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 2042
    },
    {
      "epoch": 0.4778388492573968,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 2043
    },
    {
      "epoch": 0.4780727400304058,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 2044
    },
    {
      "epoch": 0.4783066308034148,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2143,
      "step": 2045
    },
    {
      "epoch": 0.4785405215764238,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 2046
    },
    {
      "epoch": 0.4787744123494328,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2987,
      "step": 2047
    },
    {
      "epoch": 0.4790083031224418,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7777,
      "step": 2048
    },
    {
      "epoch": 0.4792421938954508,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 2049
    },
    {
      "epoch": 0.4794760846684598,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 2050
    },
    {
      "epoch": 0.4797099754414688,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 2051
    },
    {
      "epoch": 0.4799438662144778,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 2052
    },
    {
      "epoch": 0.4801777569874868,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 2053
    },
    {
      "epoch": 0.48041164776049583,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 2054
    },
    {
      "epoch": 0.48064553853350483,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 2055
    },
    {
      "epoch": 0.48087942930651384,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 2056
    },
    {
      "epoch": 0.48111332007952284,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 2057
    },
    {
      "epoch": 0.48134721085253185,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 2058
    },
    {
      "epoch": 0.48158110162554085,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3703,
      "step": 2059
    },
    {
      "epoch": 0.48181499239854986,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2903,
      "step": 2060
    },
    {
      "epoch": 0.48204888317155886,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 2061
    },
    {
      "epoch": 0.48228277394456787,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 2062
    },
    {
      "epoch": 0.48251666471757687,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 2063
    },
    {
      "epoch": 0.4827505554905859,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0913,
      "step": 2064
    },
    {
      "epoch": 0.4829844462635949,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3273,
      "step": 2065
    },
    {
      "epoch": 0.4832183370366039,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 2066
    },
    {
      "epoch": 0.4834522278096129,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 2067
    },
    {
      "epoch": 0.4836861185826219,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 2068
    },
    {
      "epoch": 0.4839200093556309,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 2069
    },
    {
      "epoch": 0.4841539001286399,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 2070
    },
    {
      "epoch": 0.4843877909016489,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 2071
    },
    {
      "epoch": 0.4846216816746579,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 2072
    },
    {
      "epoch": 0.4848555724476669,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 2073
    },
    {
      "epoch": 0.4850894632206759,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 2074
    },
    {
      "epoch": 0.4853233539936849,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 2075
    },
    {
      "epoch": 0.48555724476669393,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 2076
    },
    {
      "epoch": 0.48579113553970293,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 2077
    },
    {
      "epoch": 0.48602502631271194,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 2078
    },
    {
      "epoch": 0.48625891708572094,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 2079
    },
    {
      "epoch": 0.48649280785872995,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 2080
    },
    {
      "epoch": 0.48672669863173895,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 2081
    },
    {
      "epoch": 0.48696058940474796,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 2082
    },
    {
      "epoch": 0.48719448017775696,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 2083
    },
    {
      "epoch": 0.48742837095076597,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 2084
    },
    {
      "epoch": 0.48766226172377497,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9657,
      "step": 2085
    },
    {
      "epoch": 0.487896152496784,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1838,
      "step": 2086
    },
    {
      "epoch": 0.488130043269793,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 2087
    },
    {
      "epoch": 0.488363934042802,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1198,
      "step": 2088
    },
    {
      "epoch": 0.48859782481581104,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 2089
    },
    {
      "epoch": 0.48883171558882005,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 2090
    },
    {
      "epoch": 0.48906560636182905,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 2091
    },
    {
      "epoch": 0.48929949713483806,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 2092
    },
    {
      "epoch": 0.48953338790784706,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 2093
    },
    {
      "epoch": 0.48976727868085607,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 2094
    },
    {
      "epoch": 0.49000116945386507,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 2095
    },
    {
      "epoch": 0.4902350602268741,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 2096
    },
    {
      "epoch": 0.4904689509998831,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 2097
    },
    {
      "epoch": 0.4907028417728921,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 2098
    },
    {
      "epoch": 0.4909367325459011,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 2099
    },
    {
      "epoch": 0.4911706233189101,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 2100
    },
    {
      "epoch": 0.4911706233189101,
      "eval_runtime": 4.6025,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 2100
    },
    {
      "epoch": 0.4914045140919191,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6514,
      "step": 2101
    },
    {
      "epoch": 0.4916384048649281,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 2102
    },
    {
      "epoch": 0.4918722956379371,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 2103
    },
    {
      "epoch": 0.4921061864109461,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7308,
      "step": 2104
    },
    {
      "epoch": 0.4923400771839551,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3778,
      "step": 2105
    },
    {
      "epoch": 0.4925739679569641,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 2106
    },
    {
      "epoch": 0.4928078587299731,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5944,
      "step": 2107
    },
    {
      "epoch": 0.49304174950298213,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 2108
    },
    {
      "epoch": 0.49327564027599113,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 2109
    },
    {
      "epoch": 0.49350953104900014,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 2110
    },
    {
      "epoch": 0.49374342182200914,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 2111
    },
    {
      "epoch": 0.49397731259501815,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8073,
      "step": 2112
    },
    {
      "epoch": 0.49421120336802715,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6897,
      "step": 2113
    },
    {
      "epoch": 0.49444509414103616,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 2114
    },
    {
      "epoch": 0.49467898491404516,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 2115
    },
    {
      "epoch": 0.49491287568705417,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 2116
    },
    {
      "epoch": 0.49514676646006317,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 2117
    },
    {
      "epoch": 0.4953806572330722,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 2118
    },
    {
      "epoch": 0.4956145480060812,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2574,
      "step": 2119
    },
    {
      "epoch": 0.4958484387790902,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3926,
      "step": 2120
    },
    {
      "epoch": 0.4960823295520992,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 2121
    },
    {
      "epoch": 0.4963162203251082,
      "grad_norm": 7.5625,
      "learning_rate": 3e-05,
      "loss": 1.8714,
      "step": 2122
    },
    {
      "epoch": 0.4965501110981172,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2666,
      "step": 2123
    },
    {
      "epoch": 0.4967840018711262,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 2124
    },
    {
      "epoch": 0.4970178926441352,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.257,
      "step": 2125
    },
    {
      "epoch": 0.4972517834171442,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 2126
    },
    {
      "epoch": 0.4974856741901532,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8751,
      "step": 2127
    },
    {
      "epoch": 0.4977195649631622,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 2128
    },
    {
      "epoch": 0.4979534557361712,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 2129
    },
    {
      "epoch": 0.49818734650918023,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 2130
    },
    {
      "epoch": 0.49842123728218923,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 2131
    },
    {
      "epoch": 0.49865512805519824,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 2132
    },
    {
      "epoch": 0.49888901882820724,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 2133
    },
    {
      "epoch": 0.49912290960121625,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 2134
    },
    {
      "epoch": 0.49935680037422525,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0265,
      "step": 2135
    },
    {
      "epoch": 0.49959069114723426,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 2136
    },
    {
      "epoch": 0.49982458192024326,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 2137
    },
    {
      "epoch": 0.5000584726932522,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 2138
    },
    {
      "epoch": 0.5002923634662613,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 2139
    },
    {
      "epoch": 0.5005262542392702,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.798,
      "step": 2140
    },
    {
      "epoch": 0.5007601450122793,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 2141
    },
    {
      "epoch": 0.5009940357852882,
      "grad_norm": 9.0,
      "learning_rate": 3e-05,
      "loss": 2.1867,
      "step": 2142
    },
    {
      "epoch": 0.5012279265582973,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 2143
    },
    {
      "epoch": 0.5014618173313062,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2076,
      "step": 2144
    },
    {
      "epoch": 0.5016957081043153,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2391,
      "step": 2145
    },
    {
      "epoch": 0.5019295988773242,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 2146
    },
    {
      "epoch": 0.5021634896503333,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 2147
    },
    {
      "epoch": 0.5023973804233423,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 2148
    },
    {
      "epoch": 0.5026312711963513,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5552,
      "step": 2149
    },
    {
      "epoch": 0.5028651619693603,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 2150
    },
    {
      "epoch": 0.5030990527423693,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2707,
      "step": 2151
    },
    {
      "epoch": 0.5033329435153783,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6998,
      "step": 2152
    },
    {
      "epoch": 0.5035668342883873,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7554,
      "step": 2153
    },
    {
      "epoch": 0.5038007250613963,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 2154
    },
    {
      "epoch": 0.5040346158344053,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7488,
      "step": 2155
    },
    {
      "epoch": 0.5042685066074143,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 2156
    },
    {
      "epoch": 0.5045023973804234,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 2157
    },
    {
      "epoch": 0.5047362881534323,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5818,
      "step": 2158
    },
    {
      "epoch": 0.5049701789264414,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8136,
      "step": 2159
    },
    {
      "epoch": 0.5052040696994503,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 2160
    },
    {
      "epoch": 0.5054379604724594,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 2161
    },
    {
      "epoch": 0.5056718512454683,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3053,
      "step": 2162
    },
    {
      "epoch": 0.5059057420184774,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 2163
    },
    {
      "epoch": 0.5061396327914863,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.3793,
      "step": 2164
    },
    {
      "epoch": 0.5063735235644954,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 2165
    },
    {
      "epoch": 0.5066074143375043,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 2166
    },
    {
      "epoch": 0.5068413051105134,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 2167
    },
    {
      "epoch": 0.5070751958835223,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 2168
    },
    {
      "epoch": 0.5073090866565314,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5285,
      "step": 2169
    },
    {
      "epoch": 0.5075429774295404,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 2170
    },
    {
      "epoch": 0.5077768682025494,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 2171
    },
    {
      "epoch": 0.5080107589755584,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 2172
    },
    {
      "epoch": 0.5082446497485674,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 2173
    },
    {
      "epoch": 0.5084785405215764,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 2174
    },
    {
      "epoch": 0.5087124312945854,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 2175
    },
    {
      "epoch": 0.5089463220675944,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3825,
      "step": 2176
    },
    {
      "epoch": 0.5091802128406034,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 2177
    },
    {
      "epoch": 0.5094141036136124,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3098,
      "step": 2178
    },
    {
      "epoch": 0.5096479943866215,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7687,
      "step": 2179
    },
    {
      "epoch": 0.5098818851596304,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 2180
    },
    {
      "epoch": 0.5101157759326395,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 2181
    },
    {
      "epoch": 0.5103496667056484,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 2182
    },
    {
      "epoch": 0.5105835574786575,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 2183
    },
    {
      "epoch": 0.5108174482516664,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 2184
    },
    {
      "epoch": 0.5110513390246755,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 2185
    },
    {
      "epoch": 0.5112852297976844,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 2186
    },
    {
      "epoch": 0.5115191205706935,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.344,
      "step": 2187
    },
    {
      "epoch": 0.5117530113437025,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 2188
    },
    {
      "epoch": 0.5119869021167115,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 2189
    },
    {
      "epoch": 0.5122207928897206,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 2190
    },
    {
      "epoch": 0.5124546836627295,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8897,
      "step": 2191
    },
    {
      "epoch": 0.5126885744357386,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 2192
    },
    {
      "epoch": 0.5129224652087475,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 2193
    },
    {
      "epoch": 0.5131563559817566,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 2194
    },
    {
      "epoch": 0.5133902467547655,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7485,
      "step": 2195
    },
    {
      "epoch": 0.5136241375277746,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 2196
    },
    {
      "epoch": 0.5138580283007835,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 2197
    },
    {
      "epoch": 0.5140919190737926,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 2198
    },
    {
      "epoch": 0.5143258098468015,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 2199
    },
    {
      "epoch": 0.5145597006198106,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 2200
    },
    {
      "epoch": 0.5145597006198106,
      "eval_runtime": 4.5773,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 2200
    },
    {
      "epoch": 0.5147935913928195,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 2201
    },
    {
      "epoch": 0.5150274821658286,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 2202
    },
    {
      "epoch": 0.5152613729388376,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 2203
    },
    {
      "epoch": 0.5154952637118466,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 2204
    },
    {
      "epoch": 0.5157291544848556,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.289,
      "step": 2205
    },
    {
      "epoch": 0.5159630452578646,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 2206
    },
    {
      "epoch": 0.5161969360308736,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 2207
    },
    {
      "epoch": 0.5164308268038826,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 2208
    },
    {
      "epoch": 0.5166647175768916,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7758,
      "step": 2209
    },
    {
      "epoch": 0.5168986083499006,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 2210
    },
    {
      "epoch": 0.5171324991229096,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 2211
    },
    {
      "epoch": 0.5173663898959187,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 2212
    },
    {
      "epoch": 0.5176002806689276,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1233,
      "step": 2213
    },
    {
      "epoch": 0.5178341714419367,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 2214
    },
    {
      "epoch": 0.5180680622149456,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 2215
    },
    {
      "epoch": 0.5183019529879547,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 2216
    },
    {
      "epoch": 0.5185358437609636,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 2217
    },
    {
      "epoch": 0.5187697345339727,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2614,
      "step": 2218
    },
    {
      "epoch": 0.5190036253069816,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 2219
    },
    {
      "epoch": 0.5192375160799907,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 2220
    },
    {
      "epoch": 0.5194714068529996,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 2221
    },
    {
      "epoch": 0.5197052976260087,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 2222
    },
    {
      "epoch": 0.5199391883990176,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 2223
    },
    {
      "epoch": 0.5201730791720267,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 2224
    },
    {
      "epoch": 0.5204069699450357,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 2225
    },
    {
      "epoch": 0.5206408607180447,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 2226
    },
    {
      "epoch": 0.5208747514910537,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8721,
      "step": 2227
    },
    {
      "epoch": 0.5211086422640627,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 2228
    },
    {
      "epoch": 0.5213425330370717,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 2229
    },
    {
      "epoch": 0.5215764238100807,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 2230
    },
    {
      "epoch": 0.5218103145830897,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 2231
    },
    {
      "epoch": 0.5220442053560987,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 2232
    },
    {
      "epoch": 0.5222780961291077,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.5616,
      "step": 2233
    },
    {
      "epoch": 0.5225119869021168,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 2234
    },
    {
      "epoch": 0.5227458776751257,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 2235
    },
    {
      "epoch": 0.5229797684481348,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 2236
    },
    {
      "epoch": 0.5232136592211437,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 2237
    },
    {
      "epoch": 0.5234475499941528,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 2238
    },
    {
      "epoch": 0.5236814407671617,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 2239
    },
    {
      "epoch": 0.5239153315401708,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 2240
    },
    {
      "epoch": 0.5241492223131797,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 2241
    },
    {
      "epoch": 0.5243831130861888,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 2242
    },
    {
      "epoch": 0.5246170038591977,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 2243
    },
    {
      "epoch": 0.5248508946322068,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 2244
    },
    {
      "epoch": 0.5250847854052157,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 2245
    },
    {
      "epoch": 0.5253186761782248,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 2246
    },
    {
      "epoch": 0.5255525669512338,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 2247
    },
    {
      "epoch": 0.5257864577242428,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 2248
    },
    {
      "epoch": 0.5260203484972518,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 2249
    },
    {
      "epoch": 0.5262542392702608,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6523,
      "step": 2250
    },
    {
      "epoch": 0.5264881300432698,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1819,
      "step": 2251
    },
    {
      "epoch": 0.5267220208162788,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 2252
    },
    {
      "epoch": 0.5269559115892878,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 2253
    },
    {
      "epoch": 0.5271898023622968,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5904,
      "step": 2254
    },
    {
      "epoch": 0.5274236931353058,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 2255
    },
    {
      "epoch": 0.5276575839083149,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 2256
    },
    {
      "epoch": 0.5278914746813238,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6603,
      "step": 2257
    },
    {
      "epoch": 0.5281253654543329,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 2258
    },
    {
      "epoch": 0.5283592562273418,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 2259
    },
    {
      "epoch": 0.5285931470003509,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 2260
    },
    {
      "epoch": 0.5288270377733598,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 2261
    },
    {
      "epoch": 0.5290609285463689,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 2262
    },
    {
      "epoch": 0.5292948193193778,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 2263
    },
    {
      "epoch": 0.5295287100923869,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 2264
    },
    {
      "epoch": 0.5297626008653958,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2878,
      "step": 2265
    },
    {
      "epoch": 0.5299964916384049,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 2266
    },
    {
      "epoch": 0.5302303824114138,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 2267
    },
    {
      "epoch": 0.5304642731844229,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 2268
    },
    {
      "epoch": 0.5306981639574319,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 2269
    },
    {
      "epoch": 0.5309320547304409,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 2270
    },
    {
      "epoch": 0.5311659455034499,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7392,
      "step": 2271
    },
    {
      "epoch": 0.5313998362764589,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 2272
    },
    {
      "epoch": 0.5316337270494679,
      "grad_norm": 7.59375,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 2273
    },
    {
      "epoch": 0.5318676178224769,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 2274
    },
    {
      "epoch": 0.5321015085954859,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 2275
    },
    {
      "epoch": 0.5323353993684949,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 2276
    },
    {
      "epoch": 0.5325692901415039,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 2277
    },
    {
      "epoch": 0.532803180914513,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 2278
    },
    {
      "epoch": 0.5330370716875219,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 2279
    },
    {
      "epoch": 0.533270962460531,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3014,
      "step": 2280
    },
    {
      "epoch": 0.5335048532335399,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 2281
    },
    {
      "epoch": 0.533738744006549,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 2282
    },
    {
      "epoch": 0.5339726347795579,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 2283
    },
    {
      "epoch": 0.534206525552567,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7934,
      "step": 2284
    },
    {
      "epoch": 0.5344404163255759,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 2285
    },
    {
      "epoch": 0.534674307098585,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 2286
    },
    {
      "epoch": 0.5349081978715939,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0472,
      "step": 2287
    },
    {
      "epoch": 0.535142088644603,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 2288
    },
    {
      "epoch": 0.535375979417612,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8721,
      "step": 2289
    },
    {
      "epoch": 0.535609870190621,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 2290
    },
    {
      "epoch": 0.53584376096363,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 2291
    },
    {
      "epoch": 0.536077651736639,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 2292
    },
    {
      "epoch": 0.536311542509648,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3041,
      "step": 2293
    },
    {
      "epoch": 0.536545433282657,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 2294
    },
    {
      "epoch": 0.536779324055666,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 2295
    },
    {
      "epoch": 0.537013214828675,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7732,
      "step": 2296
    },
    {
      "epoch": 0.537247105601684,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 2297
    },
    {
      "epoch": 0.537480996374693,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 2298
    },
    {
      "epoch": 0.537714887147702,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 2299
    },
    {
      "epoch": 0.537948777920711,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 2300
    },
    {
      "epoch": 0.537948777920711,
      "eval_runtime": 4.5918,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 2300
    },
    {
      "epoch": 0.53818266869372,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 2301
    },
    {
      "epoch": 0.5384165594667291,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 2302
    },
    {
      "epoch": 0.538650450239738,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 2303
    },
    {
      "epoch": 0.5388843410127471,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 2304
    },
    {
      "epoch": 0.539118231785756,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4219,
      "step": 2305
    },
    {
      "epoch": 0.5393521225587651,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 2306
    },
    {
      "epoch": 0.539586013331774,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9751,
      "step": 2307
    },
    {
      "epoch": 0.5398199041047831,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 2308
    },
    {
      "epoch": 0.540053794877792,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 2309
    },
    {
      "epoch": 0.5402876856508011,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 2310
    },
    {
      "epoch": 0.54052157642381,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2076,
      "step": 2311
    },
    {
      "epoch": 0.5407554671968191,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 2312
    },
    {
      "epoch": 0.540989357969828,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 2313
    },
    {
      "epoch": 0.5412232487428371,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 2314
    },
    {
      "epoch": 0.5414571395158461,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 2315
    },
    {
      "epoch": 0.5416910302888551,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2357,
      "step": 2316
    },
    {
      "epoch": 0.5419249210618641,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 2317
    },
    {
      "epoch": 0.5421588118348731,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 2318
    },
    {
      "epoch": 0.5423927026078821,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 2319
    },
    {
      "epoch": 0.5426265933808911,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6204,
      "step": 2320
    },
    {
      "epoch": 0.5428604841539001,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 2321
    },
    {
      "epoch": 0.5430943749269092,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7074,
      "step": 2322
    },
    {
      "epoch": 0.5433282656999181,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 2323
    },
    {
      "epoch": 0.5435621564729272,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 2324
    },
    {
      "epoch": 0.5437960472459361,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 2325
    },
    {
      "epoch": 0.5440299380189452,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 2326
    },
    {
      "epoch": 0.5442638287919541,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 2327
    },
    {
      "epoch": 0.5444977195649632,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 2328
    },
    {
      "epoch": 0.5447316103379721,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 2329
    },
    {
      "epoch": 0.5449655011109812,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 2330
    },
    {
      "epoch": 0.5451993918839901,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 2331
    },
    {
      "epoch": 0.5454332826569992,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 2332
    },
    {
      "epoch": 0.5456671734300081,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 2333
    },
    {
      "epoch": 0.5459010642030172,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 2334
    },
    {
      "epoch": 0.5461349549760262,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 2335
    },
    {
      "epoch": 0.5463688457490352,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 2336
    },
    {
      "epoch": 0.5466027365220442,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 2337
    },
    {
      "epoch": 0.5468366272950532,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 2338
    },
    {
      "epoch": 0.5470705180680622,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 2339
    },
    {
      "epoch": 0.5473044088410712,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 2340
    },
    {
      "epoch": 0.5475382996140802,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 2341
    },
    {
      "epoch": 0.5477721903870892,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.623,
      "step": 2342
    },
    {
      "epoch": 0.5480060811600982,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7545,
      "step": 2343
    },
    {
      "epoch": 0.5482399719331073,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 2344
    },
    {
      "epoch": 0.5484738627061162,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 2345
    },
    {
      "epoch": 0.5487077534791253,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 2346
    },
    {
      "epoch": 0.5489416442521342,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 2347
    },
    {
      "epoch": 0.5491755350251433,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 2348
    },
    {
      "epoch": 0.5494094257981522,
      "grad_norm": 17.75,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 2349
    },
    {
      "epoch": 0.5496433165711613,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 2350
    },
    {
      "epoch": 0.5498772073441702,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6976,
      "step": 2351
    },
    {
      "epoch": 0.5501110981171793,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 2352
    },
    {
      "epoch": 0.5503449888901882,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 2353
    },
    {
      "epoch": 0.5505788796631973,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 2354
    },
    {
      "epoch": 0.5508127704362062,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 2355
    },
    {
      "epoch": 0.5510466612092153,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 2356
    },
    {
      "epoch": 0.5512805519822243,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 2357
    },
    {
      "epoch": 0.5515144427552333,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 2358
    },
    {
      "epoch": 0.5517483335282423,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 2359
    },
    {
      "epoch": 0.5519822243012513,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 2360
    },
    {
      "epoch": 0.5522161150742603,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 2361
    },
    {
      "epoch": 0.5524500058472693,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 2362
    },
    {
      "epoch": 0.5526838966202783,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 2363
    },
    {
      "epoch": 0.5529177873932873,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 2364
    },
    {
      "epoch": 0.5531516781662963,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 2365
    },
    {
      "epoch": 0.5533855689393054,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 2366
    },
    {
      "epoch": 0.5536194597123143,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 2367
    },
    {
      "epoch": 0.5538533504853234,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.037,
      "step": 2368
    },
    {
      "epoch": 0.5540872412583323,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 2369
    },
    {
      "epoch": 0.5543211320313414,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 2370
    },
    {
      "epoch": 0.5545550228043503,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 2371
    },
    {
      "epoch": 0.5547889135773594,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 2372
    },
    {
      "epoch": 0.5550228043503683,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 2373
    },
    {
      "epoch": 0.5552566951233774,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 2374
    },
    {
      "epoch": 0.5554905858963863,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 2375
    },
    {
      "epoch": 0.5557244766693954,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 2376
    },
    {
      "epoch": 0.5559583674424043,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 2377
    },
    {
      "epoch": 0.5561922582154134,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 2378
    },
    {
      "epoch": 0.5564261489884224,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 2379
    },
    {
      "epoch": 0.5566600397614314,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6605,
      "step": 2380
    },
    {
      "epoch": 0.5568939305344404,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3399,
      "step": 2381
    },
    {
      "epoch": 0.5571278213074494,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 2382
    },
    {
      "epoch": 0.5573617120804584,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 2383
    },
    {
      "epoch": 0.5575956028534674,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 2384
    },
    {
      "epoch": 0.5578294936264764,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 2385
    },
    {
      "epoch": 0.5580633843994854,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 2386
    },
    {
      "epoch": 0.5582972751724945,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 2387
    },
    {
      "epoch": 0.5585311659455034,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 2388
    },
    {
      "epoch": 0.5587650567185125,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 2389
    },
    {
      "epoch": 0.5589989474915215,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9831,
      "step": 2390
    },
    {
      "epoch": 0.5592328382645305,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 2391
    },
    {
      "epoch": 0.5594667290375395,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 2392
    },
    {
      "epoch": 0.5597006198105485,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 2393
    },
    {
      "epoch": 0.5599345105835575,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 2394
    },
    {
      "epoch": 0.5601684013565665,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 2395
    },
    {
      "epoch": 0.5604022921295755,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 2396
    },
    {
      "epoch": 0.5606361829025845,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 2397
    },
    {
      "epoch": 0.5608700736755935,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 2398
    },
    {
      "epoch": 0.5611039644486026,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 2399
    },
    {
      "epoch": 0.5613378552216115,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6532,
      "step": 2400
    },
    {
      "epoch": 0.5613378552216115,
      "eval_runtime": 4.6687,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 2400
    },
    {
      "epoch": 0.5615717459946206,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8028,
      "step": 2401
    },
    {
      "epoch": 0.5618056367676295,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1928,
      "step": 2402
    },
    {
      "epoch": 0.5620395275406386,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 2403
    },
    {
      "epoch": 0.5622734183136475,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 2404
    },
    {
      "epoch": 0.5625073090866566,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 2405
    },
    {
      "epoch": 0.5627411998596655,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 2406
    },
    {
      "epoch": 0.5629750906326746,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 2407
    },
    {
      "epoch": 0.5632089814056835,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 2408
    },
    {
      "epoch": 0.5634428721786926,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 2409
    },
    {
      "epoch": 0.5636767629517015,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 2410
    },
    {
      "epoch": 0.5639106537247106,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 2411
    },
    {
      "epoch": 0.5641445444977196,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 2412
    },
    {
      "epoch": 0.5643784352707286,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1506,
      "step": 2413
    },
    {
      "epoch": 0.5646123260437376,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7224,
      "step": 2414
    },
    {
      "epoch": 0.5648462168167466,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 2415
    },
    {
      "epoch": 0.5650801075897556,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 2416
    },
    {
      "epoch": 0.5653139983627646,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.681,
      "step": 2417
    },
    {
      "epoch": 0.5655478891357736,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 2418
    },
    {
      "epoch": 0.5657817799087826,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 2419
    },
    {
      "epoch": 0.5660156706817916,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 2420
    },
    {
      "epoch": 0.5662495614548007,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 2421
    },
    {
      "epoch": 0.5664834522278096,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 2422
    },
    {
      "epoch": 0.5667173430008187,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8824,
      "step": 2423
    },
    {
      "epoch": 0.5669512337738276,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 2424
    },
    {
      "epoch": 0.5671851245468367,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 2425
    },
    {
      "epoch": 0.5674190153198456,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 2426
    },
    {
      "epoch": 0.5676529060928547,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 2427
    },
    {
      "epoch": 0.5678867968658636,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 2428
    },
    {
      "epoch": 0.5681206876388727,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 2429
    },
    {
      "epoch": 0.5683545784118816,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 2430
    },
    {
      "epoch": 0.5685884691848907,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 2431
    },
    {
      "epoch": 0.5688223599578996,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 2432
    },
    {
      "epoch": 0.5690562507309087,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 2433
    },
    {
      "epoch": 0.5692901415039177,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 2434
    },
    {
      "epoch": 0.5695240322769267,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 2435
    },
    {
      "epoch": 0.5697579230499357,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 2436
    },
    {
      "epoch": 0.5699918138229447,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6064,
      "step": 2437
    },
    {
      "epoch": 0.5702257045959537,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 2438
    },
    {
      "epoch": 0.5704595953689627,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 2439
    },
    {
      "epoch": 0.5706934861419717,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 2440
    },
    {
      "epoch": 0.5709273769149807,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 2441
    },
    {
      "epoch": 0.5711612676879897,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 2442
    },
    {
      "epoch": 0.5713951584609988,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1761,
      "step": 2443
    },
    {
      "epoch": 0.5716290492340077,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 2444
    },
    {
      "epoch": 0.5718629400070168,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 2445
    },
    {
      "epoch": 0.5720968307800257,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 2446
    },
    {
      "epoch": 0.5723307215530348,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 2447
    },
    {
      "epoch": 0.5725646123260437,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1956,
      "step": 2448
    },
    {
      "epoch": 0.5727985030990528,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 2449
    },
    {
      "epoch": 0.5730323938720617,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 2450
    },
    {
      "epoch": 0.5732662846450708,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 2451
    },
    {
      "epoch": 0.5735001754180797,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7282,
      "step": 2452
    },
    {
      "epoch": 0.5737340661910888,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 2453
    },
    {
      "epoch": 0.5739679569640977,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 2454
    },
    {
      "epoch": 0.5742018477371068,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 2455
    },
    {
      "epoch": 0.5744357385101158,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 2456
    },
    {
      "epoch": 0.5746696292831248,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 2457
    },
    {
      "epoch": 0.5749035200561338,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 2458
    },
    {
      "epoch": 0.5751374108291428,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7819,
      "step": 2459
    },
    {
      "epoch": 0.5753713016021518,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 2460
    },
    {
      "epoch": 0.5756051923751608,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6976,
      "step": 2461
    },
    {
      "epoch": 0.5758390831481698,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 2462
    },
    {
      "epoch": 0.5760729739211788,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7814,
      "step": 2463
    },
    {
      "epoch": 0.5763068646941878,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6927,
      "step": 2464
    },
    {
      "epoch": 0.5765407554671969,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 2465
    },
    {
      "epoch": 0.5767746462402058,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 2466
    },
    {
      "epoch": 0.5770085370132149,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7175,
      "step": 2467
    },
    {
      "epoch": 0.5772424277862238,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 2468
    },
    {
      "epoch": 0.5774763185592329,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 2469
    },
    {
      "epoch": 0.5777102093322418,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4588,
      "step": 2470
    },
    {
      "epoch": 0.5779441001052509,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 2471
    },
    {
      "epoch": 0.5781779908782598,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 2472
    },
    {
      "epoch": 0.5784118816512689,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 2473
    },
    {
      "epoch": 0.5786457724242778,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 2474
    },
    {
      "epoch": 0.5788796631972869,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 2475
    },
    {
      "epoch": 0.5791135539702958,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 2476
    },
    {
      "epoch": 0.5793474447433049,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7095,
      "step": 2477
    },
    {
      "epoch": 0.5795813355163139,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 2478
    },
    {
      "epoch": 0.5798152262893229,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3059,
      "step": 2479
    },
    {
      "epoch": 0.5800491170623319,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 2480
    },
    {
      "epoch": 0.5802830078353409,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 2481
    },
    {
      "epoch": 0.5805168986083499,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 2482
    },
    {
      "epoch": 0.5807507893813589,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 2483
    },
    {
      "epoch": 0.5809846801543679,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 2484
    },
    {
      "epoch": 0.5812185709273769,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 2485
    },
    {
      "epoch": 0.5814524617003859,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 2486
    },
    {
      "epoch": 0.581686352473395,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 2487
    },
    {
      "epoch": 0.5819202432464039,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 2488
    },
    {
      "epoch": 0.582154134019413,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5209,
      "step": 2489
    },
    {
      "epoch": 0.5823880247924219,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 2490
    },
    {
      "epoch": 0.582621915565431,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 2491
    },
    {
      "epoch": 0.5828558063384399,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 2492
    },
    {
      "epoch": 0.583089697111449,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 2493
    },
    {
      "epoch": 0.5833235878844579,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 2494
    },
    {
      "epoch": 0.583557478657467,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 2495
    },
    {
      "epoch": 0.5837913694304759,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 2496
    },
    {
      "epoch": 0.584025260203485,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 2497
    },
    {
      "epoch": 0.584259150976494,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 2498
    },
    {
      "epoch": 0.584493041749503,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 2499
    },
    {
      "epoch": 0.584726932522512,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 2500
    },
    {
      "epoch": 0.584726932522512,
      "eval_runtime": 4.6061,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 2500
    },
    {
      "epoch": 0.584960823295521,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.331,
      "step": 2501
    },
    {
      "epoch": 0.58519471406853,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 2502
    },
    {
      "epoch": 0.585428604841539,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6212,
      "step": 2503
    },
    {
      "epoch": 0.585662495614548,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 2504
    },
    {
      "epoch": 0.585896386387557,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 2505
    },
    {
      "epoch": 0.586130277160566,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6352,
      "step": 2506
    },
    {
      "epoch": 0.586364167933575,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.655,
      "step": 2507
    },
    {
      "epoch": 0.586598058706584,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 2508
    },
    {
      "epoch": 0.586831949479593,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 2509
    },
    {
      "epoch": 0.587065840252602,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 2510
    },
    {
      "epoch": 0.5872997310256111,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 2511
    },
    {
      "epoch": 0.58753362179862,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8738,
      "step": 2512
    },
    {
      "epoch": 0.5877675125716291,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8385,
      "step": 2513
    },
    {
      "epoch": 0.588001403344638,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 2514
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.708,
      "step": 2515
    },
    {
      "epoch": 0.588469184890656,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 2516
    },
    {
      "epoch": 0.5887030756636651,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 2517
    },
    {
      "epoch": 0.588936966436674,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 2518
    },
    {
      "epoch": 0.5891708572096831,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 2519
    },
    {
      "epoch": 0.589404747982692,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 2520
    },
    {
      "epoch": 0.5896386387557011,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.078,
      "step": 2521
    },
    {
      "epoch": 0.58987252952871,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 2522
    },
    {
      "epoch": 0.5901064203017191,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2029,
      "step": 2523
    },
    {
      "epoch": 0.5903403110747281,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.408,
      "step": 2524
    },
    {
      "epoch": 0.5905742018477371,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 2525
    },
    {
      "epoch": 0.5908080926207461,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 2526
    },
    {
      "epoch": 0.5910419833937551,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 2527
    },
    {
      "epoch": 0.5912758741667641,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 2528
    },
    {
      "epoch": 0.5915097649397731,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 2529
    },
    {
      "epoch": 0.5917436557127821,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 2530
    },
    {
      "epoch": 0.5919775464857912,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4149,
      "step": 2531
    },
    {
      "epoch": 0.5922114372588001,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 2532
    },
    {
      "epoch": 0.5924453280318092,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 2533
    },
    {
      "epoch": 0.5926792188048181,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 2534
    },
    {
      "epoch": 0.5929131095778272,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 2535
    },
    {
      "epoch": 0.5931470003508361,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7202,
      "step": 2536
    },
    {
      "epoch": 0.5933808911238452,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 2537
    },
    {
      "epoch": 0.5936147818968541,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 2538
    },
    {
      "epoch": 0.5938486726698632,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 2539
    },
    {
      "epoch": 0.5940825634428721,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 2540
    },
    {
      "epoch": 0.5943164542158812,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3969,
      "step": 2541
    },
    {
      "epoch": 0.5945503449888901,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3658,
      "step": 2542
    },
    {
      "epoch": 0.5947842357618992,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 2543
    },
    {
      "epoch": 0.5950181265349082,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 2544
    },
    {
      "epoch": 0.5952520173079172,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 2545
    },
    {
      "epoch": 0.5954859080809262,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 2546
    },
    {
      "epoch": 0.5957197988539352,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 2547
    },
    {
      "epoch": 0.5959536896269442,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 2548
    },
    {
      "epoch": 0.5961875803999532,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 2549
    },
    {
      "epoch": 0.5964214711729622,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 2550
    },
    {
      "epoch": 0.5966553619459712,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 2551
    },
    {
      "epoch": 0.5968892527189802,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 2552
    },
    {
      "epoch": 0.5971231434919893,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 2553
    },
    {
      "epoch": 0.5973570342649982,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 2554
    },
    {
      "epoch": 0.5975909250380073,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 2555
    },
    {
      "epoch": 0.5978248158110162,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 2556
    },
    {
      "epoch": 0.5980587065840253,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 2557
    },
    {
      "epoch": 0.5982925973570342,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1956,
      "step": 2558
    },
    {
      "epoch": 0.5985264881300433,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 2559
    },
    {
      "epoch": 0.5987603789030522,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 2560
    },
    {
      "epoch": 0.5989942696760613,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 2561
    },
    {
      "epoch": 0.5992281604490702,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 2562
    },
    {
      "epoch": 0.5994620512220793,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7559,
      "step": 2563
    },
    {
      "epoch": 0.5996959419950882,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 2564
    },
    {
      "epoch": 0.5999298327680973,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 2565
    },
    {
      "epoch": 0.6001637235411063,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 2566
    },
    {
      "epoch": 0.6003976143141153,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 2567
    },
    {
      "epoch": 0.6006315050871243,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 2568
    },
    {
      "epoch": 0.6008653958601333,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 2569
    },
    {
      "epoch": 0.6010992866331423,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.082,
      "step": 2570
    },
    {
      "epoch": 0.6013331774061513,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 2571
    },
    {
      "epoch": 0.6015670681791603,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.229,
      "step": 2572
    },
    {
      "epoch": 0.6018009589521693,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 2573
    },
    {
      "epoch": 0.6020348497251783,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 2574
    },
    {
      "epoch": 0.6022687404981873,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3556,
      "step": 2575
    },
    {
      "epoch": 0.6025026312711963,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1028,
      "step": 2576
    },
    {
      "epoch": 0.6027365220442054,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 2577
    },
    {
      "epoch": 0.6029704128172143,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 2578
    },
    {
      "epoch": 0.6032043035902234,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 2579
    },
    {
      "epoch": 0.6034381943632323,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 2580
    },
    {
      "epoch": 0.6036720851362414,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 2581
    },
    {
      "epoch": 0.6039059759092503,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 2582
    },
    {
      "epoch": 0.6041398666822594,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3779,
      "step": 2583
    },
    {
      "epoch": 0.6043737574552683,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 2584
    },
    {
      "epoch": 0.6046076482282774,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 2585
    },
    {
      "epoch": 0.6048415390012865,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 2586
    },
    {
      "epoch": 0.6050754297742954,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 2587
    },
    {
      "epoch": 0.6053093205473045,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2639,
      "step": 2588
    },
    {
      "epoch": 0.6055432113203134,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2683,
      "step": 2589
    },
    {
      "epoch": 0.6057771020933225,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 2590
    },
    {
      "epoch": 0.6060109928663314,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 2591
    },
    {
      "epoch": 0.6062448836393405,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 2592
    },
    {
      "epoch": 0.6064787744123494,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 2593
    },
    {
      "epoch": 0.6067126651853585,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 2594
    },
    {
      "epoch": 0.6069465559583674,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 2595
    },
    {
      "epoch": 0.6071804467313765,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 2596
    },
    {
      "epoch": 0.6074143375043854,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 2597
    },
    {
      "epoch": 0.6076482282773945,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 2598
    },
    {
      "epoch": 0.6078821190504035,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 2599
    },
    {
      "epoch": 0.6081160098234125,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 2600
    },
    {
      "epoch": 0.6081160098234125,
      "eval_runtime": 4.6018,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 2600
    },
    {
      "epoch": 0.6083499005964215,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 2601
    },
    {
      "epoch": 0.6085837913694305,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8531,
      "step": 2602
    },
    {
      "epoch": 0.6088176821424395,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 2603
    },
    {
      "epoch": 0.6090515729154485,
      "grad_norm": 7.8125,
      "learning_rate": 3e-05,
      "loss": 1.8714,
      "step": 2604
    },
    {
      "epoch": 0.6092854636884575,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 2605
    },
    {
      "epoch": 0.6095193544614665,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 2606
    },
    {
      "epoch": 0.6097532452344755,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 2607
    },
    {
      "epoch": 0.6099871360074846,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 2608
    },
    {
      "epoch": 0.6102210267804935,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 2609
    },
    {
      "epoch": 0.6104549175535026,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0122,
      "step": 2610
    },
    {
      "epoch": 0.6106888083265115,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 2611
    },
    {
      "epoch": 0.6109226990995206,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 2612
    },
    {
      "epoch": 0.6111565898725295,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 2613
    },
    {
      "epoch": 0.6113904806455386,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 2614
    },
    {
      "epoch": 0.6116243714185475,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 2615
    },
    {
      "epoch": 0.6118582621915566,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8508,
      "step": 2616
    },
    {
      "epoch": 0.6120921529645655,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 2617
    },
    {
      "epoch": 0.6123260437375746,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 2618
    },
    {
      "epoch": 0.6125599345105835,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 2619
    },
    {
      "epoch": 0.6127938252835926,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 2620
    },
    {
      "epoch": 0.6130277160566016,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7524,
      "step": 2621
    },
    {
      "epoch": 0.6132616068296106,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 2622
    },
    {
      "epoch": 0.6134954976026196,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8034,
      "step": 2623
    },
    {
      "epoch": 0.6137293883756286,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 2624
    },
    {
      "epoch": 0.6139632791486376,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 2625
    },
    {
      "epoch": 0.6141971699216466,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 2626
    },
    {
      "epoch": 0.6144310606946556,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2233,
      "step": 2627
    },
    {
      "epoch": 0.6146649514676646,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.4947,
      "step": 2628
    },
    {
      "epoch": 0.6148988422406736,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 2629
    },
    {
      "epoch": 0.6151327330136827,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0928,
      "step": 2630
    },
    {
      "epoch": 0.6153666237866916,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3052,
      "step": 2631
    },
    {
      "epoch": 0.6156005145597007,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 2632
    },
    {
      "epoch": 0.6158344053327096,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2336,
      "step": 2633
    },
    {
      "epoch": 0.6160682961057187,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 2634
    },
    {
      "epoch": 0.6163021868787276,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 2635
    },
    {
      "epoch": 0.6165360776517367,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 2636
    },
    {
      "epoch": 0.6167699684247456,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2698,
      "step": 2637
    },
    {
      "epoch": 0.6170038591977547,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 2638
    },
    {
      "epoch": 0.6172377499707636,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 2639
    },
    {
      "epoch": 0.6174716407437727,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 2640
    },
    {
      "epoch": 0.6177055315167816,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8363,
      "step": 2641
    },
    {
      "epoch": 0.6179394222897907,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 2642
    },
    {
      "epoch": 0.6181733130627997,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8417,
      "step": 2643
    },
    {
      "epoch": 0.6184072038358087,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 2644
    },
    {
      "epoch": 0.6186410946088177,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 2645
    },
    {
      "epoch": 0.6188749853818267,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 2646
    },
    {
      "epoch": 0.6191088761548357,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 2647
    },
    {
      "epoch": 0.6193427669278447,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 2648
    },
    {
      "epoch": 0.6195766577008537,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 2649
    },
    {
      "epoch": 0.6198105484738627,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 2650
    },
    {
      "epoch": 0.6200444392468717,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 2651
    },
    {
      "epoch": 0.6202783300198808,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 2652
    },
    {
      "epoch": 0.6205122207928897,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 2653
    },
    {
      "epoch": 0.6207461115658988,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 2654
    },
    {
      "epoch": 0.6209800023389077,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 2655
    },
    {
      "epoch": 0.6212138931119168,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 2656
    },
    {
      "epoch": 0.6214477838849257,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 2657
    },
    {
      "epoch": 0.6216816746579348,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 2658
    },
    {
      "epoch": 0.6219155654309437,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 2659
    },
    {
      "epoch": 0.6221494562039528,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 2660
    },
    {
      "epoch": 0.6223833469769617,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2929,
      "step": 2661
    },
    {
      "epoch": 0.6226172377499708,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4265,
      "step": 2662
    },
    {
      "epoch": 0.6228511285229797,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 2663
    },
    {
      "epoch": 0.6230850192959888,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 2664
    },
    {
      "epoch": 0.6233189100689978,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 2665
    },
    {
      "epoch": 0.6235528008420068,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 2666
    },
    {
      "epoch": 0.6237866916150158,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 2667
    },
    {
      "epoch": 0.6240205823880248,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3178,
      "step": 2668
    },
    {
      "epoch": 0.6242544731610338,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 2669
    },
    {
      "epoch": 0.6244883639340428,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 2670
    },
    {
      "epoch": 0.6247222547070518,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 2671
    },
    {
      "epoch": 0.6249561454800608,
      "grad_norm": 7.65625,
      "learning_rate": 3e-05,
      "loss": 2.3304,
      "step": 2672
    },
    {
      "epoch": 0.6251900362530698,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 2673
    },
    {
      "epoch": 0.6254239270260789,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.417,
      "step": 2674
    },
    {
      "epoch": 0.6256578177990878,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3428,
      "step": 2675
    },
    {
      "epoch": 0.6258917085720969,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 2676
    },
    {
      "epoch": 0.6261255993451058,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 2677
    },
    {
      "epoch": 0.6263594901181149,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 2678
    },
    {
      "epoch": 0.6265933808911238,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 2679
    },
    {
      "epoch": 0.6268272716641329,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 2680
    },
    {
      "epoch": 0.6270611624371418,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 2681
    },
    {
      "epoch": 0.6272950532101509,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 2682
    },
    {
      "epoch": 0.6275289439831598,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 2683
    },
    {
      "epoch": 0.6277628347561689,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 2684
    },
    {
      "epoch": 0.6279967255291778,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 2685
    },
    {
      "epoch": 0.6282306163021869,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 2686
    },
    {
      "epoch": 0.6284645070751959,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 2687
    },
    {
      "epoch": 0.6286983978482049,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 2688
    },
    {
      "epoch": 0.6289322886212139,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 2689
    },
    {
      "epoch": 0.6291661793942229,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 2690
    },
    {
      "epoch": 0.6294000701672319,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 2691
    },
    {
      "epoch": 0.6296339609402409,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 2692
    },
    {
      "epoch": 0.6298678517132499,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 2693
    },
    {
      "epoch": 0.6301017424862589,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 2694
    },
    {
      "epoch": 0.6303356332592679,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 2695
    },
    {
      "epoch": 0.630569524032277,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 2696
    },
    {
      "epoch": 0.6308034148052859,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 2697
    },
    {
      "epoch": 0.631037305578295,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 2698
    },
    {
      "epoch": 0.6312711963513039,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 2699
    },
    {
      "epoch": 0.631505087124313,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 2700
    },
    {
      "epoch": 0.631505087124313,
      "eval_runtime": 4.5714,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 2700
    },
    {
      "epoch": 0.6317389778973219,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 2701
    },
    {
      "epoch": 0.631972868670331,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 2702
    },
    {
      "epoch": 0.6322067594433399,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.249,
      "step": 2703
    },
    {
      "epoch": 0.632440650216349,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 2704
    },
    {
      "epoch": 0.6326745409893579,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 2705
    },
    {
      "epoch": 0.632908431762367,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 2706
    },
    {
      "epoch": 0.633142322535376,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 2707
    },
    {
      "epoch": 0.633376213308385,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 2708
    },
    {
      "epoch": 0.633610104081394,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 2709
    },
    {
      "epoch": 0.633843994854403,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 2710
    },
    {
      "epoch": 0.634077885627412,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 2711
    },
    {
      "epoch": 0.634311776400421,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 2712
    },
    {
      "epoch": 0.63454566717343,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 2713
    },
    {
      "epoch": 0.634779557946439,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 2714
    },
    {
      "epoch": 0.635013448719448,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2465,
      "step": 2715
    },
    {
      "epoch": 0.635247339492457,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 2716
    },
    {
      "epoch": 0.635481230265466,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3717,
      "step": 2717
    },
    {
      "epoch": 0.635715121038475,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 2718
    },
    {
      "epoch": 0.635949011811484,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 2719
    },
    {
      "epoch": 0.6361829025844931,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7013,
      "step": 2720
    },
    {
      "epoch": 0.636416793357502,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 2721
    },
    {
      "epoch": 0.6366506841305111,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 2722
    },
    {
      "epoch": 0.63688457490352,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 2723
    },
    {
      "epoch": 0.6371184656765291,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 2724
    },
    {
      "epoch": 0.637352356449538,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 2725
    },
    {
      "epoch": 0.6375862472225471,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 2726
    },
    {
      "epoch": 0.637820137995556,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 2727
    },
    {
      "epoch": 0.6380540287685651,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 2728
    },
    {
      "epoch": 0.638287919541574,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7239,
      "step": 2729
    },
    {
      "epoch": 0.6385218103145831,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 2730
    },
    {
      "epoch": 0.638755701087592,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7534,
      "step": 2731
    },
    {
      "epoch": 0.6389895918606011,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 2732
    },
    {
      "epoch": 0.6392234826336101,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7597,
      "step": 2733
    },
    {
      "epoch": 0.6394573734066191,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 2734
    },
    {
      "epoch": 0.6396912641796281,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 2735
    },
    {
      "epoch": 0.6399251549526371,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 2736
    },
    {
      "epoch": 0.6401590457256461,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7947,
      "step": 2737
    },
    {
      "epoch": 0.6403929364986551,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 2738
    },
    {
      "epoch": 0.6406268272716641,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 2739
    },
    {
      "epoch": 0.6408607180446731,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 2740
    },
    {
      "epoch": 0.6410946088176821,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8838,
      "step": 2741
    },
    {
      "epoch": 0.6413284995906912,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 2742
    },
    {
      "epoch": 0.6415623903637001,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 2743
    },
    {
      "epoch": 0.6417962811367092,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 2744
    },
    {
      "epoch": 0.6420301719097181,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 2745
    },
    {
      "epoch": 0.6422640626827272,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 2746
    },
    {
      "epoch": 0.6424979534557361,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 2747
    },
    {
      "epoch": 0.6427318442287452,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1003,
      "step": 2748
    },
    {
      "epoch": 0.6429657350017541,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 2749
    },
    {
      "epoch": 0.6431996257747632,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 2750
    },
    {
      "epoch": 0.6434335165477721,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 2751
    },
    {
      "epoch": 0.6436674073207812,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 2752
    },
    {
      "epoch": 0.6439012980937902,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 2753
    },
    {
      "epoch": 0.6441351888667992,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3389,
      "step": 2754
    },
    {
      "epoch": 0.6443690796398082,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 2755
    },
    {
      "epoch": 0.6446029704128172,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 2756
    },
    {
      "epoch": 0.6448368611858262,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3917,
      "step": 2757
    },
    {
      "epoch": 0.6450707519588352,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 2758
    },
    {
      "epoch": 0.6453046427318442,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1773,
      "step": 2759
    },
    {
      "epoch": 0.6455385335048532,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 2760
    },
    {
      "epoch": 0.6457724242778622,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 2761
    },
    {
      "epoch": 0.6460063150508712,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 2762
    },
    {
      "epoch": 0.6462402058238802,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 2763
    },
    {
      "epoch": 0.6464740965968893,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 2764
    },
    {
      "epoch": 0.6467079873698982,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 2765
    },
    {
      "epoch": 0.6469418781429073,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 2766
    },
    {
      "epoch": 0.6471757689159162,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 2767
    },
    {
      "epoch": 0.6474096596889253,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 2768
    },
    {
      "epoch": 0.6476435504619342,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 2769
    },
    {
      "epoch": 0.6478774412349433,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 2770
    },
    {
      "epoch": 0.6481113320079522,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 2771
    },
    {
      "epoch": 0.6483452227809613,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 2772
    },
    {
      "epoch": 0.6485791135539702,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.3288,
      "step": 2773
    },
    {
      "epoch": 0.6488130043269793,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 2774
    },
    {
      "epoch": 0.6490468950999883,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 2775
    },
    {
      "epoch": 0.6492807858729973,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 2776
    },
    {
      "epoch": 0.6495146766460063,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.2067,
      "step": 2777
    },
    {
      "epoch": 0.6497485674190153,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 2778
    },
    {
      "epoch": 0.6499824581920243,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 2779
    },
    {
      "epoch": 0.6502163489650333,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 2780
    },
    {
      "epoch": 0.6504502397380423,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 2781
    },
    {
      "epoch": 0.6506841305110513,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 2782
    },
    {
      "epoch": 0.6509180212840603,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 2783
    },
    {
      "epoch": 0.6511519120570693,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 2784
    },
    {
      "epoch": 0.6513858028300784,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 2785
    },
    {
      "epoch": 0.6516196936030874,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 2786
    },
    {
      "epoch": 0.6518535843760964,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6897,
      "step": 2787
    },
    {
      "epoch": 0.6520874751491054,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 2788
    },
    {
      "epoch": 0.6523213659221144,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 2789
    },
    {
      "epoch": 0.6525552566951234,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 2790
    },
    {
      "epoch": 0.6527891474681324,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.494,
      "step": 2791
    },
    {
      "epoch": 0.6530230382411414,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8205,
      "step": 2792
    },
    {
      "epoch": 0.6532569290141504,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5412,
      "step": 2793
    },
    {
      "epoch": 0.6534908197871594,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 2794
    },
    {
      "epoch": 0.6537247105601685,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 2795
    },
    {
      "epoch": 0.6539586013331774,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 2796
    },
    {
      "epoch": 0.6541924921061865,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 2797
    },
    {
      "epoch": 0.6544263828791954,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6907,
      "step": 2798
    },
    {
      "epoch": 0.6546602736522045,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2654,
      "step": 2799
    },
    {
      "epoch": 0.6548941644252134,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2761,
      "step": 2800
    },
    {
      "epoch": 0.6548941644252134,
      "eval_runtime": 4.5948,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 2800
    },
    {
      "epoch": 0.6551280551982225,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 2801
    },
    {
      "epoch": 0.6553619459712314,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 2802
    },
    {
      "epoch": 0.6555958367442405,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 2803
    },
    {
      "epoch": 0.6558297275172494,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 2804
    },
    {
      "epoch": 0.6560636182902585,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 2805
    },
    {
      "epoch": 0.6562975090632674,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 2806
    },
    {
      "epoch": 0.6565313998362765,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 2807
    },
    {
      "epoch": 0.6567652906092855,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 2808
    },
    {
      "epoch": 0.6569991813822945,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 2809
    },
    {
      "epoch": 0.6572330721553035,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2171,
      "step": 2810
    },
    {
      "epoch": 0.6574669629283125,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 2811
    },
    {
      "epoch": 0.6577008537013215,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7698,
      "step": 2812
    },
    {
      "epoch": 0.6579347444743305,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 2813
    },
    {
      "epoch": 0.6581686352473395,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 2814
    },
    {
      "epoch": 0.6584025260203485,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 2815
    },
    {
      "epoch": 0.6586364167933575,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 2816
    },
    {
      "epoch": 0.6588703075663666,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4642,
      "step": 2817
    },
    {
      "epoch": 0.6591041983393755,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1514,
      "step": 2818
    },
    {
      "epoch": 0.6593380891123846,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 2819
    },
    {
      "epoch": 0.6595719798853935,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8183,
      "step": 2820
    },
    {
      "epoch": 0.6598058706584026,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3475,
      "step": 2821
    },
    {
      "epoch": 0.6600397614314115,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 2822
    },
    {
      "epoch": 0.6602736522044206,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 2823
    },
    {
      "epoch": 0.6605075429774295,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 2824
    },
    {
      "epoch": 0.6607414337504386,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 2825
    },
    {
      "epoch": 0.6609753245234475,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 2826
    },
    {
      "epoch": 0.6612092152964566,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 2827
    },
    {
      "epoch": 0.6614431060694655,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 2828
    },
    {
      "epoch": 0.6616769968424746,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 2829
    },
    {
      "epoch": 0.6619108876154836,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 2830
    },
    {
      "epoch": 0.6621447783884926,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 2831
    },
    {
      "epoch": 0.6623786691615016,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 2832
    },
    {
      "epoch": 0.6626125599345106,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 2833
    },
    {
      "epoch": 0.6628464507075196,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 2834
    },
    {
      "epoch": 0.6630803414805286,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8462,
      "step": 2835
    },
    {
      "epoch": 0.6633142322535376,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 2836
    },
    {
      "epoch": 0.6635481230265466,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 2837
    },
    {
      "epoch": 0.6637820137995556,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 2838
    },
    {
      "epoch": 0.6640159045725647,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 2839
    },
    {
      "epoch": 0.6642497953455736,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 2840
    },
    {
      "epoch": 0.6644836861185827,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 2841
    },
    {
      "epoch": 0.6647175768915916,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 2842
    },
    {
      "epoch": 0.6649514676646007,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 2843
    },
    {
      "epoch": 0.6651853584376096,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 2844
    },
    {
      "epoch": 0.6654192492106187,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 2845
    },
    {
      "epoch": 0.6656531399836276,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 2846
    },
    {
      "epoch": 0.6658870307566367,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 2847
    },
    {
      "epoch": 0.6661209215296456,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4103,
      "step": 2848
    },
    {
      "epoch": 0.6663548123026547,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 2849
    },
    {
      "epoch": 0.6665887030756636,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 2850
    },
    {
      "epoch": 0.6668225938486727,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 2851
    },
    {
      "epoch": 0.6670564846216817,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4823,
      "step": 2852
    },
    {
      "epoch": 0.6672903753946907,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 2853
    },
    {
      "epoch": 0.6675242661676997,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 2854
    },
    {
      "epoch": 0.6677581569407087,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 2855
    },
    {
      "epoch": 0.6679920477137177,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 2856
    },
    {
      "epoch": 0.6682259384867267,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 2857
    },
    {
      "epoch": 0.6684598292597357,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 2858
    },
    {
      "epoch": 0.6686937200327447,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 2859
    },
    {
      "epoch": 0.6689276108057537,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 2860
    },
    {
      "epoch": 0.6691615015787628,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8902,
      "step": 2861
    },
    {
      "epoch": 0.6693953923517717,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 2862
    },
    {
      "epoch": 0.6696292831247808,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 2863
    },
    {
      "epoch": 0.6698631738977897,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 2864
    },
    {
      "epoch": 0.6700970646707988,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 2865
    },
    {
      "epoch": 0.6703309554438077,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 2866
    },
    {
      "epoch": 0.6705648462168168,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 2867
    },
    {
      "epoch": 0.6707987369898257,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 2868
    },
    {
      "epoch": 0.6710326277628348,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 2869
    },
    {
      "epoch": 0.6712665185358437,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 2870
    },
    {
      "epoch": 0.6715004093088528,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 2871
    },
    {
      "epoch": 0.6717343000818617,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 2872
    },
    {
      "epoch": 0.6719681908548708,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 2873
    },
    {
      "epoch": 0.6722020816278798,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 2874
    },
    {
      "epoch": 0.6724359724008888,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7362,
      "step": 2875
    },
    {
      "epoch": 0.6726698631738978,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 2876
    },
    {
      "epoch": 0.6729037539469068,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 2877
    },
    {
      "epoch": 0.6731376447199158,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 2878
    },
    {
      "epoch": 0.6733715354929248,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 2879
    },
    {
      "epoch": 0.6736054262659338,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 2880
    },
    {
      "epoch": 0.6738393170389428,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 2881
    },
    {
      "epoch": 0.6740732078119518,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 2882
    },
    {
      "epoch": 0.6743070985849609,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 2883
    },
    {
      "epoch": 0.6745409893579698,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 2884
    },
    {
      "epoch": 0.6747748801309789,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 2885
    },
    {
      "epoch": 0.6750087709039878,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6997,
      "step": 2886
    },
    {
      "epoch": 0.6752426616769969,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 2887
    },
    {
      "epoch": 0.6754765524500058,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 2888
    },
    {
      "epoch": 0.6757104432230149,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 2889
    },
    {
      "epoch": 0.6759443339960238,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 2890
    },
    {
      "epoch": 0.6761782247690329,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3067,
      "step": 2891
    },
    {
      "epoch": 0.6764121155420418,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0858,
      "step": 2892
    },
    {
      "epoch": 0.6766460063150509,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 2893
    },
    {
      "epoch": 0.6768798970880598,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 2894
    },
    {
      "epoch": 0.6771137878610689,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 2895
    },
    {
      "epoch": 0.6773476786340779,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 2896
    },
    {
      "epoch": 0.6775815694070869,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 2897
    },
    {
      "epoch": 0.6778154601800959,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 2898
    },
    {
      "epoch": 0.6780493509531049,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 2899
    },
    {
      "epoch": 0.6782832417261139,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 2900
    },
    {
      "epoch": 0.6782832417261139,
      "eval_runtime": 4.6303,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 2900
    },
    {
      "epoch": 0.6785171324991229,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 2901
    },
    {
      "epoch": 0.6787510232721319,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 2902
    },
    {
      "epoch": 0.6789849140451409,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7184,
      "step": 2903
    },
    {
      "epoch": 0.6792188048181499,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 2904
    },
    {
      "epoch": 0.679452695591159,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 2905
    },
    {
      "epoch": 0.6796865863641679,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 2906
    },
    {
      "epoch": 0.679920477137177,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 2907
    },
    {
      "epoch": 0.6801543679101859,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 2908
    },
    {
      "epoch": 0.680388258683195,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3333,
      "step": 2909
    },
    {
      "epoch": 0.6806221494562039,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 2910
    },
    {
      "epoch": 0.680856040229213,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 2911
    },
    {
      "epoch": 0.6810899310022219,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 2912
    },
    {
      "epoch": 0.681323821775231,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 2913
    },
    {
      "epoch": 0.6815577125482399,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 2914
    },
    {
      "epoch": 0.681791603321249,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 2915
    },
    {
      "epoch": 0.682025494094258,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 2916
    },
    {
      "epoch": 0.682259384867267,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 2917
    },
    {
      "epoch": 0.682493275640276,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 2918
    },
    {
      "epoch": 0.682727166413285,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 2919
    },
    {
      "epoch": 0.682961057186294,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 2920
    },
    {
      "epoch": 0.683194947959303,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 2921
    },
    {
      "epoch": 0.683428838732312,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 2922
    },
    {
      "epoch": 0.683662729505321,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 2923
    },
    {
      "epoch": 0.68389662027833,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 2924
    },
    {
      "epoch": 0.684130511051339,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 2925
    },
    {
      "epoch": 0.684364401824348,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 2926
    },
    {
      "epoch": 0.684598292597357,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 2927
    },
    {
      "epoch": 0.684832183370366,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2761,
      "step": 2928
    },
    {
      "epoch": 0.6850660741433751,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8776,
      "step": 2929
    },
    {
      "epoch": 0.685299964916384,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 2930
    },
    {
      "epoch": 0.6855338556893931,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 2931
    },
    {
      "epoch": 0.685767746462402,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 2932
    },
    {
      "epoch": 0.6860016372354111,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8153,
      "step": 2933
    },
    {
      "epoch": 0.68623552800842,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 2934
    },
    {
      "epoch": 0.6864694187814291,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7999,
      "step": 2935
    },
    {
      "epoch": 0.686703309554438,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 2936
    },
    {
      "epoch": 0.6869372003274471,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 2937
    },
    {
      "epoch": 0.687171091100456,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 2938
    },
    {
      "epoch": 0.6874049818734651,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 2939
    },
    {
      "epoch": 0.687638872646474,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.4501,
      "step": 2940
    },
    {
      "epoch": 0.6878727634194831,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 2941
    },
    {
      "epoch": 0.6881066541924921,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 2942
    },
    {
      "epoch": 0.6883405449655011,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 2943
    },
    {
      "epoch": 0.6885744357385101,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 2944
    },
    {
      "epoch": 0.6888083265115191,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7261,
      "step": 2945
    },
    {
      "epoch": 0.6890422172845281,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 2946
    },
    {
      "epoch": 0.6892761080575371,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 2947
    },
    {
      "epoch": 0.6895099988305461,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3139,
      "step": 2948
    },
    {
      "epoch": 0.6897438896035551,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 2949
    },
    {
      "epoch": 0.6899777803765641,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 2950
    },
    {
      "epoch": 0.6902116711495732,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 2951
    },
    {
      "epoch": 0.6904455619225821,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 2952
    },
    {
      "epoch": 0.6906794526955912,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 2953
    },
    {
      "epoch": 0.6909133434686001,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 2954
    },
    {
      "epoch": 0.6911472342416092,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 2955
    },
    {
      "epoch": 0.6913811250146181,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 2956
    },
    {
      "epoch": 0.6916150157876272,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 2957
    },
    {
      "epoch": 0.6918489065606361,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 2958
    },
    {
      "epoch": 0.6920827973336452,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 2959
    },
    {
      "epoch": 0.6923166881066541,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 2960
    },
    {
      "epoch": 0.6925505788796632,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 2961
    },
    {
      "epoch": 0.6927844696526722,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 2962
    },
    {
      "epoch": 0.6930183604256812,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 2963
    },
    {
      "epoch": 0.6932522511986902,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 2964
    },
    {
      "epoch": 0.6934861419716992,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 2965
    },
    {
      "epoch": 0.6937200327447082,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 2966
    },
    {
      "epoch": 0.6939539235177172,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 2967
    },
    {
      "epoch": 0.6941878142907262,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 2968
    },
    {
      "epoch": 0.6944217050637352,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 2969
    },
    {
      "epoch": 0.6946555958367442,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7067,
      "step": 2970
    },
    {
      "epoch": 0.6948894866097532,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 2971
    },
    {
      "epoch": 0.6951233773827622,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 2972
    },
    {
      "epoch": 0.6953572681557713,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 2973
    },
    {
      "epoch": 0.6955911589287802,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 2974
    },
    {
      "epoch": 0.6958250497017893,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 2975
    },
    {
      "epoch": 0.6960589404747982,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7234,
      "step": 2976
    },
    {
      "epoch": 0.6962928312478073,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 2977
    },
    {
      "epoch": 0.6965267220208162,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8929,
      "step": 2978
    },
    {
      "epoch": 0.6967606127938253,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 2979
    },
    {
      "epoch": 0.6969945035668342,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 2980
    },
    {
      "epoch": 0.6972283943398433,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 2981
    },
    {
      "epoch": 0.6974622851128522,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 2982
    },
    {
      "epoch": 0.6976961758858613,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6793,
      "step": 2983
    },
    {
      "epoch": 0.6979300666588704,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.337,
      "step": 2984
    },
    {
      "epoch": 0.6981639574318793,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 2985
    },
    {
      "epoch": 0.6983978482048884,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9786,
      "step": 2986
    },
    {
      "epoch": 0.6986317389778973,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 2987
    },
    {
      "epoch": 0.6988656297509064,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7321,
      "step": 2988
    },
    {
      "epoch": 0.6990995205239153,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 2989
    },
    {
      "epoch": 0.6993334112969244,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 2990
    },
    {
      "epoch": 0.6995673020699333,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 2991
    },
    {
      "epoch": 0.6998011928429424,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 2992
    },
    {
      "epoch": 0.7000350836159513,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 2993
    },
    {
      "epoch": 0.7002689743889604,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 2994
    },
    {
      "epoch": 0.7005028651619694,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 2995
    },
    {
      "epoch": 0.7007367559349784,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 2996
    },
    {
      "epoch": 0.7009706467079874,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7789,
      "step": 2997
    },
    {
      "epoch": 0.7012045374809964,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 2998
    },
    {
      "epoch": 0.7014384282540054,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 2999
    },
    {
      "epoch": 0.7016723190270144,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 3000
    },
    {
      "epoch": 0.7016723190270144,
      "eval_runtime": 4.7063,
      "eval_samples_per_second": 0.212,
      "eval_steps_per_second": 0.212,
      "step": 3000
    },
    {
      "epoch": 0.7019062098000234,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0112,
      "step": 3001
    },
    {
      "epoch": 0.7021401005730324,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6394,
      "step": 3002
    },
    {
      "epoch": 0.7023739913460414,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 3003
    },
    {
      "epoch": 0.7026078821190505,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 3004
    },
    {
      "epoch": 0.7028417728920594,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 3005
    },
    {
      "epoch": 0.7030756636650685,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 3006
    },
    {
      "epoch": 0.7033095544380774,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 3007
    },
    {
      "epoch": 0.7035434452110865,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 3008
    },
    {
      "epoch": 0.7037773359840954,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 3009
    },
    {
      "epoch": 0.7040112267571045,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.844,
      "step": 3010
    },
    {
      "epoch": 0.7042451175301134,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 3011
    },
    {
      "epoch": 0.7044790083031225,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 3012
    },
    {
      "epoch": 0.7047128990761314,
      "grad_norm": 13.3125,
      "learning_rate": 3e-05,
      "loss": 2.4246,
      "step": 3013
    },
    {
      "epoch": 0.7049467898491405,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 3014
    },
    {
      "epoch": 0.7051806806221494,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5568,
      "step": 3015
    },
    {
      "epoch": 0.7054145713951585,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 3016
    },
    {
      "epoch": 0.7056484621681675,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 3017
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 3018
    },
    {
      "epoch": 0.7061162437141855,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 3019
    },
    {
      "epoch": 0.7063501344871945,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1779,
      "step": 3020
    },
    {
      "epoch": 0.7065840252602035,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 3021
    },
    {
      "epoch": 0.7068179160332125,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 3022
    },
    {
      "epoch": 0.7070518068062215,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 3023
    },
    {
      "epoch": 0.7072856975792305,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 3024
    },
    {
      "epoch": 0.7075195883522395,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 3025
    },
    {
      "epoch": 0.7077534791252486,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 3026
    },
    {
      "epoch": 0.7079873698982575,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7634,
      "step": 3027
    },
    {
      "epoch": 0.7082212606712666,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 3028
    },
    {
      "epoch": 0.7084551514442755,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 3029
    },
    {
      "epoch": 0.7086890422172846,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 3030
    },
    {
      "epoch": 0.7089229329902935,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 3031
    },
    {
      "epoch": 0.7091568237633026,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 3032
    },
    {
      "epoch": 0.7093907145363115,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 3033
    },
    {
      "epoch": 0.7096246053093206,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 3034
    },
    {
      "epoch": 0.7098584960823295,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 3035
    },
    {
      "epoch": 0.7100923868553386,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 3036
    },
    {
      "epoch": 0.7103262776283475,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 3037
    },
    {
      "epoch": 0.7105601684013566,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 3038
    },
    {
      "epoch": 0.7107940591743656,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 3039
    },
    {
      "epoch": 0.7110279499473746,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 3040
    },
    {
      "epoch": 0.7112618407203836,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4337,
      "step": 3041
    },
    {
      "epoch": 0.7114957314933926,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 3042
    },
    {
      "epoch": 0.7117296222664016,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 3043
    },
    {
      "epoch": 0.7119635130394106,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3236,
      "step": 3044
    },
    {
      "epoch": 0.7121974038124196,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 3045
    },
    {
      "epoch": 0.7124312945854286,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 3046
    },
    {
      "epoch": 0.7126651853584376,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 3047
    },
    {
      "epoch": 0.7128990761314467,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 3048
    },
    {
      "epoch": 0.7131329669044556,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 3049
    },
    {
      "epoch": 0.7133668576774647,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 3050
    },
    {
      "epoch": 0.7136007484504736,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 3051
    },
    {
      "epoch": 0.7138346392234827,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 3052
    },
    {
      "epoch": 0.7140685299964916,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 3053
    },
    {
      "epoch": 0.7143024207695007,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6798,
      "step": 3054
    },
    {
      "epoch": 0.7145363115425096,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7524,
      "step": 3055
    },
    {
      "epoch": 0.7147702023155187,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 3056
    },
    {
      "epoch": 0.7150040930885276,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 3057
    },
    {
      "epoch": 0.7152379838615367,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8503,
      "step": 3058
    },
    {
      "epoch": 0.7154718746345456,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6618,
      "step": 3059
    },
    {
      "epoch": 0.7157057654075547,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4343,
      "step": 3060
    },
    {
      "epoch": 0.7159396561805637,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 3061
    },
    {
      "epoch": 0.7161735469535727,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 3062
    },
    {
      "epoch": 0.7164074377265817,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 3063
    },
    {
      "epoch": 0.7166413284995907,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 3064
    },
    {
      "epoch": 0.7168752192725997,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 3065
    },
    {
      "epoch": 0.7171091100456087,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 3066
    },
    {
      "epoch": 0.7173430008186177,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 3067
    },
    {
      "epoch": 0.7175768915916267,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 3068
    },
    {
      "epoch": 0.7178107823646357,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 3069
    },
    {
      "epoch": 0.7180446731376448,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 3070
    },
    {
      "epoch": 0.7182785639106537,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 3071
    },
    {
      "epoch": 0.7185124546836628,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 3072
    },
    {
      "epoch": 0.7187463454566717,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 3073
    },
    {
      "epoch": 0.7189802362296808,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 3074
    },
    {
      "epoch": 0.7192141270026897,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 3075
    },
    {
      "epoch": 0.7194480177756988,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7718,
      "step": 3076
    },
    {
      "epoch": 0.7196819085487077,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 3077
    },
    {
      "epoch": 0.7199157993217168,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 3078
    },
    {
      "epoch": 0.7201496900947257,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 3079
    },
    {
      "epoch": 0.7203835808677348,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 3080
    },
    {
      "epoch": 0.7206174716407437,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6125,
      "step": 3081
    },
    {
      "epoch": 0.7208513624137528,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 3082
    },
    {
      "epoch": 0.7210852531867618,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 3083
    },
    {
      "epoch": 0.7213191439597708,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 3084
    },
    {
      "epoch": 0.7215530347327798,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 3085
    },
    {
      "epoch": 0.7217869255057888,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 3086
    },
    {
      "epoch": 0.7220208162787978,
      "grad_norm": 14.1875,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 3087
    },
    {
      "epoch": 0.7222547070518068,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 3088
    },
    {
      "epoch": 0.7224885978248158,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 3089
    },
    {
      "epoch": 0.7227224885978248,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 3090
    },
    {
      "epoch": 0.7229563793708338,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 3091
    },
    {
      "epoch": 0.7231902701438429,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 3092
    },
    {
      "epoch": 0.7234241609168518,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 3093
    },
    {
      "epoch": 0.7236580516898609,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 3094
    },
    {
      "epoch": 0.7238919424628698,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 3095
    },
    {
      "epoch": 0.7241258332358789,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.6622,
      "step": 3096
    },
    {
      "epoch": 0.7243597240088878,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 3097
    },
    {
      "epoch": 0.7245936147818969,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 3098
    },
    {
      "epoch": 0.7248275055549058,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 3099
    },
    {
      "epoch": 0.7250613963279149,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 3100
    },
    {
      "epoch": 0.7250613963279149,
      "eval_runtime": 4.584,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 3100
    },
    {
      "epoch": 0.7252952871009238,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 3101
    },
    {
      "epoch": 0.7255291778739329,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 3102
    },
    {
      "epoch": 0.7257630686469418,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 3103
    },
    {
      "epoch": 0.7259969594199509,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3626,
      "step": 3104
    },
    {
      "epoch": 0.7262308501929599,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 3105
    },
    {
      "epoch": 0.7264647409659689,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 3106
    },
    {
      "epoch": 0.7266986317389779,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 3107
    },
    {
      "epoch": 0.7269325225119869,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 3108
    },
    {
      "epoch": 0.7271664132849959,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 3109
    },
    {
      "epoch": 0.7274003040580049,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 3110
    },
    {
      "epoch": 0.7276341948310139,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 3111
    },
    {
      "epoch": 0.7278680856040229,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7108,
      "step": 3112
    },
    {
      "epoch": 0.7281019763770319,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 3113
    },
    {
      "epoch": 0.728335867150041,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 3114
    },
    {
      "epoch": 0.7285697579230499,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 3115
    },
    {
      "epoch": 0.728803648696059,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7052,
      "step": 3116
    },
    {
      "epoch": 0.7290375394690679,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2184,
      "step": 3117
    },
    {
      "epoch": 0.729271430242077,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8043,
      "step": 3118
    },
    {
      "epoch": 0.7295053210150859,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6847,
      "step": 3119
    },
    {
      "epoch": 0.729739211788095,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 3120
    },
    {
      "epoch": 0.7299731025611039,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 3121
    },
    {
      "epoch": 0.730206993334113,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 3122
    },
    {
      "epoch": 0.7304408841071219,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 3123
    },
    {
      "epoch": 0.730674774880131,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 3124
    },
    {
      "epoch": 0.7309086656531399,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 3125
    },
    {
      "epoch": 0.731142556426149,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 3126
    },
    {
      "epoch": 0.731376447199158,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 3127
    },
    {
      "epoch": 0.731610337972167,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 3128
    },
    {
      "epoch": 0.731844228745176,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 3129
    },
    {
      "epoch": 0.732078119518185,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 3130
    },
    {
      "epoch": 0.732312010291194,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 3131
    },
    {
      "epoch": 0.732545901064203,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 3132
    },
    {
      "epoch": 0.732779791837212,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 3133
    },
    {
      "epoch": 0.733013682610221,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 3134
    },
    {
      "epoch": 0.73324757338323,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 3135
    },
    {
      "epoch": 0.733481464156239,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 3136
    },
    {
      "epoch": 0.733715354929248,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 3137
    },
    {
      "epoch": 0.7339492457022571,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 3138
    },
    {
      "epoch": 0.734183136475266,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 3139
    },
    {
      "epoch": 0.7344170272482751,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 3140
    },
    {
      "epoch": 0.734650918021284,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 3141
    },
    {
      "epoch": 0.7348848087942931,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 3142
    },
    {
      "epoch": 0.735118699567302,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 3143
    },
    {
      "epoch": 0.7353525903403111,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 3144
    },
    {
      "epoch": 0.73558648111332,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 3145
    },
    {
      "epoch": 0.7358203718863291,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 3146
    },
    {
      "epoch": 0.736054262659338,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 3147
    },
    {
      "epoch": 0.7362881534323471,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0528,
      "step": 3148
    },
    {
      "epoch": 0.736522044205356,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 3149
    },
    {
      "epoch": 0.7367559349783651,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 3150
    },
    {
      "epoch": 0.7369898257513741,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 3151
    },
    {
      "epoch": 0.7372237165243831,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 3152
    },
    {
      "epoch": 0.7374576072973921,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 3153
    },
    {
      "epoch": 0.7376914980704011,
      "grad_norm": 8.375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 3154
    },
    {
      "epoch": 0.7379253888434101,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 3155
    },
    {
      "epoch": 0.7381592796164191,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 3156
    },
    {
      "epoch": 0.7383931703894281,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 3157
    },
    {
      "epoch": 0.7386270611624371,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7033,
      "step": 3158
    },
    {
      "epoch": 0.7388609519354461,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.9184,
      "step": 3159
    },
    {
      "epoch": 0.7390948427084552,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 3160
    },
    {
      "epoch": 0.7393287334814641,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 3161
    },
    {
      "epoch": 0.7395626242544732,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 3162
    },
    {
      "epoch": 0.7397965150274821,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 3163
    },
    {
      "epoch": 0.7400304058004912,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 3164
    },
    {
      "epoch": 0.7402642965735001,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 3165
    },
    {
      "epoch": 0.7404981873465092,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 3166
    },
    {
      "epoch": 0.7407320781195181,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 3167
    },
    {
      "epoch": 0.7409659688925272,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 3168
    },
    {
      "epoch": 0.7411998596655361,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 3169
    },
    {
      "epoch": 0.7414337504385452,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 3170
    },
    {
      "epoch": 0.7416676412115542,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 3171
    },
    {
      "epoch": 0.7419015319845632,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 3172
    },
    {
      "epoch": 0.7421354227575722,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 3173
    },
    {
      "epoch": 0.7423693135305812,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 3174
    },
    {
      "epoch": 0.7426032043035902,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2691,
      "step": 3175
    },
    {
      "epoch": 0.7428370950765992,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 3176
    },
    {
      "epoch": 0.7430709858496082,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 3177
    },
    {
      "epoch": 0.7433048766226172,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 3178
    },
    {
      "epoch": 0.7435387673956262,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 3179
    },
    {
      "epoch": 0.7437726581686352,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 3180
    },
    {
      "epoch": 0.7440065489416442,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 3181
    },
    {
      "epoch": 0.7442404397146533,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 3182
    },
    {
      "epoch": 0.7444743304876623,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 3183
    },
    {
      "epoch": 0.7447082212606713,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4107,
      "step": 3184
    },
    {
      "epoch": 0.7449421120336803,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 3185
    },
    {
      "epoch": 0.7451760028066893,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 3186
    },
    {
      "epoch": 0.7454098935796983,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 3187
    },
    {
      "epoch": 0.7456437843527073,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 3188
    },
    {
      "epoch": 0.7458776751257163,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 3189
    },
    {
      "epoch": 0.7461115658987253,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 3190
    },
    {
      "epoch": 0.7463454566717344,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 3191
    },
    {
      "epoch": 0.7465793474447433,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 3192
    },
    {
      "epoch": 0.7468132382177524,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 3193
    },
    {
      "epoch": 0.7470471289907613,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 3194
    },
    {
      "epoch": 0.7472810197637704,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4126,
      "step": 3195
    },
    {
      "epoch": 0.7475149105367793,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 3196
    },
    {
      "epoch": 0.7477488013097884,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 3197
    },
    {
      "epoch": 0.7479826920827973,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 3198
    },
    {
      "epoch": 0.7482165828558064,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 3199
    },
    {
      "epoch": 0.7484504736288153,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 3200
    },
    {
      "epoch": 0.7484504736288153,
      "eval_runtime": 4.6043,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 3200
    },
    {
      "epoch": 0.7486843644018244,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9976,
      "step": 3201
    },
    {
      "epoch": 0.7489182551748333,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2849,
      "step": 3202
    },
    {
      "epoch": 0.7491521459478424,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 3203
    },
    {
      "epoch": 0.7493860367208514,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 3204
    },
    {
      "epoch": 0.7496199274938604,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 3205
    },
    {
      "epoch": 0.7498538182668694,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 3206
    },
    {
      "epoch": 0.7500877090398784,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 3207
    },
    {
      "epoch": 0.7503215998128874,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 3208
    },
    {
      "epoch": 0.7505554905858964,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3697,
      "step": 3209
    },
    {
      "epoch": 0.7507893813589054,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 3210
    },
    {
      "epoch": 0.7510232721319144,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 3211
    },
    {
      "epoch": 0.7512571629049234,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 3212
    },
    {
      "epoch": 0.7514910536779325,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 3213
    },
    {
      "epoch": 0.7517249444509414,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 3214
    },
    {
      "epoch": 0.7519588352239505,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 3215
    },
    {
      "epoch": 0.7521927259969594,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 3216
    },
    {
      "epoch": 0.7524266167699685,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 3217
    },
    {
      "epoch": 0.7526605075429774,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 3218
    },
    {
      "epoch": 0.7528943983159865,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 3219
    },
    {
      "epoch": 0.7531282890889954,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 3220
    },
    {
      "epoch": 0.7533621798620045,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 3221
    },
    {
      "epoch": 0.7535960706350134,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.255,
      "step": 3222
    },
    {
      "epoch": 0.7538299614080225,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 3223
    },
    {
      "epoch": 0.7540638521810314,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 3224
    },
    {
      "epoch": 0.7542977429540405,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 3225
    },
    {
      "epoch": 0.7545316337270495,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 3226
    },
    {
      "epoch": 0.7547655245000585,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 3227
    },
    {
      "epoch": 0.7549994152730675,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 3228
    },
    {
      "epoch": 0.7552333060460765,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 3229
    },
    {
      "epoch": 0.7554671968190855,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8536,
      "step": 3230
    },
    {
      "epoch": 0.7557010875920945,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 3231
    },
    {
      "epoch": 0.7559349783651035,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 3232
    },
    {
      "epoch": 0.7561688691381125,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 3233
    },
    {
      "epoch": 0.7564027599111215,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2746,
      "step": 3234
    },
    {
      "epoch": 0.7566366506841306,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 3235
    },
    {
      "epoch": 0.7568705414571395,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 3236
    },
    {
      "epoch": 0.7571044322301486,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 3237
    },
    {
      "epoch": 0.7573383230031575,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 3238
    },
    {
      "epoch": 0.7575722137761666,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 3239
    },
    {
      "epoch": 0.7578061045491755,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 3240
    },
    {
      "epoch": 0.7580399953221846,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 3241
    },
    {
      "epoch": 0.7582738860951935,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 3242
    },
    {
      "epoch": 0.7585077768682026,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 3243
    },
    {
      "epoch": 0.7587416676412115,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6436,
      "step": 3244
    },
    {
      "epoch": 0.7589755584142206,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 3245
    },
    {
      "epoch": 0.7592094491872295,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 3246
    },
    {
      "epoch": 0.7594433399602386,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 3247
    },
    {
      "epoch": 0.7596772307332476,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 3248
    },
    {
      "epoch": 0.7599111215062566,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 3249
    },
    {
      "epoch": 0.7601450122792656,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 3250
    },
    {
      "epoch": 0.7603789030522746,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 3251
    },
    {
      "epoch": 0.7606127938252836,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 3252
    },
    {
      "epoch": 0.7608466845982926,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 3253
    },
    {
      "epoch": 0.7610805753713016,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5447,
      "step": 3254
    },
    {
      "epoch": 0.7613144661443106,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 3255
    },
    {
      "epoch": 0.7615483569173196,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 3256
    },
    {
      "epoch": 0.7617822476903287,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 3257
    },
    {
      "epoch": 0.7620161384633376,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 3258
    },
    {
      "epoch": 0.7622500292363467,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 3259
    },
    {
      "epoch": 0.7624839200093556,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 3260
    },
    {
      "epoch": 0.7627178107823647,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 3261
    },
    {
      "epoch": 0.7629517015553736,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 3262
    },
    {
      "epoch": 0.7631855923283827,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 3263
    },
    {
      "epoch": 0.7634194831013916,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 3264
    },
    {
      "epoch": 0.7636533738744007,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3053,
      "step": 3265
    },
    {
      "epoch": 0.7638872646474096,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 3266
    },
    {
      "epoch": 0.7641211554204187,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 3267
    },
    {
      "epoch": 0.7643550461934276,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 3268
    },
    {
      "epoch": 0.7645889369664367,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 3269
    },
    {
      "epoch": 0.7648228277394457,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 3270
    },
    {
      "epoch": 0.7650567185124547,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 3271
    },
    {
      "epoch": 0.7652906092854637,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 3272
    },
    {
      "epoch": 0.7655245000584727,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 3273
    },
    {
      "epoch": 0.7657583908314817,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 3274
    },
    {
      "epoch": 0.7659922816044907,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 3275
    },
    {
      "epoch": 0.7662261723774997,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 3276
    },
    {
      "epoch": 0.7664600631505087,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 3277
    },
    {
      "epoch": 0.7666939539235177,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 3278
    },
    {
      "epoch": 0.7669278446965268,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 3279
    },
    {
      "epoch": 0.7671617354695357,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 3280
    },
    {
      "epoch": 0.7673956262425448,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 3281
    },
    {
      "epoch": 0.7676295170155537,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.462,
      "step": 3282
    },
    {
      "epoch": 0.7678634077885628,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6998,
      "step": 3283
    },
    {
      "epoch": 0.7680972985615717,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3147,
      "step": 3284
    },
    {
      "epoch": 0.7683311893345808,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2596,
      "step": 3285
    },
    {
      "epoch": 0.7685650801075897,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7449,
      "step": 3286
    },
    {
      "epoch": 0.7687989708805988,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 3287
    },
    {
      "epoch": 0.7690328616536077,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 3288
    },
    {
      "epoch": 0.7692667524266168,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2848,
      "step": 3289
    },
    {
      "epoch": 0.7695006431996257,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 3290
    },
    {
      "epoch": 0.7697345339726348,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 3291
    },
    {
      "epoch": 0.7699684247456438,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 3292
    },
    {
      "epoch": 0.7702023155186528,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 3293
    },
    {
      "epoch": 0.7704362062916618,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 3294
    },
    {
      "epoch": 0.7706700970646708,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 3295
    },
    {
      "epoch": 0.7709039878376798,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 3296
    },
    {
      "epoch": 0.7711378786106888,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 3297
    },
    {
      "epoch": 0.7713717693836978,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 3298
    },
    {
      "epoch": 0.7716056601567068,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3489,
      "step": 3299
    },
    {
      "epoch": 0.7718395509297158,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 3300
    },
    {
      "epoch": 0.7718395509297158,
      "eval_runtime": 4.6656,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 3300
    },
    {
      "epoch": 0.7720734417027248,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 3301
    },
    {
      "epoch": 0.7723073324757338,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 3302
    },
    {
      "epoch": 0.7725412232487429,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3054,
      "step": 3303
    },
    {
      "epoch": 0.7727751140217518,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7296,
      "step": 3304
    },
    {
      "epoch": 0.7730090047947609,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 3305
    },
    {
      "epoch": 0.7732428955677698,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 3306
    },
    {
      "epoch": 0.7734767863407789,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 3307
    },
    {
      "epoch": 0.7737106771137878,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 3308
    },
    {
      "epoch": 0.7739445678867969,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 3309
    },
    {
      "epoch": 0.7741784586598058,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 3310
    },
    {
      "epoch": 0.7744123494328149,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 3311
    },
    {
      "epoch": 0.7746462402058238,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.4009,
      "step": 3312
    },
    {
      "epoch": 0.7748801309788329,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 3313
    },
    {
      "epoch": 0.7751140217518419,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 3314
    },
    {
      "epoch": 0.7753479125248509,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2334,
      "step": 3315
    },
    {
      "epoch": 0.7755818032978599,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 3316
    },
    {
      "epoch": 0.7758156940708689,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 3317
    },
    {
      "epoch": 0.7760495848438779,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 3318
    },
    {
      "epoch": 0.7762834756168869,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6331,
      "step": 3319
    },
    {
      "epoch": 0.7765173663898959,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 3320
    },
    {
      "epoch": 0.7767512571629049,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1715,
      "step": 3321
    },
    {
      "epoch": 0.7769851479359139,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 3322
    },
    {
      "epoch": 0.777219038708923,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 3323
    },
    {
      "epoch": 0.7774529294819319,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 3324
    },
    {
      "epoch": 0.777686820254941,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 3325
    },
    {
      "epoch": 0.7779207110279499,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7266,
      "step": 3326
    },
    {
      "epoch": 0.778154601800959,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 3327
    },
    {
      "epoch": 0.7783884925739679,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 3328
    },
    {
      "epoch": 0.778622383346977,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 3329
    },
    {
      "epoch": 0.7788562741199859,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 3330
    },
    {
      "epoch": 0.779090164892995,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 3331
    },
    {
      "epoch": 0.7793240556660039,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2761,
      "step": 3332
    },
    {
      "epoch": 0.779557946439013,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 3333
    },
    {
      "epoch": 0.7797918372120219,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0791,
      "step": 3334
    },
    {
      "epoch": 0.780025727985031,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 3335
    },
    {
      "epoch": 0.78025961875804,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 3336
    },
    {
      "epoch": 0.780493509531049,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5154,
      "step": 3337
    },
    {
      "epoch": 0.780727400304058,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6593,
      "step": 3338
    },
    {
      "epoch": 0.780961291077067,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 3339
    },
    {
      "epoch": 0.781195181850076,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 3340
    },
    {
      "epoch": 0.781429072623085,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 3341
    },
    {
      "epoch": 0.781662963396094,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 3342
    },
    {
      "epoch": 0.781896854169103,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.4278,
      "step": 3343
    },
    {
      "epoch": 0.782130744942112,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 3344
    },
    {
      "epoch": 0.782364635715121,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 3345
    },
    {
      "epoch": 0.78259852648813,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 3346
    },
    {
      "epoch": 0.7828324172611391,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 3347
    },
    {
      "epoch": 0.783066308034148,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 3348
    },
    {
      "epoch": 0.7833001988071571,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 3349
    },
    {
      "epoch": 0.783534089580166,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 3350
    },
    {
      "epoch": 0.7837679803531751,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.495,
      "step": 3351
    },
    {
      "epoch": 0.784001871126184,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 3352
    },
    {
      "epoch": 0.7842357618991931,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 3353
    },
    {
      "epoch": 0.784469652672202,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 3354
    },
    {
      "epoch": 0.7847035434452111,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8318,
      "step": 3355
    },
    {
      "epoch": 0.78493743421822,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 3356
    },
    {
      "epoch": 0.7851713249912291,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 3357
    },
    {
      "epoch": 0.785405215764238,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 3358
    },
    {
      "epoch": 0.7856391065372471,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6514,
      "step": 3359
    },
    {
      "epoch": 0.7858729973102561,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 3360
    },
    {
      "epoch": 0.7861068880832651,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 3361
    },
    {
      "epoch": 0.7863407788562741,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 3362
    },
    {
      "epoch": 0.7865746696292831,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 3363
    },
    {
      "epoch": 0.7868085604022921,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 3364
    },
    {
      "epoch": 0.7870424511753011,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3457,
      "step": 3365
    },
    {
      "epoch": 0.7872763419483101,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 3366
    },
    {
      "epoch": 0.7875102327213191,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2143,
      "step": 3367
    },
    {
      "epoch": 0.7877441234943281,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7587,
      "step": 3368
    },
    {
      "epoch": 0.7879780142673372,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 3369
    },
    {
      "epoch": 0.7882119050403461,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.204,
      "step": 3370
    },
    {
      "epoch": 0.7884457958133552,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5542,
      "step": 3371
    },
    {
      "epoch": 0.7886796865863641,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 3372
    },
    {
      "epoch": 0.7889135773593732,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 3373
    },
    {
      "epoch": 0.7891474681323821,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 3374
    },
    {
      "epoch": 0.7893813589053912,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 3375
    },
    {
      "epoch": 0.7896152496784001,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 3376
    },
    {
      "epoch": 0.7898491404514092,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 3377
    },
    {
      "epoch": 0.7900830312244181,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 3378
    },
    {
      "epoch": 0.7903169219974272,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 3379
    },
    {
      "epoch": 0.7905508127704362,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 3380
    },
    {
      "epoch": 0.7907847035434452,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 3381
    },
    {
      "epoch": 0.7910185943164543,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 3382
    },
    {
      "epoch": 0.7912524850894632,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2344,
      "step": 3383
    },
    {
      "epoch": 0.7914863758624723,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 3384
    },
    {
      "epoch": 0.7917202666354812,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 3385
    },
    {
      "epoch": 0.7919541574084903,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 3386
    },
    {
      "epoch": 0.7921880481814992,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 3387
    },
    {
      "epoch": 0.7924219389545083,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0544,
      "step": 3388
    },
    {
      "epoch": 0.7926558297275172,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 3389
    },
    {
      "epoch": 0.7928897205005263,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 3390
    },
    {
      "epoch": 0.7931236112735353,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2305,
      "step": 3391
    },
    {
      "epoch": 0.7933575020465443,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 3392
    },
    {
      "epoch": 0.7935913928195533,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 3393
    },
    {
      "epoch": 0.7938252835925623,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4346,
      "step": 3394
    },
    {
      "epoch": 0.7940591743655713,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 3395
    },
    {
      "epoch": 0.7942930651385803,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 3396
    },
    {
      "epoch": 0.7945269559115893,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 3397
    },
    {
      "epoch": 0.7947608466845983,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 3398
    },
    {
      "epoch": 0.7949947374576073,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 3399
    },
    {
      "epoch": 0.7952286282306164,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 3400
    },
    {
      "epoch": 0.7952286282306164,
      "eval_runtime": 4.5985,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 3400
    },
    {
      "epoch": 0.7954625190036253,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 3401
    },
    {
      "epoch": 0.7956964097766344,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 3402
    },
    {
      "epoch": 0.7959303005496433,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 3403
    },
    {
      "epoch": 0.7961641913226524,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 3404
    },
    {
      "epoch": 0.7963980820956613,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 3405
    },
    {
      "epoch": 0.7966319728686704,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6108,
      "step": 3406
    },
    {
      "epoch": 0.7968658636416793,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 3407
    },
    {
      "epoch": 0.7970997544146884,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 3408
    },
    {
      "epoch": 0.7973336451876973,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 3409
    },
    {
      "epoch": 0.7975675359607064,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 3410
    },
    {
      "epoch": 0.7978014267337153,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 3411
    },
    {
      "epoch": 0.7980353175067244,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 3412
    },
    {
      "epoch": 0.7982692082797334,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5784,
      "step": 3413
    },
    {
      "epoch": 0.7985030990527424,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 3414
    },
    {
      "epoch": 0.7987369898257514,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 3415
    },
    {
      "epoch": 0.7989708805987604,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 3416
    },
    {
      "epoch": 0.7992047713717694,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 3417
    },
    {
      "epoch": 0.7994386621447784,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 3418
    },
    {
      "epoch": 0.7996725529177874,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 3419
    },
    {
      "epoch": 0.7999064436907964,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 3420
    },
    {
      "epoch": 0.8001403344638054,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 3421
    },
    {
      "epoch": 0.8003742252368145,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 3422
    },
    {
      "epoch": 0.8006081160098234,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 3423
    },
    {
      "epoch": 0.8008420067828325,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 3424
    },
    {
      "epoch": 0.8010758975558414,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2237,
      "step": 3425
    },
    {
      "epoch": 0.8013097883288505,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 3426
    },
    {
      "epoch": 0.8015436791018594,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 3427
    },
    {
      "epoch": 0.8017775698748685,
      "grad_norm": 7.28125,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 3428
    },
    {
      "epoch": 0.8020114606478774,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7845,
      "step": 3429
    },
    {
      "epoch": 0.8022453514208865,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6175,
      "step": 3430
    },
    {
      "epoch": 0.8024792421938954,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 3431
    },
    {
      "epoch": 0.8027131329669045,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 3432
    },
    {
      "epoch": 0.8029470237399134,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 3433
    },
    {
      "epoch": 0.8031809145129225,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 3434
    },
    {
      "epoch": 0.8034148052859315,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 3435
    },
    {
      "epoch": 0.8036486960589405,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 3436
    },
    {
      "epoch": 0.8038825868319495,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 3437
    },
    {
      "epoch": 0.8041164776049585,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 3438
    },
    {
      "epoch": 0.8043503683779675,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 3439
    },
    {
      "epoch": 0.8045842591509765,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 3440
    },
    {
      "epoch": 0.8048181499239855,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 3441
    },
    {
      "epoch": 0.8050520406969945,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 3442
    },
    {
      "epoch": 0.8052859314700035,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 3443
    },
    {
      "epoch": 0.8055198222430126,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8183,
      "step": 3444
    },
    {
      "epoch": 0.8057537130160215,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 3445
    },
    {
      "epoch": 0.8059876037890306,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 3446
    },
    {
      "epoch": 0.8062214945620395,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 3447
    },
    {
      "epoch": 0.8064553853350486,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 3448
    },
    {
      "epoch": 0.8066892761080575,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 3449
    },
    {
      "epoch": 0.8069231668810666,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 3450
    },
    {
      "epoch": 0.8071570576540755,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 3451
    },
    {
      "epoch": 0.8073909484270846,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1584,
      "step": 3452
    },
    {
      "epoch": 0.8076248392000935,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 3453
    },
    {
      "epoch": 0.8078587299731026,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5033,
      "step": 3454
    },
    {
      "epoch": 0.8080926207461115,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 3455
    },
    {
      "epoch": 0.8083265115191206,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 3456
    },
    {
      "epoch": 0.8085604022921296,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5657,
      "step": 3457
    },
    {
      "epoch": 0.8087942930651386,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 3458
    },
    {
      "epoch": 0.8090281838381476,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 3459
    },
    {
      "epoch": 0.8092620746111566,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 3460
    },
    {
      "epoch": 0.8094959653841656,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 3461
    },
    {
      "epoch": 0.8097298561571746,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 3462
    },
    {
      "epoch": 0.8099637469301836,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3174,
      "step": 3463
    },
    {
      "epoch": 0.8101976377031926,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 3464
    },
    {
      "epoch": 0.8104315284762016,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 3465
    },
    {
      "epoch": 0.8106654192492107,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 3466
    },
    {
      "epoch": 0.8108993100222196,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 3467
    },
    {
      "epoch": 0.8111332007952287,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 3468
    },
    {
      "epoch": 0.8113670915682376,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 3469
    },
    {
      "epoch": 0.8116009823412467,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 3470
    },
    {
      "epoch": 0.8118348731142556,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 3471
    },
    {
      "epoch": 0.8120687638872647,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 3472
    },
    {
      "epoch": 0.8123026546602736,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 3473
    },
    {
      "epoch": 0.8125365454332827,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 3474
    },
    {
      "epoch": 0.8127704362062916,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 3475
    },
    {
      "epoch": 0.8130043269793007,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 3476
    },
    {
      "epoch": 0.8132382177523096,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 3477
    },
    {
      "epoch": 0.8134721085253187,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6323,
      "step": 3478
    },
    {
      "epoch": 0.8137059992983277,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 3479
    },
    {
      "epoch": 0.8139398900713367,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 3480
    },
    {
      "epoch": 0.8141737808443457,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 3481
    },
    {
      "epoch": 0.8144076716173547,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 3482
    },
    {
      "epoch": 0.8146415623903637,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 3483
    },
    {
      "epoch": 0.8148754531633727,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6718,
      "step": 3484
    },
    {
      "epoch": 0.8151093439363817,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 3485
    },
    {
      "epoch": 0.8153432347093907,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 3486
    },
    {
      "epoch": 0.8155771254823997,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 3487
    },
    {
      "epoch": 0.8158110162554087,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 3488
    },
    {
      "epoch": 0.8160449070284177,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 3489
    },
    {
      "epoch": 0.8162787978014268,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 3490
    },
    {
      "epoch": 0.8165126885744357,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 3491
    },
    {
      "epoch": 0.8167465793474448,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 3492
    },
    {
      "epoch": 0.8169804701204537,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 3493
    },
    {
      "epoch": 0.8172143608934628,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 3494
    },
    {
      "epoch": 0.8174482516664717,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 3495
    },
    {
      "epoch": 0.8176821424394808,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 3496
    },
    {
      "epoch": 0.8179160332124897,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 3497
    },
    {
      "epoch": 0.8181499239854988,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 3498
    },
    {
      "epoch": 0.8183838147585077,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 3499
    },
    {
      "epoch": 0.8186177055315168,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 3500
    },
    {
      "epoch": 0.8186177055315168,
      "eval_runtime": 4.6927,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 3500
    },
    {
      "epoch": 0.8188515963045258,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2105,
      "step": 3501
    },
    {
      "epoch": 0.8190854870775348,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 3502
    },
    {
      "epoch": 0.8193193778505438,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 3503
    },
    {
      "epoch": 0.8195532686235528,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 3504
    },
    {
      "epoch": 0.8197871593965618,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 3505
    },
    {
      "epoch": 0.8200210501695708,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 3506
    },
    {
      "epoch": 0.8202549409425798,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 3507
    },
    {
      "epoch": 0.8204888317155888,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 3508
    },
    {
      "epoch": 0.8207227224885978,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 3509
    },
    {
      "epoch": 0.8209566132616068,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 3510
    },
    {
      "epoch": 0.8211905040346158,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.3951,
      "step": 3511
    },
    {
      "epoch": 0.8214243948076249,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6697,
      "step": 3512
    },
    {
      "epoch": 0.8216582855806338,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 3513
    },
    {
      "epoch": 0.8218921763536429,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 3514
    },
    {
      "epoch": 0.8221260671266518,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 3515
    },
    {
      "epoch": 0.8223599578996609,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 3516
    },
    {
      "epoch": 0.8225938486726698,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9667,
      "step": 3517
    },
    {
      "epoch": 0.8228277394456789,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 3518
    },
    {
      "epoch": 0.8230616302186878,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 3519
    },
    {
      "epoch": 0.8232955209916969,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 3520
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 3521
    },
    {
      "epoch": 0.8237633025377149,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7201,
      "step": 3522
    },
    {
      "epoch": 0.8239971933107239,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9301,
      "step": 3523
    },
    {
      "epoch": 0.8242310840837329,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 3524
    },
    {
      "epoch": 0.8244649748567419,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 3525
    },
    {
      "epoch": 0.8246988656297509,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3994,
      "step": 3526
    },
    {
      "epoch": 0.8249327564027599,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0348,
      "step": 3527
    },
    {
      "epoch": 0.8251666471757689,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 3528
    },
    {
      "epoch": 0.8254005379487779,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 3529
    },
    {
      "epoch": 0.8256344287217869,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 3530
    },
    {
      "epoch": 0.8258683194947959,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 3531
    },
    {
      "epoch": 0.826102210267805,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 3532
    },
    {
      "epoch": 0.8263361010408139,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 3533
    },
    {
      "epoch": 0.826569991813823,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7178,
      "step": 3534
    },
    {
      "epoch": 0.8268038825868319,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 3535
    },
    {
      "epoch": 0.827037773359841,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 3536
    },
    {
      "epoch": 0.8272716641328499,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 3537
    },
    {
      "epoch": 0.827505554905859,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 3538
    },
    {
      "epoch": 0.8277394456788679,
      "grad_norm": 11.875,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 3539
    },
    {
      "epoch": 0.827973336451877,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0639,
      "step": 3540
    },
    {
      "epoch": 0.8282072272248859,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 3541
    },
    {
      "epoch": 0.828441117997895,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 3542
    },
    {
      "epoch": 0.8286750087709039,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 3543
    },
    {
      "epoch": 0.828908899543913,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 3544
    },
    {
      "epoch": 0.829142790316922,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 3545
    },
    {
      "epoch": 0.829376681089931,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 3546
    },
    {
      "epoch": 0.82961057186294,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 3547
    },
    {
      "epoch": 0.829844462635949,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 3548
    },
    {
      "epoch": 0.830078353408958,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 3549
    },
    {
      "epoch": 0.830312244181967,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 3550
    },
    {
      "epoch": 0.830546134954976,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 3551
    },
    {
      "epoch": 0.830780025727985,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 3552
    },
    {
      "epoch": 0.831013916500994,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 3553
    },
    {
      "epoch": 0.831247807274003,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 3554
    },
    {
      "epoch": 0.831481698047012,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 3555
    },
    {
      "epoch": 0.8317155888200211,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2198,
      "step": 3556
    },
    {
      "epoch": 0.83194947959303,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 3557
    },
    {
      "epoch": 0.8321833703660391,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 3558
    },
    {
      "epoch": 0.832417261139048,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 3559
    },
    {
      "epoch": 0.8326511519120571,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 3560
    },
    {
      "epoch": 0.832885042685066,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 3561
    },
    {
      "epoch": 0.8331189334580751,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 3562
    },
    {
      "epoch": 0.833352824231084,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 3563
    },
    {
      "epoch": 0.8335867150040931,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 3564
    },
    {
      "epoch": 0.833820605777102,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 3565
    },
    {
      "epoch": 0.8340544965501111,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 3566
    },
    {
      "epoch": 0.83428838732312,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 3567
    },
    {
      "epoch": 0.8345222780961291,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 3568
    },
    {
      "epoch": 0.8347561688691381,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 3569
    },
    {
      "epoch": 0.8349900596421471,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 3570
    },
    {
      "epoch": 0.8352239504151561,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 3571
    },
    {
      "epoch": 0.8354578411881651,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 3572
    },
    {
      "epoch": 0.8356917319611741,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 3573
    },
    {
      "epoch": 0.8359256227341831,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 3574
    },
    {
      "epoch": 0.8361595135071921,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.038,
      "step": 3575
    },
    {
      "epoch": 0.8363934042802011,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 3576
    },
    {
      "epoch": 0.8366272950532101,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 3577
    },
    {
      "epoch": 0.8368611858262192,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 3578
    },
    {
      "epoch": 0.8370950765992281,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 3579
    },
    {
      "epoch": 0.8373289673722372,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 3580
    },
    {
      "epoch": 0.8375628581452462,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 3581
    },
    {
      "epoch": 0.8377967489182552,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 3582
    },
    {
      "epoch": 0.8380306396912642,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 3583
    },
    {
      "epoch": 0.8382645304642732,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 3584
    },
    {
      "epoch": 0.8384984212372822,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 3585
    },
    {
      "epoch": 0.8387323120102912,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 3586
    },
    {
      "epoch": 0.8389662027833003,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 3587
    },
    {
      "epoch": 0.8392000935563092,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 3588
    },
    {
      "epoch": 0.8394339843293183,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 3589
    },
    {
      "epoch": 0.8396678751023272,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8776,
      "step": 3590
    },
    {
      "epoch": 0.8399017658753363,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 3591
    },
    {
      "epoch": 0.8401356566483452,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 3592
    },
    {
      "epoch": 0.8403695474213543,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 3593
    },
    {
      "epoch": 0.8406034381943632,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 3594
    },
    {
      "epoch": 0.8408373289673723,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 3595
    },
    {
      "epoch": 0.8410712197403812,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 3596
    },
    {
      "epoch": 0.8413051105133903,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8075,
      "step": 3597
    },
    {
      "epoch": 0.8415390012863992,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 3598
    },
    {
      "epoch": 0.8417728920594083,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.5972,
      "step": 3599
    },
    {
      "epoch": 0.8420067828324173,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 3600
    },
    {
      "epoch": 0.8420067828324173,
      "eval_runtime": 4.569,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 3600
    },
    {
      "epoch": 0.8422406736054263,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 3601
    },
    {
      "epoch": 0.8424745643784353,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 3602
    },
    {
      "epoch": 0.8427084551514443,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 3603
    },
    {
      "epoch": 0.8429423459244533,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 3604
    },
    {
      "epoch": 0.8431762366974623,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2951,
      "step": 3605
    },
    {
      "epoch": 0.8434101274704713,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 3606
    },
    {
      "epoch": 0.8436440182434803,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 3607
    },
    {
      "epoch": 0.8438779090164893,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 3608
    },
    {
      "epoch": 0.8441117997894984,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 3609
    },
    {
      "epoch": 0.8443456905625073,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 3610
    },
    {
      "epoch": 0.8445795813355164,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9087,
      "step": 3611
    },
    {
      "epoch": 0.8448134721085253,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 3612
    },
    {
      "epoch": 0.8450473628815344,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 3613
    },
    {
      "epoch": 0.8452812536545433,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 3614
    },
    {
      "epoch": 0.8455151444275524,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 3615
    },
    {
      "epoch": 0.8457490352005613,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 3616
    },
    {
      "epoch": 0.8459829259735704,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4693,
      "step": 3617
    },
    {
      "epoch": 0.8462168167465793,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 3618
    },
    {
      "epoch": 0.8464507075195884,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.578,
      "step": 3619
    },
    {
      "epoch": 0.8466845982925973,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 3620
    },
    {
      "epoch": 0.8469184890656064,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 3621
    },
    {
      "epoch": 0.8471523798386154,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 3622
    },
    {
      "epoch": 0.8473862706116244,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 3623
    },
    {
      "epoch": 0.8476201613846334,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 3624
    },
    {
      "epoch": 0.8478540521576424,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 3625
    },
    {
      "epoch": 0.8480879429306514,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 3626
    },
    {
      "epoch": 0.8483218337036604,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7162,
      "step": 3627
    },
    {
      "epoch": 0.8485557244766694,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 3628
    },
    {
      "epoch": 0.8487896152496784,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 3629
    },
    {
      "epoch": 0.8490235060226874,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 3630
    },
    {
      "epoch": 0.8492573967956965,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 3631
    },
    {
      "epoch": 0.8494912875687054,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 3632
    },
    {
      "epoch": 0.8497251783417145,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 3633
    },
    {
      "epoch": 0.8499590691147234,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 3634
    },
    {
      "epoch": 0.8501929598877325,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 3635
    },
    {
      "epoch": 0.8504268506607414,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2526,
      "step": 3636
    },
    {
      "epoch": 0.8506607414337505,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 3637
    },
    {
      "epoch": 0.8508946322067594,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7503,
      "step": 3638
    },
    {
      "epoch": 0.8511285229797685,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 3639
    },
    {
      "epoch": 0.8513624137527774,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 3640
    },
    {
      "epoch": 0.8515963045257865,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 3641
    },
    {
      "epoch": 0.8518301952987954,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 3642
    },
    {
      "epoch": 0.8520640860718045,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 3643
    },
    {
      "epoch": 0.8522979768448135,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3278,
      "step": 3644
    },
    {
      "epoch": 0.8525318676178225,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 3645
    },
    {
      "epoch": 0.8527657583908315,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 3646
    },
    {
      "epoch": 0.8529996491638405,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 3647
    },
    {
      "epoch": 0.8532335399368495,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 3648
    },
    {
      "epoch": 0.8534674307098585,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6481,
      "step": 3649
    },
    {
      "epoch": 0.8537013214828675,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 3650
    },
    {
      "epoch": 0.8539352122558765,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 3651
    },
    {
      "epoch": 0.8541691030288855,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.844,
      "step": 3652
    },
    {
      "epoch": 0.8544029938018946,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 3653
    },
    {
      "epoch": 0.8546368845749035,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 3654
    },
    {
      "epoch": 0.8548707753479126,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 3655
    },
    {
      "epoch": 0.8551046661209215,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 3656
    },
    {
      "epoch": 0.8553385568939306,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 3657
    },
    {
      "epoch": 0.8555724476669395,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 3658
    },
    {
      "epoch": 0.8558063384399486,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6898,
      "step": 3659
    },
    {
      "epoch": 0.8560402292129575,
      "grad_norm": 7.90625,
      "learning_rate": 3e-05,
      "loss": 1.7619,
      "step": 3660
    },
    {
      "epoch": 0.8562741199859666,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 3661
    },
    {
      "epoch": 0.8565080107589755,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2994,
      "step": 3662
    },
    {
      "epoch": 0.8567419015319846,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 3663
    },
    {
      "epoch": 0.8569757923049935,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 3664
    },
    {
      "epoch": 0.8572096830780026,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 3665
    },
    {
      "epoch": 0.8574435738510116,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 3666
    },
    {
      "epoch": 0.8576774646240206,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7421,
      "step": 3667
    },
    {
      "epoch": 0.8579113553970296,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 3668
    },
    {
      "epoch": 0.8581452461700386,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 3669
    },
    {
      "epoch": 0.8583791369430476,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 3670
    },
    {
      "epoch": 0.8586130277160566,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8891,
      "step": 3671
    },
    {
      "epoch": 0.8588469184890656,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 3672
    },
    {
      "epoch": 0.8590808092620746,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.216,
      "step": 3673
    },
    {
      "epoch": 0.8593147000350836,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 3674
    },
    {
      "epoch": 0.8595485908080926,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 3675
    },
    {
      "epoch": 0.8597824815811016,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 3676
    },
    {
      "epoch": 0.8600163723541107,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 3677
    },
    {
      "epoch": 0.8602502631271196,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 3678
    },
    {
      "epoch": 0.8604841539001287,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 3679
    },
    {
      "epoch": 0.8607180446731376,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 3680
    },
    {
      "epoch": 0.8609519354461467,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 3681
    },
    {
      "epoch": 0.8611858262191556,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 3682
    },
    {
      "epoch": 0.8614197169921647,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 3683
    },
    {
      "epoch": 0.8616536077651736,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 3684
    },
    {
      "epoch": 0.8618874985381827,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 3685
    },
    {
      "epoch": 0.8621213893111916,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 3686
    },
    {
      "epoch": 0.8623552800842007,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 3687
    },
    {
      "epoch": 0.8625891708572097,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 3688
    },
    {
      "epoch": 0.8628230616302187,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 3689
    },
    {
      "epoch": 0.8630569524032277,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 3690
    },
    {
      "epoch": 0.8632908431762367,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 3691
    },
    {
      "epoch": 0.8635247339492457,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 3692
    },
    {
      "epoch": 0.8637586247222547,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 3693
    },
    {
      "epoch": 0.8639925154952637,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 3694
    },
    {
      "epoch": 0.8642264062682727,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 3695
    },
    {
      "epoch": 0.8644602970412817,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 3696
    },
    {
      "epoch": 0.8646941878142907,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 3697
    },
    {
      "epoch": 0.8649280785872997,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 3698
    },
    {
      "epoch": 0.8651619693603088,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 3699
    },
    {
      "epoch": 0.8653958601333177,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9756,
      "step": 3700
    },
    {
      "epoch": 0.8653958601333177,
      "eval_runtime": 4.6165,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 3700
    },
    {
      "epoch": 0.8656297509063268,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 3701
    },
    {
      "epoch": 0.8658636416793357,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7297,
      "step": 3702
    },
    {
      "epoch": 0.8660975324523448,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 3703
    },
    {
      "epoch": 0.8663314232253537,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 3704
    },
    {
      "epoch": 0.8665653139983628,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 3705
    },
    {
      "epoch": 0.8667992047713717,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 3706
    },
    {
      "epoch": 0.8670330955443808,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 3707
    },
    {
      "epoch": 0.8672669863173897,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 3708
    },
    {
      "epoch": 0.8675008770903988,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 3709
    },
    {
      "epoch": 0.8677347678634078,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 3710
    },
    {
      "epoch": 0.8679686586364168,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5962,
      "step": 3711
    },
    {
      "epoch": 0.8682025494094258,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 3712
    },
    {
      "epoch": 0.8684364401824348,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1953,
      "step": 3713
    },
    {
      "epoch": 0.8686703309554438,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 3714
    },
    {
      "epoch": 0.8689042217284528,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 3715
    },
    {
      "epoch": 0.8691381125014618,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 3716
    },
    {
      "epoch": 0.8693720032744708,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 3717
    },
    {
      "epoch": 0.8696058940474798,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 3718
    },
    {
      "epoch": 0.8698397848204888,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.2097,
      "step": 3719
    },
    {
      "epoch": 0.8700736755934978,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7023,
      "step": 3720
    },
    {
      "epoch": 0.8703075663665069,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 3721
    },
    {
      "epoch": 0.8705414571395158,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 3722
    },
    {
      "epoch": 0.8707753479125249,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 3723
    },
    {
      "epoch": 0.8710092386855338,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 3724
    },
    {
      "epoch": 0.8712431294585429,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 3725
    },
    {
      "epoch": 0.8714770202315518,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.375,
      "step": 3726
    },
    {
      "epoch": 0.8717109110045609,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 3727
    },
    {
      "epoch": 0.8719448017775698,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 3728
    },
    {
      "epoch": 0.8721786925505789,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 3729
    },
    {
      "epoch": 0.8724125833235878,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 3730
    },
    {
      "epoch": 0.8726464740965969,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.4743,
      "step": 3731
    },
    {
      "epoch": 0.8728803648696059,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 3732
    },
    {
      "epoch": 0.8731142556426149,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 3733
    },
    {
      "epoch": 0.8733481464156239,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 3734
    },
    {
      "epoch": 0.8735820371886329,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 3735
    },
    {
      "epoch": 0.8738159279616419,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 3736
    },
    {
      "epoch": 0.8740498187346509,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 3737
    },
    {
      "epoch": 0.8742837095076599,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 3738
    },
    {
      "epoch": 0.8745176002806689,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 3739
    },
    {
      "epoch": 0.8747514910536779,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 3740
    },
    {
      "epoch": 0.874985381826687,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 3741
    },
    {
      "epoch": 0.8752192725996959,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7771,
      "step": 3742
    },
    {
      "epoch": 0.875453163372705,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1954,
      "step": 3743
    },
    {
      "epoch": 0.8756870541457139,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 3744
    },
    {
      "epoch": 0.875920944918723,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 3745
    },
    {
      "epoch": 0.8761548356917319,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 3746
    },
    {
      "epoch": 0.876388726464741,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 3747
    },
    {
      "epoch": 0.8766226172377499,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 3748
    },
    {
      "epoch": 0.876856508010759,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 3749
    },
    {
      "epoch": 0.8770903987837679,
      "grad_norm": 8.5,
      "learning_rate": 3e-05,
      "loss": 2.4284,
      "step": 3750
    },
    {
      "epoch": 0.877324289556777,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 3751
    },
    {
      "epoch": 0.8775581803297859,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 3752
    },
    {
      "epoch": 0.877792071102795,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 3753
    },
    {
      "epoch": 0.878025961875804,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 3754
    },
    {
      "epoch": 0.878259852648813,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 3755
    },
    {
      "epoch": 0.878493743421822,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 3756
    },
    {
      "epoch": 0.878727634194831,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 3757
    },
    {
      "epoch": 0.87896152496784,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 3758
    },
    {
      "epoch": 0.879195415740849,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6591,
      "step": 3759
    },
    {
      "epoch": 0.879429306513858,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 3760
    },
    {
      "epoch": 0.879663197286867,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 3761
    },
    {
      "epoch": 0.879897088059876,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 3762
    },
    {
      "epoch": 0.880130978832885,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 3763
    },
    {
      "epoch": 0.880364869605894,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 3764
    },
    {
      "epoch": 0.8805987603789031,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 3765
    },
    {
      "epoch": 0.880832651151912,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 3766
    },
    {
      "epoch": 0.8810665419249211,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 3767
    },
    {
      "epoch": 0.88130043269793,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 3768
    },
    {
      "epoch": 0.8815343234709391,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6259,
      "step": 3769
    },
    {
      "epoch": 0.881768214243948,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 3770
    },
    {
      "epoch": 0.8820021050169571,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 3771
    },
    {
      "epoch": 0.882235995789966,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2991,
      "step": 3772
    },
    {
      "epoch": 0.8824698865629751,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 3773
    },
    {
      "epoch": 0.882703777335984,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 3774
    },
    {
      "epoch": 0.8829376681089931,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 3775
    },
    {
      "epoch": 0.883171558882002,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5829,
      "step": 3776
    },
    {
      "epoch": 0.8834054496550111,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 3777
    },
    {
      "epoch": 0.8836393404280201,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 3778
    },
    {
      "epoch": 0.8838732312010291,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 3779
    },
    {
      "epoch": 0.8841071219740382,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 3780
    },
    {
      "epoch": 0.8843410127470471,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 3781
    },
    {
      "epoch": 0.8845749035200562,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3645,
      "step": 3782
    },
    {
      "epoch": 0.8848087942930651,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 3783
    },
    {
      "epoch": 0.8850426850660742,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 3784
    },
    {
      "epoch": 0.8852765758390831,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 3785
    },
    {
      "epoch": 0.8855104666120922,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 3786
    },
    {
      "epoch": 0.8857443573851012,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 3787
    },
    {
      "epoch": 0.8859782481581102,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7814,
      "step": 3788
    },
    {
      "epoch": 0.8862121389311192,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 3789
    },
    {
      "epoch": 0.8864460297041282,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 3790
    },
    {
      "epoch": 0.8866799204771372,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 3791
    },
    {
      "epoch": 0.8869138112501462,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 3792
    },
    {
      "epoch": 0.8871477020231552,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 3793
    },
    {
      "epoch": 0.8873815927961642,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 3794
    },
    {
      "epoch": 0.8876154835691732,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 3795
    },
    {
      "epoch": 0.8878493743421823,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 3796
    },
    {
      "epoch": 0.8880832651151912,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 3797
    },
    {
      "epoch": 0.8883171558882003,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 3798
    },
    {
      "epoch": 0.8885510466612092,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.832,
      "step": 3799
    },
    {
      "epoch": 0.8887849374342183,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2985,
      "step": 3800
    },
    {
      "epoch": 0.8887849374342183,
      "eval_runtime": 4.6038,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 3800
    },
    {
      "epoch": 0.8890188282072272,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 3801
    },
    {
      "epoch": 0.8892527189802363,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 3802
    },
    {
      "epoch": 0.8894866097532452,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 3803
    },
    {
      "epoch": 0.8897205005262543,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8617,
      "step": 3804
    },
    {
      "epoch": 0.8899543912992632,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 3805
    },
    {
      "epoch": 0.8901882820722723,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 3806
    },
    {
      "epoch": 0.8904221728452812,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 3807
    },
    {
      "epoch": 0.8906560636182903,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 3808
    },
    {
      "epoch": 0.8908899543912993,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 3809
    },
    {
      "epoch": 0.8911238451643083,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 3810
    },
    {
      "epoch": 0.8913577359373173,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 3811
    },
    {
      "epoch": 0.8915916267103263,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 3812
    },
    {
      "epoch": 0.8918255174833353,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 3813
    },
    {
      "epoch": 0.8920594082563443,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 3814
    },
    {
      "epoch": 0.8922932990293533,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 3815
    },
    {
      "epoch": 0.8925271898023623,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 3816
    },
    {
      "epoch": 0.8927610805753713,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3181,
      "step": 3817
    },
    {
      "epoch": 0.8929949713483804,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 3818
    },
    {
      "epoch": 0.8932288621213893,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 3819
    },
    {
      "epoch": 0.8934627528943984,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 3820
    },
    {
      "epoch": 0.8936966436674073,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 3821
    },
    {
      "epoch": 0.8939305344404164,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8634,
      "step": 3822
    },
    {
      "epoch": 0.8941644252134253,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 3823
    },
    {
      "epoch": 0.8943983159864344,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 3824
    },
    {
      "epoch": 0.8946322067594433,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 3825
    },
    {
      "epoch": 0.8948660975324524,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 3826
    },
    {
      "epoch": 0.8950999883054613,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 3827
    },
    {
      "epoch": 0.8953338790784704,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 3828
    },
    {
      "epoch": 0.8955677698514793,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7407,
      "step": 3829
    },
    {
      "epoch": 0.8958016606244884,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2677,
      "step": 3830
    },
    {
      "epoch": 0.8960355513974974,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 3831
    },
    {
      "epoch": 0.8962694421705064,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 3832
    },
    {
      "epoch": 0.8965033329435154,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6821,
      "step": 3833
    },
    {
      "epoch": 0.8967372237165244,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 3834
    },
    {
      "epoch": 0.8969711144895334,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 3835
    },
    {
      "epoch": 0.8972050052625424,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5385,
      "step": 3836
    },
    {
      "epoch": 0.8974388960355514,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 3837
    },
    {
      "epoch": 0.8976727868085604,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2132,
      "step": 3838
    },
    {
      "epoch": 0.8979066775815694,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 3839
    },
    {
      "epoch": 0.8981405683545785,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3278,
      "step": 3840
    },
    {
      "epoch": 0.8983744591275874,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 3841
    },
    {
      "epoch": 0.8986083499005965,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 3842
    },
    {
      "epoch": 0.8988422406736054,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7206,
      "step": 3843
    },
    {
      "epoch": 0.8990761314466145,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 3844
    },
    {
      "epoch": 0.8993100222196234,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8728,
      "step": 3845
    },
    {
      "epoch": 0.8995439129926325,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 3846
    },
    {
      "epoch": 0.8997778037656414,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 3847
    },
    {
      "epoch": 0.9000116945386505,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.412,
      "step": 3848
    },
    {
      "epoch": 0.9002455853116594,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 3849
    },
    {
      "epoch": 0.9004794760846685,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6823,
      "step": 3850
    },
    {
      "epoch": 0.9007133668576774,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 3851
    },
    {
      "epoch": 0.9009472576306865,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 3852
    },
    {
      "epoch": 0.9011811484036955,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 3853
    },
    {
      "epoch": 0.9014150391767045,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 3854
    },
    {
      "epoch": 0.9016489299497135,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 3855
    },
    {
      "epoch": 0.9018828207227225,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.476,
      "step": 3856
    },
    {
      "epoch": 0.9021167114957315,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 3857
    },
    {
      "epoch": 0.9023506022687405,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 3858
    },
    {
      "epoch": 0.9025844930417495,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 3859
    },
    {
      "epoch": 0.9028183838147585,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 3860
    },
    {
      "epoch": 0.9030522745877675,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 3861
    },
    {
      "epoch": 0.9032861653607765,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 3862
    },
    {
      "epoch": 0.9035200561337855,
      "grad_norm": 8.9375,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 3863
    },
    {
      "epoch": 0.9037539469067946,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 3864
    },
    {
      "epoch": 0.9039878376798035,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 3865
    },
    {
      "epoch": 0.9042217284528126,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2481,
      "step": 3866
    },
    {
      "epoch": 0.9044556192258215,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 3867
    },
    {
      "epoch": 0.9046895099988306,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 3868
    },
    {
      "epoch": 0.9049234007718395,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 3869
    },
    {
      "epoch": 0.9051572915448486,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 3870
    },
    {
      "epoch": 0.9053911823178575,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 3871
    },
    {
      "epoch": 0.9056250730908666,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 3872
    },
    {
      "epoch": 0.9058589638638755,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6367,
      "step": 3873
    },
    {
      "epoch": 0.9060928546368846,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 3874
    },
    {
      "epoch": 0.9063267454098936,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 3875
    },
    {
      "epoch": 0.9065606361829026,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 3876
    },
    {
      "epoch": 0.9067945269559116,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 3877
    },
    {
      "epoch": 0.9070284177289206,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 3878
    },
    {
      "epoch": 0.9072623085019296,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 3879
    },
    {
      "epoch": 0.9074961992749386,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8131,
      "step": 3880
    },
    {
      "epoch": 0.9077300900479476,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 3881
    },
    {
      "epoch": 0.9079639808209566,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 3882
    },
    {
      "epoch": 0.9081978715939656,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 3883
    },
    {
      "epoch": 0.9084317623669746,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2447,
      "step": 3884
    },
    {
      "epoch": 0.9086656531399836,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 3885
    },
    {
      "epoch": 0.9088995439129927,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 3886
    },
    {
      "epoch": 0.9091334346860016,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 3887
    },
    {
      "epoch": 0.9093673254590107,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8644,
      "step": 3888
    },
    {
      "epoch": 0.9096012162320196,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 3889
    },
    {
      "epoch": 0.9098351070050287,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 3890
    },
    {
      "epoch": 0.9100689977780376,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 3891
    },
    {
      "epoch": 0.9103028885510467,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 3892
    },
    {
      "epoch": 0.9105367793240556,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 3893
    },
    {
      "epoch": 0.9107706700970647,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 3894
    },
    {
      "epoch": 0.9110045608700736,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 3895
    },
    {
      "epoch": 0.9112384516430827,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.6178,
      "step": 3896
    },
    {
      "epoch": 0.9114723424160917,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 3897
    },
    {
      "epoch": 0.9117062331891007,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 3898
    },
    {
      "epoch": 0.9119401239621097,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 3899
    },
    {
      "epoch": 0.9121740147351187,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 3900
    },
    {
      "epoch": 0.9121740147351187,
      "eval_runtime": 4.6058,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 3900
    },
    {
      "epoch": 0.9124079055081277,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 3901
    },
    {
      "epoch": 0.9126417962811367,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8711,
      "step": 3902
    },
    {
      "epoch": 0.9128756870541457,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 3903
    },
    {
      "epoch": 0.9131095778271547,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 3904
    },
    {
      "epoch": 0.9133434686001637,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 3905
    },
    {
      "epoch": 0.9135773593731727,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7732,
      "step": 3906
    },
    {
      "epoch": 0.9138112501461817,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6246,
      "step": 3907
    },
    {
      "epoch": 0.9140451409191908,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 3908
    },
    {
      "epoch": 0.9142790316921997,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 3909
    },
    {
      "epoch": 0.9145129224652088,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 3910
    },
    {
      "epoch": 0.9147468132382177,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 3911
    },
    {
      "epoch": 0.9149807040112268,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 3912
    },
    {
      "epoch": 0.9152145947842357,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3496,
      "step": 3913
    },
    {
      "epoch": 0.9154484855572448,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6524,
      "step": 3914
    },
    {
      "epoch": 0.9156823763302537,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 3915
    },
    {
      "epoch": 0.9159162671032628,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 3916
    },
    {
      "epoch": 0.9161501578762717,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 3917
    },
    {
      "epoch": 0.9163840486492808,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 3918
    },
    {
      "epoch": 0.9166179394222898,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6687,
      "step": 3919
    },
    {
      "epoch": 0.9168518301952988,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 3920
    },
    {
      "epoch": 0.9170857209683078,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 3921
    },
    {
      "epoch": 0.9173196117413168,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 3922
    },
    {
      "epoch": 0.9175535025143258,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 3923
    },
    {
      "epoch": 0.9177873932873348,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 3924
    },
    {
      "epoch": 0.9180212840603438,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 3925
    },
    {
      "epoch": 0.9182551748333528,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 3926
    },
    {
      "epoch": 0.9184890656063618,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 3927
    },
    {
      "epoch": 0.9187229563793708,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 3928
    },
    {
      "epoch": 0.9189568471523798,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 3929
    },
    {
      "epoch": 0.9191907379253889,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 3930
    },
    {
      "epoch": 0.9194246286983978,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 3931
    },
    {
      "epoch": 0.9196585194714069,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 3932
    },
    {
      "epoch": 0.9198924102444158,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 3933
    },
    {
      "epoch": 0.9201263010174249,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 3934
    },
    {
      "epoch": 0.9203601917904338,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2496,
      "step": 3935
    },
    {
      "epoch": 0.9205940825634429,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3345,
      "step": 3936
    },
    {
      "epoch": 0.9208279733364518,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 3937
    },
    {
      "epoch": 0.9210618641094609,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 3938
    },
    {
      "epoch": 0.9212957548824698,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 3939
    },
    {
      "epoch": 0.9215296456554789,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 3940
    },
    {
      "epoch": 0.9217635364284879,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 3941
    },
    {
      "epoch": 0.9219974272014969,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 3942
    },
    {
      "epoch": 0.9222313179745059,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7694,
      "step": 3943
    },
    {
      "epoch": 0.9224652087475149,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 3944
    },
    {
      "epoch": 0.9226990995205239,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6679,
      "step": 3945
    },
    {
      "epoch": 0.9229329902935329,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 3946
    },
    {
      "epoch": 0.9231668810665419,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8053,
      "step": 3947
    },
    {
      "epoch": 0.9234007718395509,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 3948
    },
    {
      "epoch": 0.9236346626125599,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2295,
      "step": 3949
    },
    {
      "epoch": 0.923868553385569,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 3950
    },
    {
      "epoch": 0.9241024441585779,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 3951
    },
    {
      "epoch": 0.924336334931587,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7064,
      "step": 3952
    },
    {
      "epoch": 0.9245702257045959,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 3953
    },
    {
      "epoch": 0.924804116477605,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 3954
    },
    {
      "epoch": 0.9250380072506139,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 3955
    },
    {
      "epoch": 0.925271898023623,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 3956
    },
    {
      "epoch": 0.9255057887966319,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 3957
    },
    {
      "epoch": 0.925739679569641,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 3958
    },
    {
      "epoch": 0.9259735703426499,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2832,
      "step": 3959
    },
    {
      "epoch": 0.926207461115659,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1255,
      "step": 3960
    },
    {
      "epoch": 0.9264413518886679,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 3961
    },
    {
      "epoch": 0.926675242661677,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 3962
    },
    {
      "epoch": 0.926909133434686,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 3963
    },
    {
      "epoch": 0.927143024207695,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 3964
    },
    {
      "epoch": 0.927376914980704,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 3965
    },
    {
      "epoch": 0.927610805753713,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 3966
    },
    {
      "epoch": 0.927844696526722,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.6175,
      "step": 3967
    },
    {
      "epoch": 0.928078587299731,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 3968
    },
    {
      "epoch": 0.92831247807274,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3145,
      "step": 3969
    },
    {
      "epoch": 0.928546368845749,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 3970
    },
    {
      "epoch": 0.928780259618758,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8501,
      "step": 3971
    },
    {
      "epoch": 0.929014150391767,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 3972
    },
    {
      "epoch": 0.929248041164776,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 3973
    },
    {
      "epoch": 0.929481931937785,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 3974
    },
    {
      "epoch": 0.929715822710794,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 3975
    },
    {
      "epoch": 0.9299497134838031,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 3976
    },
    {
      "epoch": 0.930183604256812,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 3977
    },
    {
      "epoch": 0.9304174950298211,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 3978
    },
    {
      "epoch": 0.9306513858028301,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 3979
    },
    {
      "epoch": 0.9308852765758391,
      "grad_norm": 7.59375,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 3980
    },
    {
      "epoch": 0.9311191673488481,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 3981
    },
    {
      "epoch": 0.9313530581218571,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 3982
    },
    {
      "epoch": 0.9315869488948662,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 3983
    },
    {
      "epoch": 0.9318208396678751,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.219,
      "step": 3984
    },
    {
      "epoch": 0.9320547304408842,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 3985
    },
    {
      "epoch": 0.9322886212138931,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 3986
    },
    {
      "epoch": 0.9325225119869022,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 3987
    },
    {
      "epoch": 0.9327564027599111,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9759,
      "step": 3988
    },
    {
      "epoch": 0.9329902935329202,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 3989
    },
    {
      "epoch": 0.9332241843059291,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7375,
      "step": 3990
    },
    {
      "epoch": 0.9334580750789382,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 3991
    },
    {
      "epoch": 0.9336919658519471,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 3992
    },
    {
      "epoch": 0.9339258566249562,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9569,
      "step": 3993
    },
    {
      "epoch": 0.9341597473979651,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.659,
      "step": 3994
    },
    {
      "epoch": 0.9343936381709742,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 3995
    },
    {
      "epoch": 0.9346275289439832,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 3996
    },
    {
      "epoch": 0.9348614197169922,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 3997
    },
    {
      "epoch": 0.9350953104900012,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0827,
      "step": 3998
    },
    {
      "epoch": 0.9353292012630102,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 3999
    },
    {
      "epoch": 0.9355630920360192,
      "grad_norm": 9.9375,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 4000
    },
    {
      "epoch": 0.9355630920360192,
      "eval_runtime": 4.598,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 4000
    },
    {
      "epoch": 0.9357969828090282,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8011,
      "step": 4001
    },
    {
      "epoch": 0.9360308735820372,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 4002
    },
    {
      "epoch": 0.9362647643550462,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.4373,
      "step": 4003
    },
    {
      "epoch": 0.9364986551280552,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 4004
    },
    {
      "epoch": 0.9367325459010643,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 4005
    },
    {
      "epoch": 0.9369664366740732,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 4006
    },
    {
      "epoch": 0.9372003274470823,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6175,
      "step": 4007
    },
    {
      "epoch": 0.9374342182200912,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 4008
    },
    {
      "epoch": 0.9376681089931003,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 4009
    },
    {
      "epoch": 0.9379019997661092,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 4010
    },
    {
      "epoch": 0.9381358905391183,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 4011
    },
    {
      "epoch": 0.9383697813121272,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 4012
    },
    {
      "epoch": 0.9386036720851363,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 4013
    },
    {
      "epoch": 0.9388375628581452,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 4014
    },
    {
      "epoch": 0.9390714536311543,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 4015
    },
    {
      "epoch": 0.9393053444041632,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 4016
    },
    {
      "epoch": 0.9395392351771723,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 4017
    },
    {
      "epoch": 0.9397731259501813,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 4018
    },
    {
      "epoch": 0.9400070167231903,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 4019
    },
    {
      "epoch": 0.9402409074961993,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2597,
      "step": 4020
    },
    {
      "epoch": 0.9404747982692083,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 4021
    },
    {
      "epoch": 0.9407086890422173,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 4022
    },
    {
      "epoch": 0.9409425798152263,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 4023
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 4024
    },
    {
      "epoch": 0.9414103613612443,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 4025
    },
    {
      "epoch": 0.9416442521342533,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 4026
    },
    {
      "epoch": 0.9418781429072624,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 4027
    },
    {
      "epoch": 0.9421120336802713,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 4028
    },
    {
      "epoch": 0.9423459244532804,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 4029
    },
    {
      "epoch": 0.9425798152262893,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 4030
    },
    {
      "epoch": 0.9428137059992984,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 4031
    },
    {
      "epoch": 0.9430475967723073,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 4032
    },
    {
      "epoch": 0.9432814875453164,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 4033
    },
    {
      "epoch": 0.9435153783183253,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 4034
    },
    {
      "epoch": 0.9437492690913344,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 4035
    },
    {
      "epoch": 0.9439831598643433,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6552,
      "step": 4036
    },
    {
      "epoch": 0.9442170506373524,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 4037
    },
    {
      "epoch": 0.9444509414103613,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 4038
    },
    {
      "epoch": 0.9446848321833704,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 4039
    },
    {
      "epoch": 0.9449187229563794,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3031,
      "step": 4040
    },
    {
      "epoch": 0.9451526137293884,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 4041
    },
    {
      "epoch": 0.9453865045023974,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 4042
    },
    {
      "epoch": 0.9456203952754064,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 4043
    },
    {
      "epoch": 0.9458542860484154,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 4044
    },
    {
      "epoch": 0.9460881768214244,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2933,
      "step": 4045
    },
    {
      "epoch": 0.9463220675944334,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 4046
    },
    {
      "epoch": 0.9465559583674424,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 4047
    },
    {
      "epoch": 0.9467898491404514,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 4048
    },
    {
      "epoch": 0.9470237399134604,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9313,
      "step": 4049
    },
    {
      "epoch": 0.9472576306864694,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 4050
    },
    {
      "epoch": 0.9474915214594785,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 4051
    },
    {
      "epoch": 0.9477254122324874,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 4052
    },
    {
      "epoch": 0.9479593030054965,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 4053
    },
    {
      "epoch": 0.9481931937785054,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 4054
    },
    {
      "epoch": 0.9484270845515145,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 4055
    },
    {
      "epoch": 0.9486609753245234,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 4056
    },
    {
      "epoch": 0.9488948660975325,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 4057
    },
    {
      "epoch": 0.9491287568705414,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6053,
      "step": 4058
    },
    {
      "epoch": 0.9493626476435505,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 4059
    },
    {
      "epoch": 0.9495965384165594,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 4060
    },
    {
      "epoch": 0.9498304291895685,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7548,
      "step": 4061
    },
    {
      "epoch": 0.9500643199625775,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 4062
    },
    {
      "epoch": 0.9502982107355865,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 4063
    },
    {
      "epoch": 0.9505321015085955,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5147,
      "step": 4064
    },
    {
      "epoch": 0.9507659922816045,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3113,
      "step": 4065
    },
    {
      "epoch": 0.9509998830546135,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 4066
    },
    {
      "epoch": 0.9512337738276225,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 4067
    },
    {
      "epoch": 0.9514676646006315,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 4068
    },
    {
      "epoch": 0.9517015553736405,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 4069
    },
    {
      "epoch": 0.9519354461466495,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 4070
    },
    {
      "epoch": 0.9521693369196585,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 4071
    },
    {
      "epoch": 0.9524032276926675,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 4072
    },
    {
      "epoch": 0.9526371184656766,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 4073
    },
    {
      "epoch": 0.9528710092386855,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 4074
    },
    {
      "epoch": 0.9531049000116946,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 4075
    },
    {
      "epoch": 0.9533387907847035,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 4076
    },
    {
      "epoch": 0.9535726815577126,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 4077
    },
    {
      "epoch": 0.9538065723307215,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 4078
    },
    {
      "epoch": 0.9540404631037306,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 4079
    },
    {
      "epoch": 0.9542743538767395,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 4080
    },
    {
      "epoch": 0.9545082446497486,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 4081
    },
    {
      "epoch": 0.9547421354227575,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 4082
    },
    {
      "epoch": 0.9549760261957666,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 4083
    },
    {
      "epoch": 0.9552099169687756,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 4084
    },
    {
      "epoch": 0.9554438077417846,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 4085
    },
    {
      "epoch": 0.9556776985147936,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 4086
    },
    {
      "epoch": 0.9559115892878026,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7281,
      "step": 4087
    },
    {
      "epoch": 0.9561454800608116,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4179,
      "step": 4088
    },
    {
      "epoch": 0.9563793708338206,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 4089
    },
    {
      "epoch": 0.9566132616068296,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 4090
    },
    {
      "epoch": 0.9568471523798386,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6947,
      "step": 4091
    },
    {
      "epoch": 0.9570810431528476,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 4092
    },
    {
      "epoch": 0.9573149339258566,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8501,
      "step": 4093
    },
    {
      "epoch": 0.9575488246988656,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 4094
    },
    {
      "epoch": 0.9577827154718747,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 4095
    },
    {
      "epoch": 0.9580166062448836,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8305,
      "step": 4096
    },
    {
      "epoch": 0.9582504970178927,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7607,
      "step": 4097
    },
    {
      "epoch": 0.9584843877909016,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1768,
      "step": 4098
    },
    {
      "epoch": 0.9587182785639107,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 4099
    },
    {
      "epoch": 0.9589521693369196,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 4100
    },
    {
      "epoch": 0.9589521693369196,
      "eval_runtime": 4.6191,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 4100
    },
    {
      "epoch": 0.9591860601099287,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 4101
    },
    {
      "epoch": 0.9594199508829376,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 4102
    },
    {
      "epoch": 0.9596538416559467,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 4103
    },
    {
      "epoch": 0.9598877324289556,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 4104
    },
    {
      "epoch": 0.9601216232019647,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 4105
    },
    {
      "epoch": 0.9603555139749737,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 4106
    },
    {
      "epoch": 0.9605894047479827,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 4107
    },
    {
      "epoch": 0.9608232955209917,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2612,
      "step": 4108
    },
    {
      "epoch": 0.9610571862940007,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.491,
      "step": 4109
    },
    {
      "epoch": 0.9612910770670097,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 4110
    },
    {
      "epoch": 0.9615249678400187,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2476,
      "step": 4111
    },
    {
      "epoch": 0.9617588586130277,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 4112
    },
    {
      "epoch": 0.9619927493860367,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 4113
    },
    {
      "epoch": 0.9622266401590457,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 4114
    },
    {
      "epoch": 0.9624605309320547,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 4115
    },
    {
      "epoch": 0.9626944217050637,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 4116
    },
    {
      "epoch": 0.9629283124780728,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 4117
    },
    {
      "epoch": 0.9631622032510817,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 4118
    },
    {
      "epoch": 0.9633960940240908,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 4119
    },
    {
      "epoch": 0.9636299847970997,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 4120
    },
    {
      "epoch": 0.9638638755701088,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3805,
      "step": 4121
    },
    {
      "epoch": 0.9640977663431177,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 4122
    },
    {
      "epoch": 0.9643316571161268,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 4123
    },
    {
      "epoch": 0.9645655478891357,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 4124
    },
    {
      "epoch": 0.9647994386621448,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2801,
      "step": 4125
    },
    {
      "epoch": 0.9650333294351537,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 4126
    },
    {
      "epoch": 0.9652672202081628,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1548,
      "step": 4127
    },
    {
      "epoch": 0.9655011109811718,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 4128
    },
    {
      "epoch": 0.9657350017541808,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 4129
    },
    {
      "epoch": 0.9659688925271898,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 4130
    },
    {
      "epoch": 0.9662027833001988,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6936,
      "step": 4131
    },
    {
      "epoch": 0.9664366740732078,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 4132
    },
    {
      "epoch": 0.9666705648462168,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 4133
    },
    {
      "epoch": 0.9669044556192258,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6604,
      "step": 4134
    },
    {
      "epoch": 0.9671383463922348,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3286,
      "step": 4135
    },
    {
      "epoch": 0.9673722371652438,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 4136
    },
    {
      "epoch": 0.9676061279382528,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 4137
    },
    {
      "epoch": 0.9678400187112618,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0299,
      "step": 4138
    },
    {
      "epoch": 0.9680739094842709,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 4139
    },
    {
      "epoch": 0.9683078002572798,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 4140
    },
    {
      "epoch": 0.9685416910302889,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 4141
    },
    {
      "epoch": 0.9687755818032978,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 4142
    },
    {
      "epoch": 0.9690094725763069,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 4143
    },
    {
      "epoch": 0.9692433633493158,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 4144
    },
    {
      "epoch": 0.9694772541223249,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7552,
      "step": 4145
    },
    {
      "epoch": 0.9697111448953338,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 4146
    },
    {
      "epoch": 0.9699450356683429,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.788,
      "step": 4147
    },
    {
      "epoch": 0.9701789264413518,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 4148
    },
    {
      "epoch": 0.9704128172143609,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 4149
    },
    {
      "epoch": 0.9706467079873699,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 4150
    },
    {
      "epoch": 0.9708805987603789,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2703,
      "step": 4151
    },
    {
      "epoch": 0.9711144895333879,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 4152
    },
    {
      "epoch": 0.9713483803063969,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 4153
    },
    {
      "epoch": 0.9715822710794059,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 4154
    },
    {
      "epoch": 0.9718161618524149,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7528,
      "step": 4155
    },
    {
      "epoch": 0.9720500526254239,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 4156
    },
    {
      "epoch": 0.9722839433984329,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 4157
    },
    {
      "epoch": 0.9725178341714419,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 4158
    },
    {
      "epoch": 0.972751724944451,
      "grad_norm": 7.03125,
      "learning_rate": 3e-05,
      "loss": 1.8903,
      "step": 4159
    },
    {
      "epoch": 0.9729856157174599,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 4160
    },
    {
      "epoch": 0.973219506490469,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 4161
    },
    {
      "epoch": 0.9734533972634779,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 4162
    },
    {
      "epoch": 0.973687288036487,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 4163
    },
    {
      "epoch": 0.9739211788094959,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7421,
      "step": 4164
    },
    {
      "epoch": 0.974155069582505,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 4165
    },
    {
      "epoch": 0.9743889603555139,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 4166
    },
    {
      "epoch": 0.974622851128523,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 4167
    },
    {
      "epoch": 0.9748567419015319,
      "grad_norm": 7.28125,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 4168
    },
    {
      "epoch": 0.975090632674541,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 4169
    },
    {
      "epoch": 0.9753245234475499,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 4170
    },
    {
      "epoch": 0.975558414220559,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 4171
    },
    {
      "epoch": 0.975792304993568,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 4172
    },
    {
      "epoch": 0.976026195766577,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 4173
    },
    {
      "epoch": 0.976260086539586,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 4174
    },
    {
      "epoch": 0.976493977312595,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 4175
    },
    {
      "epoch": 0.976727868085604,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 4176
    },
    {
      "epoch": 0.976961758858613,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7201,
      "step": 4177
    },
    {
      "epoch": 0.9771956496316221,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 4178
    },
    {
      "epoch": 0.977429540404631,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.511,
      "step": 4179
    },
    {
      "epoch": 0.9776634311776401,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 4180
    },
    {
      "epoch": 0.977897321950649,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 4181
    },
    {
      "epoch": 0.9781312127236581,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4019,
      "step": 4182
    },
    {
      "epoch": 0.978365103496667,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 4183
    },
    {
      "epoch": 0.9785989942696761,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2331,
      "step": 4184
    },
    {
      "epoch": 0.9788328850426851,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2558,
      "step": 4185
    },
    {
      "epoch": 0.9790667758156941,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 4186
    },
    {
      "epoch": 0.9793006665887031,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 4187
    },
    {
      "epoch": 0.9795345573617121,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 4188
    },
    {
      "epoch": 0.9797684481347211,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 4189
    },
    {
      "epoch": 0.9800023389077301,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2558,
      "step": 4190
    },
    {
      "epoch": 0.9802362296807391,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 4191
    },
    {
      "epoch": 0.9804701204537482,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 4192
    },
    {
      "epoch": 0.9807040112267571,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1884,
      "step": 4193
    },
    {
      "epoch": 0.9809379019997662,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 4194
    },
    {
      "epoch": 0.9811717927727751,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3559,
      "step": 4195
    },
    {
      "epoch": 0.9814056835457842,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 4196
    },
    {
      "epoch": 0.9816395743187931,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 4197
    },
    {
      "epoch": 0.9818734650918022,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1556,
      "step": 4198
    },
    {
      "epoch": 0.9821073558648111,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 4199
    },
    {
      "epoch": 0.9823412466378202,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 4200
    },
    {
      "epoch": 0.9823412466378202,
      "eval_runtime": 4.6317,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 4200
    },
    {
      "epoch": 0.9825751374108291,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 4201
    },
    {
      "epoch": 0.9828090281838382,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 4202
    },
    {
      "epoch": 0.9830429189568471,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.5088,
      "step": 4203
    },
    {
      "epoch": 0.9832768097298562,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 4204
    },
    {
      "epoch": 0.9835107005028652,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8201,
      "step": 4205
    },
    {
      "epoch": 0.9837445912758742,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 4206
    },
    {
      "epoch": 0.9839784820488832,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.668,
      "step": 4207
    },
    {
      "epoch": 0.9842123728218922,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 4208
    },
    {
      "epoch": 0.9844462635949012,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8347,
      "step": 4209
    },
    {
      "epoch": 0.9846801543679102,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 4210
    },
    {
      "epoch": 0.9849140451409192,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 4211
    },
    {
      "epoch": 0.9851479359139282,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 4212
    },
    {
      "epoch": 0.9853818266869372,
      "grad_norm": 7.21875,
      "learning_rate": 3e-05,
      "loss": 2.3747,
      "step": 4213
    },
    {
      "epoch": 0.9856157174599463,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 4214
    },
    {
      "epoch": 0.9858496082329552,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 4215
    },
    {
      "epoch": 0.9860834990059643,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 4216
    },
    {
      "epoch": 0.9863173897789732,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 4217
    },
    {
      "epoch": 0.9865512805519823,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 4218
    },
    {
      "epoch": 0.9867851713249912,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 4219
    },
    {
      "epoch": 0.9870190620980003,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 4220
    },
    {
      "epoch": 0.9872529528710092,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 4221
    },
    {
      "epoch": 0.9874868436440183,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 4222
    },
    {
      "epoch": 0.9877207344170272,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 4223
    },
    {
      "epoch": 0.9879546251900363,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8634,
      "step": 4224
    },
    {
      "epoch": 0.9881885159630452,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1754,
      "step": 4225
    },
    {
      "epoch": 0.9884224067360543,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 4226
    },
    {
      "epoch": 0.9886562975090633,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 4227
    },
    {
      "epoch": 0.9888901882820723,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 4228
    },
    {
      "epoch": 0.9891240790550813,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 4229
    },
    {
      "epoch": 0.9893579698280903,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.662,
      "step": 4230
    },
    {
      "epoch": 0.9895918606010993,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 4231
    },
    {
      "epoch": 0.9898257513741083,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5052,
      "step": 4232
    },
    {
      "epoch": 0.9900596421471173,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 4233
    },
    {
      "epoch": 0.9902935329201263,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 4234
    },
    {
      "epoch": 0.9905274236931353,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5352,
      "step": 4235
    },
    {
      "epoch": 0.9907613144661443,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 4236
    },
    {
      "epoch": 0.9909952052391533,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 4237
    },
    {
      "epoch": 0.9912290960121624,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 4238
    },
    {
      "epoch": 0.9914629867851713,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 4239
    },
    {
      "epoch": 0.9916968775581804,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 4240
    },
    {
      "epoch": 0.9919307683311893,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 4241
    },
    {
      "epoch": 0.9921646591041984,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 4242
    },
    {
      "epoch": 0.9923985498772073,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 4243
    },
    {
      "epoch": 0.9926324406502164,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 4244
    },
    {
      "epoch": 0.9928663314232253,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 4245
    },
    {
      "epoch": 0.9931002221962344,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 4246
    },
    {
      "epoch": 0.9933341129692433,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 4247
    },
    {
      "epoch": 0.9935680037422524,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 4248
    },
    {
      "epoch": 0.9938018945152614,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 4249
    },
    {
      "epoch": 0.9940357852882704,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 4250
    },
    {
      "epoch": 0.9942696760612794,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 4251
    },
    {
      "epoch": 0.9945035668342884,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7224,
      "step": 4252
    },
    {
      "epoch": 0.9947374576072974,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 4253
    },
    {
      "epoch": 0.9949713483803064,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 4254
    },
    {
      "epoch": 0.9952052391533154,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 4255
    },
    {
      "epoch": 0.9954391299263244,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 4256
    },
    {
      "epoch": 0.9956730206993334,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 4257
    },
    {
      "epoch": 0.9959069114723424,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 4258
    },
    {
      "epoch": 0.9961408022453514,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 4259
    },
    {
      "epoch": 0.9963746930183605,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 4260
    },
    {
      "epoch": 0.9966085837913694,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 4261
    },
    {
      "epoch": 0.9968424745643785,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 4262
    },
    {
      "epoch": 0.9970763653373874,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 4263
    },
    {
      "epoch": 0.9973102561103965,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 4264
    },
    {
      "epoch": 0.9975441468834054,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 4265
    },
    {
      "epoch": 0.9977780376564145,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5606,
      "step": 4266
    },
    {
      "epoch": 0.9980119284294234,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3384,
      "step": 4267
    },
    {
      "epoch": 0.9982458192024325,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 4268
    },
    {
      "epoch": 0.9984797099754414,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 4269
    },
    {
      "epoch": 0.9987136007484505,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7332,
      "step": 4270
    },
    {
      "epoch": 0.9989474915214595,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 4271
    },
    {
      "epoch": 0.9991813822944685,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 4272
    },
    {
      "epoch": 0.9994152730674775,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 4273
    },
    {
      "epoch": 0.9996491638404865,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 4274
    },
    {
      "epoch": 0.9998830546134955,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 4275
    },
    {
      "epoch": 1.000233890773009,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 3.4164,
      "step": 4276
    },
    {
      "epoch": 1.0004677815460181,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 4277
    },
    {
      "epoch": 1.000701672319027,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 4278
    },
    {
      "epoch": 1.000935563092036,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.201,
      "step": 4279
    },
    {
      "epoch": 1.001169453865045,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2987,
      "step": 4280
    },
    {
      "epoch": 1.0014033446380541,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 4281
    },
    {
      "epoch": 1.001637235411063,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 4282
    },
    {
      "epoch": 1.001871126184072,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 4283
    },
    {
      "epoch": 1.002105016957081,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 4284
    },
    {
      "epoch": 1.0023389077300902,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 4285
    },
    {
      "epoch": 1.002572798503099,
      "grad_norm": 10.25,
      "learning_rate": 3e-05,
      "loss": 2.3295,
      "step": 4286
    },
    {
      "epoch": 1.002806689276108,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 4287
    },
    {
      "epoch": 1.0030405800491171,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 4288
    },
    {
      "epoch": 1.0032744708221262,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 4289
    },
    {
      "epoch": 1.003508361595135,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 4290
    },
    {
      "epoch": 1.003742252368144,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 4291
    },
    {
      "epoch": 1.0039761431411531,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 4292
    },
    {
      "epoch": 1.0042100339141622,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 4293
    },
    {
      "epoch": 1.004443924687171,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 4294
    },
    {
      "epoch": 1.00467781546018,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 4295
    },
    {
      "epoch": 1.0049117062331892,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 4296
    },
    {
      "epoch": 1.0051455970061982,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 4297
    },
    {
      "epoch": 1.005379487779207,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6313,
      "step": 4298
    },
    {
      "epoch": 1.005613378552216,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 4299
    },
    {
      "epoch": 1.0058472693252252,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 4300
    },
    {
      "epoch": 1.0058472693252252,
      "eval_runtime": 4.664,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 4300
    },
    {
      "epoch": 1.0060811600982342,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 4301
    },
    {
      "epoch": 1.006315050871243,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 4302
    },
    {
      "epoch": 1.0065489416442521,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6788,
      "step": 4303
    },
    {
      "epoch": 1.0067828324172612,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7339,
      "step": 4304
    },
    {
      "epoch": 1.0070167231902702,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 4305
    },
    {
      "epoch": 1.007250613963279,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 4306
    },
    {
      "epoch": 1.0074845047362881,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 4307
    },
    {
      "epoch": 1.0077183955092972,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 4308
    },
    {
      "epoch": 1.0079522862823063,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 4309
    },
    {
      "epoch": 1.008186177055315,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 4310
    },
    {
      "epoch": 1.0084200678283242,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 4311
    },
    {
      "epoch": 1.0086539586013332,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4135,
      "step": 4312
    },
    {
      "epoch": 1.0088878493743423,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8608,
      "step": 4313
    },
    {
      "epoch": 1.0091217401473511,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 4314
    },
    {
      "epoch": 1.0093556309203602,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 4315
    },
    {
      "epoch": 1.0095895216933692,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.5943,
      "step": 4316
    },
    {
      "epoch": 1.0098234124663783,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 4317
    },
    {
      "epoch": 1.0100573032393871,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 4318
    },
    {
      "epoch": 1.0102911940123962,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 4319
    },
    {
      "epoch": 1.0105250847854053,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8318,
      "step": 4320
    },
    {
      "epoch": 1.0107589755584143,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 4321
    },
    {
      "epoch": 1.0109928663314232,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.3671,
      "step": 4322
    },
    {
      "epoch": 1.0112267571044322,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 4323
    },
    {
      "epoch": 1.0114606478774413,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 4324
    },
    {
      "epoch": 1.0116945386504503,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.424,
      "step": 4325
    },
    {
      "epoch": 1.0119284294234592,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 4326
    },
    {
      "epoch": 1.0121623201964682,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 4327
    },
    {
      "epoch": 1.0123962109694773,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 4328
    },
    {
      "epoch": 1.0126301017424864,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 4329
    },
    {
      "epoch": 1.0128639925154952,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 4330
    },
    {
      "epoch": 1.0130978832885043,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6184,
      "step": 4331
    },
    {
      "epoch": 1.0133317740615133,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 4332
    },
    {
      "epoch": 1.0135656648345224,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 4333
    },
    {
      "epoch": 1.0137995556075312,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 4334
    },
    {
      "epoch": 1.0140334463805403,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 4335
    },
    {
      "epoch": 1.0142673371535493,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 4336
    },
    {
      "epoch": 1.0145012279265584,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 4337
    },
    {
      "epoch": 1.0147351186995672,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 4338
    },
    {
      "epoch": 1.0149690094725763,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7936,
      "step": 4339
    },
    {
      "epoch": 1.0152029002455853,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 4340
    },
    {
      "epoch": 1.0154367910185944,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6043,
      "step": 4341
    },
    {
      "epoch": 1.0156706817916032,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 4342
    },
    {
      "epoch": 1.0159045725646123,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 4343
    },
    {
      "epoch": 1.0161384633376214,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.7868,
      "step": 4344
    },
    {
      "epoch": 1.0163723541106304,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 4345
    },
    {
      "epoch": 1.0166062448836393,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 4346
    },
    {
      "epoch": 1.0168401356566483,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2468,
      "step": 4347
    },
    {
      "epoch": 1.0170740264296574,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 4348
    },
    {
      "epoch": 1.0173079172026664,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 4349
    },
    {
      "epoch": 1.0175418079756753,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 4350
    },
    {
      "epoch": 1.0177756987486843,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 4351
    },
    {
      "epoch": 1.0180095895216934,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 4352
    },
    {
      "epoch": 1.0182434802947025,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 4353
    },
    {
      "epoch": 1.0184773710677113,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 4354
    },
    {
      "epoch": 1.0187112618407204,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 4355
    },
    {
      "epoch": 1.0189451526137294,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 4356
    },
    {
      "epoch": 1.0191790433867385,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 4357
    },
    {
      "epoch": 1.0194129341597473,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 4358
    },
    {
      "epoch": 1.0196468249327564,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7039,
      "step": 4359
    },
    {
      "epoch": 1.0198807157057654,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 4360
    },
    {
      "epoch": 1.0201146064787745,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 4361
    },
    {
      "epoch": 1.0203484972517833,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 4362
    },
    {
      "epoch": 1.0205823880247924,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 4363
    },
    {
      "epoch": 1.0208162787978015,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 4364
    },
    {
      "epoch": 1.0210501695708105,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 4365
    },
    {
      "epoch": 1.0212840603438194,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 4366
    },
    {
      "epoch": 1.0215179511168284,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 4367
    },
    {
      "epoch": 1.0217518418898375,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 4368
    },
    {
      "epoch": 1.0219857326628465,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 4369
    },
    {
      "epoch": 1.0222196234358554,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 4370
    },
    {
      "epoch": 1.0224535142088644,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0676,
      "step": 4371
    },
    {
      "epoch": 1.0226874049818735,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 4372
    },
    {
      "epoch": 1.0229212957548826,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 4373
    },
    {
      "epoch": 1.0231551865278914,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 4374
    },
    {
      "epoch": 1.0233890773009005,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 4375
    },
    {
      "epoch": 1.0236229680739095,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 4376
    },
    {
      "epoch": 1.0238568588469186,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 4377
    },
    {
      "epoch": 1.0240907496199274,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 4378
    },
    {
      "epoch": 1.0243246403929365,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7467,
      "step": 4379
    },
    {
      "epoch": 1.0245585311659455,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 4380
    },
    {
      "epoch": 1.0247924219389546,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 4381
    },
    {
      "epoch": 1.0250263127119634,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 4382
    },
    {
      "epoch": 1.0252602034849725,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 4383
    },
    {
      "epoch": 1.0254940942579815,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6773,
      "step": 4384
    },
    {
      "epoch": 1.0257279850309906,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 4385
    },
    {
      "epoch": 1.0259618758039994,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 4386
    },
    {
      "epoch": 1.0261957665770085,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 4387
    },
    {
      "epoch": 1.0264296573500176,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 4388
    },
    {
      "epoch": 1.0266635481230266,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 4389
    },
    {
      "epoch": 1.0268974388960355,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 4390
    },
    {
      "epoch": 1.0271313296690445,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4027,
      "step": 4391
    },
    {
      "epoch": 1.0273652204420536,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 4392
    },
    {
      "epoch": 1.0275991112150626,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 4393
    },
    {
      "epoch": 1.0278330019880715,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 4394
    },
    {
      "epoch": 1.0280668927610805,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 4395
    },
    {
      "epoch": 1.0283007835340896,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 4396
    },
    {
      "epoch": 1.0285346743070987,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7872,
      "step": 4397
    },
    {
      "epoch": 1.0287685650801075,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 4398
    },
    {
      "epoch": 1.0290024558531166,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 4399
    },
    {
      "epoch": 1.0292363466261256,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 4400
    },
    {
      "epoch": 1.0292363466261256,
      "eval_runtime": 4.6608,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 4400
    },
    {
      "epoch": 1.0294702373991347,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8607,
      "step": 4401
    },
    {
      "epoch": 1.0297041281721435,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 4402
    },
    {
      "epoch": 1.0299380189451526,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 4403
    },
    {
      "epoch": 1.0301719097181616,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1656,
      "step": 4404
    },
    {
      "epoch": 1.0304058004911707,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7284,
      "step": 4405
    },
    {
      "epoch": 1.0306396912641795,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.885,
      "step": 4406
    },
    {
      "epoch": 1.0308735820371886,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 4407
    },
    {
      "epoch": 1.0311074728101977,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 4408
    },
    {
      "epoch": 1.0313413635832067,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 4409
    },
    {
      "epoch": 1.0315752543562156,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0492,
      "step": 4410
    },
    {
      "epoch": 1.0318091451292246,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 4411
    },
    {
      "epoch": 1.0320430359022337,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 4412
    },
    {
      "epoch": 1.0322769266752427,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 4413
    },
    {
      "epoch": 1.0325108174482516,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 4414
    },
    {
      "epoch": 1.0327447082212606,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 4415
    },
    {
      "epoch": 1.0329785989942697,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 4416
    },
    {
      "epoch": 1.0332124897672788,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 4417
    },
    {
      "epoch": 1.0334463805402876,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 4418
    },
    {
      "epoch": 1.0336802713132967,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 4419
    },
    {
      "epoch": 1.0339141620863057,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 4420
    },
    {
      "epoch": 1.0341480528593148,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7013,
      "step": 4421
    },
    {
      "epoch": 1.0343819436323236,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 4422
    },
    {
      "epoch": 1.0346158344053327,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 4423
    },
    {
      "epoch": 1.0348497251783417,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3272,
      "step": 4424
    },
    {
      "epoch": 1.0350836159513508,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 4425
    },
    {
      "epoch": 1.0353175067243596,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 4426
    },
    {
      "epoch": 1.0355513974973687,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.6176,
      "step": 4427
    },
    {
      "epoch": 1.0357852882703777,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 4428
    },
    {
      "epoch": 1.0360191790433868,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 4429
    },
    {
      "epoch": 1.0362530698163956,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 4430
    },
    {
      "epoch": 1.0364869605894047,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 4431
    },
    {
      "epoch": 1.0367208513624138,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 4432
    },
    {
      "epoch": 1.0369547421354228,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 4433
    },
    {
      "epoch": 1.0371886329084317,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 4434
    },
    {
      "epoch": 1.0374225236814407,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 4435
    },
    {
      "epoch": 1.0376564144544498,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 4436
    },
    {
      "epoch": 1.0378903052274588,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 4437
    },
    {
      "epoch": 1.0381241960004677,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 4438
    },
    {
      "epoch": 1.0383580867734767,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 4439
    },
    {
      "epoch": 1.0385919775464858,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 4440
    },
    {
      "epoch": 1.0388258683194949,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 4441
    },
    {
      "epoch": 1.0390597590925037,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.656,
      "step": 4442
    },
    {
      "epoch": 1.0392936498655128,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 4443
    },
    {
      "epoch": 1.0395275406385218,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 4444
    },
    {
      "epoch": 1.0397614314115309,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 4445
    },
    {
      "epoch": 1.0399953221845397,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 4446
    },
    {
      "epoch": 1.0402292129575488,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 4447
    },
    {
      "epoch": 1.0404631037305578,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 4448
    },
    {
      "epoch": 1.040696994503567,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 4449
    },
    {
      "epoch": 1.0409308852765757,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 4450
    },
    {
      "epoch": 1.0411647760495848,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 4451
    },
    {
      "epoch": 1.0413986668225939,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.211,
      "step": 4452
    },
    {
      "epoch": 1.041632557595603,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9751,
      "step": 4453
    },
    {
      "epoch": 1.0418664483686118,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 4454
    },
    {
      "epoch": 1.0421003391416208,
      "grad_norm": 2.515625,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 4455
    },
    {
      "epoch": 1.0423342299146299,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 4456
    },
    {
      "epoch": 1.042568120687639,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 4457
    },
    {
      "epoch": 1.0428020114606478,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 4458
    },
    {
      "epoch": 1.0430359022336568,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 4459
    },
    {
      "epoch": 1.043269793006666,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 4460
    },
    {
      "epoch": 1.043503683779675,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7656,
      "step": 4461
    },
    {
      "epoch": 1.0437375745526838,
      "grad_norm": 8.3125,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 4462
    },
    {
      "epoch": 1.0439714653256928,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 4463
    },
    {
      "epoch": 1.044205356098702,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 4464
    },
    {
      "epoch": 1.044439246871711,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 4465
    },
    {
      "epoch": 1.04467313764472,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 4466
    },
    {
      "epoch": 1.0449070284177289,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 4467
    },
    {
      "epoch": 1.045140919190738,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 4468
    },
    {
      "epoch": 1.045374809963747,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8385,
      "step": 4469
    },
    {
      "epoch": 1.0456087007367558,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 4470
    },
    {
      "epoch": 1.0458425915097649,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 4471
    },
    {
      "epoch": 1.046076482282774,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 4472
    },
    {
      "epoch": 1.046310373055783,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 4473
    },
    {
      "epoch": 1.046544263828792,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 4474
    },
    {
      "epoch": 1.046778154601801,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 4475
    },
    {
      "epoch": 1.04701204537481,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 4476
    },
    {
      "epoch": 1.047245936147819,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 4477
    },
    {
      "epoch": 1.0474798269208279,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.4497,
      "step": 4478
    },
    {
      "epoch": 1.047713717693837,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5467,
      "step": 4479
    },
    {
      "epoch": 1.047947608466846,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 4480
    },
    {
      "epoch": 1.048181499239855,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 4481
    },
    {
      "epoch": 1.048415390012864,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 4482
    },
    {
      "epoch": 1.048649280785873,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6729,
      "step": 4483
    },
    {
      "epoch": 1.048883171558882,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 4484
    },
    {
      "epoch": 1.049117062331891,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 4485
    },
    {
      "epoch": 1.0493509531049,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 4486
    },
    {
      "epoch": 1.049584843877909,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6149,
      "step": 4487
    },
    {
      "epoch": 1.049818734650918,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 4488
    },
    {
      "epoch": 1.050052625423927,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5065,
      "step": 4489
    },
    {
      "epoch": 1.0502865161969361,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 4490
    },
    {
      "epoch": 1.050520406969945,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 4491
    },
    {
      "epoch": 1.050754297742954,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 4492
    },
    {
      "epoch": 1.050988188515963,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 4493
    },
    {
      "epoch": 1.0512220792889722,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.2294,
      "step": 4494
    },
    {
      "epoch": 1.051455970061981,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 4495
    },
    {
      "epoch": 1.05168986083499,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 4496
    },
    {
      "epoch": 1.0519237516079991,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 4497
    },
    {
      "epoch": 1.0521576423810082,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 4498
    },
    {
      "epoch": 1.052391533154017,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7267,
      "step": 4499
    },
    {
      "epoch": 1.052625423927026,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 4500
    },
    {
      "epoch": 1.052625423927026,
      "eval_runtime": 4.6048,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 4500
    },
    {
      "epoch": 1.0528593147000351,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6319,
      "step": 4501
    },
    {
      "epoch": 1.0530932054730442,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 4502
    },
    {
      "epoch": 1.053327096246053,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7489,
      "step": 4503
    },
    {
      "epoch": 1.053560987019062,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 4504
    },
    {
      "epoch": 1.0537948777920711,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 4505
    },
    {
      "epoch": 1.0540287685650802,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 4506
    },
    {
      "epoch": 1.054262659338089,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 4507
    },
    {
      "epoch": 1.054496550111098,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 4508
    },
    {
      "epoch": 1.0547304408841072,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 4509
    },
    {
      "epoch": 1.0549643316571162,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 4510
    },
    {
      "epoch": 1.055198222430125,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 4511
    },
    {
      "epoch": 1.0554321132031341,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 4512
    },
    {
      "epoch": 1.0556660039761432,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 4513
    },
    {
      "epoch": 1.0558998947491522,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 4514
    },
    {
      "epoch": 1.056133785522161,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 4515
    },
    {
      "epoch": 1.0563676762951701,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 4516
    },
    {
      "epoch": 1.0566015670681792,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 4517
    },
    {
      "epoch": 1.0568354578411883,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 4518
    },
    {
      "epoch": 1.057069348614197,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 4519
    },
    {
      "epoch": 1.0573032393872062,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 4520
    },
    {
      "epoch": 1.0575371301602152,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 4521
    },
    {
      "epoch": 1.0577710209332243,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 4522
    },
    {
      "epoch": 1.0580049117062331,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 4523
    },
    {
      "epoch": 1.0582388024792422,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 4524
    },
    {
      "epoch": 1.0584726932522512,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 4525
    },
    {
      "epoch": 1.0587065840252603,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9733,
      "step": 4526
    },
    {
      "epoch": 1.0589404747982691,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 4527
    },
    {
      "epoch": 1.0591743655712782,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1341,
      "step": 4528
    },
    {
      "epoch": 1.0594082563442873,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 4529
    },
    {
      "epoch": 1.0596421471172963,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 4530
    },
    {
      "epoch": 1.0598760378903052,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 4531
    },
    {
      "epoch": 1.0601099286633142,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 4532
    },
    {
      "epoch": 1.0603438194363233,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 4533
    },
    {
      "epoch": 1.0605777102093323,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 4534
    },
    {
      "epoch": 1.0608116009823412,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 4535
    },
    {
      "epoch": 1.0610454917553502,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2552,
      "step": 4536
    },
    {
      "epoch": 1.0612793825283593,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 4537
    },
    {
      "epoch": 1.0615132733013684,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 4538
    },
    {
      "epoch": 1.0617471640743772,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 4539
    },
    {
      "epoch": 1.0619810548473863,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 4540
    },
    {
      "epoch": 1.0622149456203953,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 4541
    },
    {
      "epoch": 1.0624488363934044,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 4542
    },
    {
      "epoch": 1.0626827271664132,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 4543
    },
    {
      "epoch": 1.0629166179394223,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 4544
    },
    {
      "epoch": 1.0631505087124313,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 4545
    },
    {
      "epoch": 1.0633843994854404,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 4546
    },
    {
      "epoch": 1.0636182902584492,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 4547
    },
    {
      "epoch": 1.0638521810314583,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 4548
    },
    {
      "epoch": 1.0640860718044673,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 4549
    },
    {
      "epoch": 1.0643199625774764,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 4550
    },
    {
      "epoch": 1.0645538533504852,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 4551
    },
    {
      "epoch": 1.0647877441234943,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 4552
    },
    {
      "epoch": 1.0650216348965034,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 4553
    },
    {
      "epoch": 1.0652555256695124,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 4554
    },
    {
      "epoch": 1.0654894164425213,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.8734,
      "step": 4555
    },
    {
      "epoch": 1.0657233072155303,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 4556
    },
    {
      "epoch": 1.0659571979885394,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 4557
    },
    {
      "epoch": 1.0661910887615484,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7814,
      "step": 4558
    },
    {
      "epoch": 1.0664249795345573,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 4559
    },
    {
      "epoch": 1.0666588703075663,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 4560
    },
    {
      "epoch": 1.0668927610805754,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 4561
    },
    {
      "epoch": 1.0671266518535845,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 4562
    },
    {
      "epoch": 1.0673605426265933,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 4563
    },
    {
      "epoch": 1.0675944333996024,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9381,
      "step": 4564
    },
    {
      "epoch": 1.0678283241726114,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 4565
    },
    {
      "epoch": 1.0680622149456205,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 4566
    },
    {
      "epoch": 1.0682961057186293,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 4567
    },
    {
      "epoch": 1.0685299964916384,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 4568
    },
    {
      "epoch": 1.0687638872646474,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 4569
    },
    {
      "epoch": 1.0689977780376565,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 4570
    },
    {
      "epoch": 1.0692316688106653,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 4571
    },
    {
      "epoch": 1.0694655595836744,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 4572
    },
    {
      "epoch": 1.0696994503566835,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 4573
    },
    {
      "epoch": 1.0699333411296925,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 4574
    },
    {
      "epoch": 1.0701672319027014,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 4575
    },
    {
      "epoch": 1.0704011226757104,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6874,
      "step": 4576
    },
    {
      "epoch": 1.0706350134487195,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6077,
      "step": 4577
    },
    {
      "epoch": 1.0708689042217285,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 4578
    },
    {
      "epoch": 1.0711027949947374,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 4579
    },
    {
      "epoch": 1.0713366857677464,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 4580
    },
    {
      "epoch": 1.0715705765407555,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 4581
    },
    {
      "epoch": 1.0718044673137646,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 4582
    },
    {
      "epoch": 1.0720383580867734,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 4583
    },
    {
      "epoch": 1.0722722488597825,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 4584
    },
    {
      "epoch": 1.0725061396327915,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 4585
    },
    {
      "epoch": 1.0727400304058006,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 4586
    },
    {
      "epoch": 1.0729739211788094,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 4587
    },
    {
      "epoch": 1.0732078119518185,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 4588
    },
    {
      "epoch": 1.0734417027248275,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 4589
    },
    {
      "epoch": 1.0736755934978366,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 4590
    },
    {
      "epoch": 1.0739094842708454,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 4591
    },
    {
      "epoch": 1.0741433750438545,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 4592
    },
    {
      "epoch": 1.0743772658168635,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 4593
    },
    {
      "epoch": 1.0746111565898726,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 4594
    },
    {
      "epoch": 1.0748450473628814,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 4595
    },
    {
      "epoch": 1.0750789381358905,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 4596
    },
    {
      "epoch": 1.0753128289088996,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 4597
    },
    {
      "epoch": 1.0755467196819086,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 4598
    },
    {
      "epoch": 1.0757806104549175,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8156,
      "step": 4599
    },
    {
      "epoch": 1.0760145012279265,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 4600
    },
    {
      "epoch": 1.0760145012279265,
      "eval_runtime": 4.6164,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 4600
    },
    {
      "epoch": 1.0762483920009356,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5765,
      "step": 4601
    },
    {
      "epoch": 1.0764822827739446,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 4602
    },
    {
      "epoch": 1.0767161735469535,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 4603
    },
    {
      "epoch": 1.0769500643199625,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 4604
    },
    {
      "epoch": 1.0771839550929716,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7266,
      "step": 4605
    },
    {
      "epoch": 1.0774178458659807,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 4606
    },
    {
      "epoch": 1.0776517366389895,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 4607
    },
    {
      "epoch": 1.0778856274119986,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 4608
    },
    {
      "epoch": 1.0781195181850076,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 4609
    },
    {
      "epoch": 1.0783534089580167,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 4610
    },
    {
      "epoch": 1.0785872997310255,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 4611
    },
    {
      "epoch": 1.0788211905040346,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7604,
      "step": 4612
    },
    {
      "epoch": 1.0790550812770436,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 4613
    },
    {
      "epoch": 1.0792889720500527,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6026,
      "step": 4614
    },
    {
      "epoch": 1.0795228628230615,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 4615
    },
    {
      "epoch": 1.0797567535960706,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 4616
    },
    {
      "epoch": 1.0799906443690797,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 4617
    },
    {
      "epoch": 1.0802245351420887,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 4618
    },
    {
      "epoch": 1.0804584259150976,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0437,
      "step": 4619
    },
    {
      "epoch": 1.0806923166881066,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7242,
      "step": 4620
    },
    {
      "epoch": 1.0809262074611157,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 4621
    },
    {
      "epoch": 1.0811600982341247,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 4622
    },
    {
      "epoch": 1.0813939890071336,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 4623
    },
    {
      "epoch": 1.0816278797801426,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8676,
      "step": 4624
    },
    {
      "epoch": 1.0818617705531517,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 4625
    },
    {
      "epoch": 1.0820956613261608,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8707,
      "step": 4626
    },
    {
      "epoch": 1.0823295520991696,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 4627
    },
    {
      "epoch": 1.0825634428721786,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 4628
    },
    {
      "epoch": 1.0827973336451877,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.775,
      "step": 4629
    },
    {
      "epoch": 1.0830312244181968,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 4630
    },
    {
      "epoch": 1.0832651151912056,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 4631
    },
    {
      "epoch": 1.0834990059642147,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 4632
    },
    {
      "epoch": 1.0837328967372237,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 4633
    },
    {
      "epoch": 1.0839667875102328,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.444,
      "step": 4634
    },
    {
      "epoch": 1.0842006782832416,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 4635
    },
    {
      "epoch": 1.0844345690562507,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 4636
    },
    {
      "epoch": 1.0846684598292597,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 4637
    },
    {
      "epoch": 1.0849023506022688,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 4638
    },
    {
      "epoch": 1.0851362413752776,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2738,
      "step": 4639
    },
    {
      "epoch": 1.0853701321482867,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6453,
      "step": 4640
    },
    {
      "epoch": 1.0856040229212958,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4506,
      "step": 4641
    },
    {
      "epoch": 1.0858379136943048,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 4642
    },
    {
      "epoch": 1.0860718044673137,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 4643
    },
    {
      "epoch": 1.0863056952403227,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 4644
    },
    {
      "epoch": 1.0865395860133318,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 4645
    },
    {
      "epoch": 1.0867734767863408,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8451,
      "step": 4646
    },
    {
      "epoch": 1.0870073675593497,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.827,
      "step": 4647
    },
    {
      "epoch": 1.0872412583323587,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 4648
    },
    {
      "epoch": 1.0874751491053678,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 4649
    },
    {
      "epoch": 1.0877090398783769,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 4650
    },
    {
      "epoch": 1.0879429306513857,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 4651
    },
    {
      "epoch": 1.0881768214243948,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5878,
      "step": 4652
    },
    {
      "epoch": 1.0884107121974038,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 4653
    },
    {
      "epoch": 1.0886446029704129,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 4654
    },
    {
      "epoch": 1.088878493743422,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 4655
    },
    {
      "epoch": 1.0891123845164308,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 4656
    },
    {
      "epoch": 1.0893462752894398,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 4657
    },
    {
      "epoch": 1.089580166062449,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 4658
    },
    {
      "epoch": 1.0898140568354577,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 4659
    },
    {
      "epoch": 1.0900479476084668,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 4660
    },
    {
      "epoch": 1.0902818383814759,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7432,
      "step": 4661
    },
    {
      "epoch": 1.090515729154485,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.716,
      "step": 4662
    },
    {
      "epoch": 1.090749619927494,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 4663
    },
    {
      "epoch": 1.0909835107005028,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 4664
    },
    {
      "epoch": 1.0912174014735119,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 4665
    },
    {
      "epoch": 1.091451292246521,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 4666
    },
    {
      "epoch": 1.0916851830195298,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 4667
    },
    {
      "epoch": 1.0919190737925388,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 4668
    },
    {
      "epoch": 1.092152964565548,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6746,
      "step": 4669
    },
    {
      "epoch": 1.092386855338557,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 4670
    },
    {
      "epoch": 1.092620746111566,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6244,
      "step": 4671
    },
    {
      "epoch": 1.0928546368845748,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 4672
    },
    {
      "epoch": 1.093088527657584,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 4673
    },
    {
      "epoch": 1.093322418430593,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 4674
    },
    {
      "epoch": 1.0935563092036018,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 4675
    },
    {
      "epoch": 1.0937901999766109,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 4676
    },
    {
      "epoch": 1.09402409074962,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 4677
    },
    {
      "epoch": 1.094257981522629,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 4678
    },
    {
      "epoch": 1.094491872295638,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 4679
    },
    {
      "epoch": 1.0947257630686469,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 4680
    },
    {
      "epoch": 1.094959653841656,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 4681
    },
    {
      "epoch": 1.095193544614665,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2465,
      "step": 4682
    },
    {
      "epoch": 1.0954274353876738,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 4683
    },
    {
      "epoch": 1.095661326160683,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8257,
      "step": 4684
    },
    {
      "epoch": 1.095895216933692,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 4685
    },
    {
      "epoch": 1.096129107706701,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 4686
    },
    {
      "epoch": 1.09636299847971,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 4687
    },
    {
      "epoch": 1.096596889252719,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 4688
    },
    {
      "epoch": 1.096830780025728,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 4689
    },
    {
      "epoch": 1.097064670798737,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 4690
    },
    {
      "epoch": 1.0972985615717459,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 4691
    },
    {
      "epoch": 1.097532452344755,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5043,
      "step": 4692
    },
    {
      "epoch": 1.097766343117764,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 4693
    },
    {
      "epoch": 1.098000233890773,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 4694
    },
    {
      "epoch": 1.0982341246637821,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 4695
    },
    {
      "epoch": 1.098468015436791,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 4696
    },
    {
      "epoch": 1.0987019062098,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8409,
      "step": 4697
    },
    {
      "epoch": 1.098935796982809,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 4698
    },
    {
      "epoch": 1.0991696877558181,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 4699
    },
    {
      "epoch": 1.099403578528827,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 4700
    },
    {
      "epoch": 1.099403578528827,
      "eval_runtime": 4.6326,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 4700
    },
    {
      "epoch": 1.099637469301836,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 4701
    },
    {
      "epoch": 1.099871360074845,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 4702
    },
    {
      "epoch": 1.1001052508478542,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 4703
    },
    {
      "epoch": 1.100339141620863,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 4704
    },
    {
      "epoch": 1.100573032393872,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 4705
    },
    {
      "epoch": 1.1008069231668811,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 4706
    },
    {
      "epoch": 1.1010408139398902,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 4707
    },
    {
      "epoch": 1.101274704712899,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 4708
    },
    {
      "epoch": 1.101508595485908,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.689,
      "step": 4709
    },
    {
      "epoch": 1.1017424862589171,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 4710
    },
    {
      "epoch": 1.1019763770319262,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 4711
    },
    {
      "epoch": 1.102210267804935,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 4712
    },
    {
      "epoch": 1.102444158577944,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 4713
    },
    {
      "epoch": 1.1026780493509531,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 4714
    },
    {
      "epoch": 1.1029119401239622,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 4715
    },
    {
      "epoch": 1.103145830896971,
      "grad_norm": 7.9375,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 4716
    },
    {
      "epoch": 1.10337972166998,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 4717
    },
    {
      "epoch": 1.1036136124429892,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 4718
    },
    {
      "epoch": 1.1038475032159982,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8462,
      "step": 4719
    },
    {
      "epoch": 1.104081393989007,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 4720
    },
    {
      "epoch": 1.1043152847620161,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 4721
    },
    {
      "epoch": 1.1045491755350252,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 4722
    },
    {
      "epoch": 1.1047830663080342,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5737,
      "step": 4723
    },
    {
      "epoch": 1.105016957081043,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 4724
    },
    {
      "epoch": 1.1052508478540521,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8087,
      "step": 4725
    },
    {
      "epoch": 1.1054847386270612,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 4726
    },
    {
      "epoch": 1.1057186294000703,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 4727
    },
    {
      "epoch": 1.105952520173079,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 4728
    },
    {
      "epoch": 1.1061864109460882,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2506,
      "step": 4729
    },
    {
      "epoch": 1.1064203017190972,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 4730
    },
    {
      "epoch": 1.1066541924921063,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 4731
    },
    {
      "epoch": 1.1068880832651151,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 4732
    },
    {
      "epoch": 1.1071219740381242,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 4733
    },
    {
      "epoch": 1.1073558648111332,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 4734
    },
    {
      "epoch": 1.1075897555841423,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 4735
    },
    {
      "epoch": 1.1078236463571511,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 4736
    },
    {
      "epoch": 1.1080575371301602,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 4737
    },
    {
      "epoch": 1.1082914279031693,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 4738
    },
    {
      "epoch": 1.1085253186761783,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9638,
      "step": 4739
    },
    {
      "epoch": 1.1087592094491872,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 4740
    },
    {
      "epoch": 1.1089931002221962,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 4741
    },
    {
      "epoch": 1.1092269909952053,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 4742
    },
    {
      "epoch": 1.1094608817682143,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 4743
    },
    {
      "epoch": 1.1096947725412232,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5439,
      "step": 4744
    },
    {
      "epoch": 1.1099286633142322,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 4745
    },
    {
      "epoch": 1.1101625540872413,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 4746
    },
    {
      "epoch": 1.1103964448602504,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6805,
      "step": 4747
    },
    {
      "epoch": 1.1106303356332592,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 4748
    },
    {
      "epoch": 1.1108642264062683,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 4749
    },
    {
      "epoch": 1.1110981171792773,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 4750
    },
    {
      "epoch": 1.1113320079522864,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6236,
      "step": 4751
    },
    {
      "epoch": 1.1115658987252952,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 4752
    },
    {
      "epoch": 1.1117997894983043,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 4753
    },
    {
      "epoch": 1.1120336802713133,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1963,
      "step": 4754
    },
    {
      "epoch": 1.1122675710443224,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 4755
    },
    {
      "epoch": 1.1125014618173312,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 4756
    },
    {
      "epoch": 1.1127353525903403,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 4757
    },
    {
      "epoch": 1.1129692433633493,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 4758
    },
    {
      "epoch": 1.1132031341363584,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 4759
    },
    {
      "epoch": 1.1134370249093672,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.3608,
      "step": 4760
    },
    {
      "epoch": 1.1136709156823763,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 4761
    },
    {
      "epoch": 1.1139048064553854,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 4762
    },
    {
      "epoch": 1.1141386972283944,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 4763
    },
    {
      "epoch": 1.1143725880014033,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 4764
    },
    {
      "epoch": 1.1146064787744123,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8524,
      "step": 4765
    },
    {
      "epoch": 1.1148403695474214,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 4766
    },
    {
      "epoch": 1.1150742603204304,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 4767
    },
    {
      "epoch": 1.1153081510934393,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 4768
    },
    {
      "epoch": 1.1155420418664483,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5579,
      "step": 4769
    },
    {
      "epoch": 1.1157759326394574,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 4770
    },
    {
      "epoch": 1.1160098234124665,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 4771
    },
    {
      "epoch": 1.1162437141854753,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 4772
    },
    {
      "epoch": 1.1164776049584844,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3543,
      "step": 4773
    },
    {
      "epoch": 1.1167114957314934,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 4774
    },
    {
      "epoch": 1.1169453865045025,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 4775
    },
    {
      "epoch": 1.1171792772775113,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7997,
      "step": 4776
    },
    {
      "epoch": 1.1174131680505204,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 4777
    },
    {
      "epoch": 1.1176470588235294,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 4778
    },
    {
      "epoch": 1.1178809495965385,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 4779
    },
    {
      "epoch": 1.1181148403695473,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 4780
    },
    {
      "epoch": 1.1183487311425564,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 4781
    },
    {
      "epoch": 1.1185826219155655,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 4782
    },
    {
      "epoch": 1.1188165126885745,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 4783
    },
    {
      "epoch": 1.1190504034615834,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 4784
    },
    {
      "epoch": 1.1192842942345924,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 4785
    },
    {
      "epoch": 1.1195181850076015,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 4786
    },
    {
      "epoch": 1.1197520757806105,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 4787
    },
    {
      "epoch": 1.1199859665536194,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7932,
      "step": 4788
    },
    {
      "epoch": 1.1202198573266284,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 4789
    },
    {
      "epoch": 1.1204537480996375,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6748,
      "step": 4790
    },
    {
      "epoch": 1.1206876388726466,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7733,
      "step": 4791
    },
    {
      "epoch": 1.1209215296456554,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.434,
      "step": 4792
    },
    {
      "epoch": 1.1211554204186645,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 4793
    },
    {
      "epoch": 1.1213893111916735,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 4794
    },
    {
      "epoch": 1.1216232019646826,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6288,
      "step": 4795
    },
    {
      "epoch": 1.1218570927376914,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 4796
    },
    {
      "epoch": 1.1220909835107005,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 4797
    },
    {
      "epoch": 1.1223248742837095,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7845,
      "step": 4798
    },
    {
      "epoch": 1.1225587650567186,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 4799
    },
    {
      "epoch": 1.1227926558297274,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7595,
      "step": 4800
    },
    {
      "epoch": 1.1227926558297274,
      "eval_runtime": 4.604,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 4800
    },
    {
      "epoch": 1.1230265466027365,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 4801
    },
    {
      "epoch": 1.1232604373757455,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 4802
    },
    {
      "epoch": 1.1234943281487546,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8174,
      "step": 4803
    },
    {
      "epoch": 1.1237282189217634,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 4804
    },
    {
      "epoch": 1.1239621096947725,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 4805
    },
    {
      "epoch": 1.1241960004677816,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 4806
    },
    {
      "epoch": 1.1244298912407906,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 4807
    },
    {
      "epoch": 1.1246637820137995,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 4808
    },
    {
      "epoch": 1.1248976727868085,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 4809
    },
    {
      "epoch": 1.1251315635598176,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 4810
    },
    {
      "epoch": 1.1253654543328266,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 4811
    },
    {
      "epoch": 1.1255993451058355,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 4812
    },
    {
      "epoch": 1.1258332358788445,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 4813
    },
    {
      "epoch": 1.1260671266518536,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6971,
      "step": 4814
    },
    {
      "epoch": 1.1263010174248627,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6776,
      "step": 4815
    },
    {
      "epoch": 1.1265349081978715,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7258,
      "step": 4816
    },
    {
      "epoch": 1.1267687989708806,
      "grad_norm": 14.0,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 4817
    },
    {
      "epoch": 1.1270026897438896,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6532,
      "step": 4818
    },
    {
      "epoch": 1.1272365805168987,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 4819
    },
    {
      "epoch": 1.1274704712899075,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 4820
    },
    {
      "epoch": 1.1277043620629166,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 4821
    },
    {
      "epoch": 1.1279382528359256,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 4822
    },
    {
      "epoch": 1.1281721436089347,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3593,
      "step": 4823
    },
    {
      "epoch": 1.1284060343819435,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 4824
    },
    {
      "epoch": 1.1286399251549526,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 4825
    },
    {
      "epoch": 1.1288738159279617,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 4826
    },
    {
      "epoch": 1.1291077067009707,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8417,
      "step": 4827
    },
    {
      "epoch": 1.1293415974739798,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 4828
    },
    {
      "epoch": 1.1295754882469886,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2952,
      "step": 4829
    },
    {
      "epoch": 1.1298093790199977,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 4830
    },
    {
      "epoch": 1.1300432697930067,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 4831
    },
    {
      "epoch": 1.1302771605660156,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7851,
      "step": 4832
    },
    {
      "epoch": 1.1305110513390246,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 4833
    },
    {
      "epoch": 1.1307449421120337,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 4834
    },
    {
      "epoch": 1.1309788328850428,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 4835
    },
    {
      "epoch": 1.1312127236580518,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 4836
    },
    {
      "epoch": 1.1314466144310606,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 4837
    },
    {
      "epoch": 1.1316805052040697,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 4838
    },
    {
      "epoch": 1.1319143959770788,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 4839
    },
    {
      "epoch": 1.1321482867500876,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 4840
    },
    {
      "epoch": 1.1323821775230967,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 4841
    },
    {
      "epoch": 1.1326160682961057,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 4842
    },
    {
      "epoch": 1.1328499590691148,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 4843
    },
    {
      "epoch": 1.1330838498421238,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 4844
    },
    {
      "epoch": 1.1333177406151327,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 4845
    },
    {
      "epoch": 1.1335516313881417,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 4846
    },
    {
      "epoch": 1.1337855221611508,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 4847
    },
    {
      "epoch": 1.1340194129341596,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 4848
    },
    {
      "epoch": 1.1342533037071687,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 4849
    },
    {
      "epoch": 1.1344871944801778,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 4850
    },
    {
      "epoch": 1.1347210852531868,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 4851
    },
    {
      "epoch": 1.1349549760261959,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 4852
    },
    {
      "epoch": 1.1351888667992047,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.603,
      "step": 4853
    },
    {
      "epoch": 1.1354227575722138,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 4854
    },
    {
      "epoch": 1.1356566483452228,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 4855
    },
    {
      "epoch": 1.1358905391182317,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 4856
    },
    {
      "epoch": 1.1361244298912407,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6752,
      "step": 4857
    },
    {
      "epoch": 1.1363583206642498,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 4858
    },
    {
      "epoch": 1.1365922114372589,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 4859
    },
    {
      "epoch": 1.136826102210268,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 4860
    },
    {
      "epoch": 1.1370599929832768,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 4861
    },
    {
      "epoch": 1.1372938837562858,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 4862
    },
    {
      "epoch": 1.1375277745292949,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 4863
    },
    {
      "epoch": 1.1377616653023037,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 4864
    },
    {
      "epoch": 1.1379955560753128,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 4865
    },
    {
      "epoch": 1.1382294468483218,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 4866
    },
    {
      "epoch": 1.138463337621331,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 4867
    },
    {
      "epoch": 1.13869722839434,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6678,
      "step": 4868
    },
    {
      "epoch": 1.1389311191673488,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 4869
    },
    {
      "epoch": 1.1391650099403579,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 4870
    },
    {
      "epoch": 1.139398900713367,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 4871
    },
    {
      "epoch": 1.1396327914863758,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7017,
      "step": 4872
    },
    {
      "epoch": 1.1398666822593848,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 4873
    },
    {
      "epoch": 1.1401005730323939,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 4874
    },
    {
      "epoch": 1.140334463805403,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6587,
      "step": 4875
    },
    {
      "epoch": 1.140568354578412,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4846,
      "step": 4876
    },
    {
      "epoch": 1.1408022453514208,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 4877
    },
    {
      "epoch": 1.14103613612443,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8178,
      "step": 4878
    },
    {
      "epoch": 1.141270026897439,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 4879
    },
    {
      "epoch": 1.1415039176704478,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 4880
    },
    {
      "epoch": 1.1417378084434568,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 4881
    },
    {
      "epoch": 1.141971699216466,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 4882
    },
    {
      "epoch": 1.142205589989475,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 4883
    },
    {
      "epoch": 1.142439480762484,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 4884
    },
    {
      "epoch": 1.1426733715354929,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.5123,
      "step": 4885
    },
    {
      "epoch": 1.142907262308502,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 4886
    },
    {
      "epoch": 1.143141153081511,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 4887
    },
    {
      "epoch": 1.1433750438545198,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 4888
    },
    {
      "epoch": 1.1436089346275289,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7646,
      "step": 4889
    },
    {
      "epoch": 1.143842825400538,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 4890
    },
    {
      "epoch": 1.144076716173547,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 4891
    },
    {
      "epoch": 1.144310606946556,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 4892
    },
    {
      "epoch": 1.144544497719565,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 4893
    },
    {
      "epoch": 1.144778388492574,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 4894
    },
    {
      "epoch": 1.145012279265583,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 4895
    },
    {
      "epoch": 1.1452461700385919,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 4896
    },
    {
      "epoch": 1.145480060811601,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 4897
    },
    {
      "epoch": 1.14571395158461,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 4898
    },
    {
      "epoch": 1.145947842357619,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 4899
    },
    {
      "epoch": 1.146181733130628,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.827,
      "step": 4900
    },
    {
      "epoch": 1.146181733130628,
      "eval_runtime": 4.6962,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 4900
    },
    {
      "epoch": 1.146415623903637,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 4901
    },
    {
      "epoch": 1.146649514676646,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 4902
    },
    {
      "epoch": 1.146883405449655,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 4903
    },
    {
      "epoch": 1.147117296222664,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 4904
    },
    {
      "epoch": 1.147351186995673,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 4905
    },
    {
      "epoch": 1.147585077768682,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 4906
    },
    {
      "epoch": 1.147818968541691,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.755,
      "step": 4907
    },
    {
      "epoch": 1.1480528593147001,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 4908
    },
    {
      "epoch": 1.148286750087709,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 4909
    },
    {
      "epoch": 1.148520640860718,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 4910
    },
    {
      "epoch": 1.148754531633727,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 4911
    },
    {
      "epoch": 1.148988422406736,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6627,
      "step": 4912
    },
    {
      "epoch": 1.149222313179745,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 4913
    },
    {
      "epoch": 1.149456203952754,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 4914
    },
    {
      "epoch": 1.1496900947257631,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 4915
    },
    {
      "epoch": 1.1499239854987722,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 4916
    },
    {
      "epoch": 1.150157876271781,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 4917
    },
    {
      "epoch": 1.15039176704479,
      "grad_norm": 8.75,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 4918
    },
    {
      "epoch": 1.1506256578177991,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 4919
    },
    {
      "epoch": 1.150859548590808,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7156,
      "step": 4920
    },
    {
      "epoch": 1.151093439363817,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 4921
    },
    {
      "epoch": 1.151327330136826,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 4922
    },
    {
      "epoch": 1.1515612209098351,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6221,
      "step": 4923
    },
    {
      "epoch": 1.1517951116828442,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 4924
    },
    {
      "epoch": 1.152029002455853,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 4925
    },
    {
      "epoch": 1.152262893228862,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 4926
    },
    {
      "epoch": 1.1524967840018712,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 4927
    },
    {
      "epoch": 1.1527306747748802,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8251,
      "step": 4928
    },
    {
      "epoch": 1.152964565547889,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 4929
    },
    {
      "epoch": 1.1531984563208981,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 4930
    },
    {
      "epoch": 1.1534323470939072,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 4931
    },
    {
      "epoch": 1.1536662378669162,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 4932
    },
    {
      "epoch": 1.153900128639925,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5444,
      "step": 4933
    },
    {
      "epoch": 1.1541340194129341,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 4934
    },
    {
      "epoch": 1.1543679101859432,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 4935
    },
    {
      "epoch": 1.1546018009589523,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 4936
    },
    {
      "epoch": 1.154835691731961,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 4937
    },
    {
      "epoch": 1.1550695825049702,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 4938
    },
    {
      "epoch": 1.1553034732779792,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 4939
    },
    {
      "epoch": 1.1555373640509883,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 4940
    },
    {
      "epoch": 1.1557712548239971,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 4941
    },
    {
      "epoch": 1.1560051455970062,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 4942
    },
    {
      "epoch": 1.1562390363700152,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 4943
    },
    {
      "epoch": 1.1564729271430243,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 4944
    },
    {
      "epoch": 1.1567068179160331,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 4945
    },
    {
      "epoch": 1.1569407086890422,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 2.2793,
      "step": 4946
    },
    {
      "epoch": 1.1571745994620513,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 4947
    },
    {
      "epoch": 1.1574084902350603,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 4948
    },
    {
      "epoch": 1.1576423810080692,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 4949
    },
    {
      "epoch": 1.1578762717810782,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 4950
    },
    {
      "epoch": 1.1581101625540873,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 4951
    },
    {
      "epoch": 1.1583440533270963,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.886,
      "step": 4952
    },
    {
      "epoch": 1.1585779441001052,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 2.396,
      "step": 4953
    },
    {
      "epoch": 1.1588118348731142,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 4954
    },
    {
      "epoch": 1.1590457256461233,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 4955
    },
    {
      "epoch": 1.1592796164191324,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9998,
      "step": 4956
    },
    {
      "epoch": 1.1595135071921412,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 4957
    },
    {
      "epoch": 1.1597473979651503,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 4958
    },
    {
      "epoch": 1.1599812887381593,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2048,
      "step": 4959
    },
    {
      "epoch": 1.1602151795111684,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 4960
    },
    {
      "epoch": 1.1604490702841772,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 4961
    },
    {
      "epoch": 1.1606829610571863,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6925,
      "step": 4962
    },
    {
      "epoch": 1.1609168518301953,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 4963
    },
    {
      "epoch": 1.1611507426032044,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 4964
    },
    {
      "epoch": 1.1613846333762132,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7203,
      "step": 4965
    },
    {
      "epoch": 1.1616185241492223,
      "grad_norm": 7.65625,
      "learning_rate": 3e-05,
      "loss": 2.3598,
      "step": 4966
    },
    {
      "epoch": 1.1618524149222313,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 4967
    },
    {
      "epoch": 1.1620863056952404,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 4968
    },
    {
      "epoch": 1.1623201964682492,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 4969
    },
    {
      "epoch": 1.1625540872412583,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 4970
    },
    {
      "epoch": 1.1627879780142674,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 4971
    },
    {
      "epoch": 1.1630218687872764,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8187,
      "step": 4972
    },
    {
      "epoch": 1.1632557595602853,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.588,
      "step": 4973
    },
    {
      "epoch": 1.1634896503332943,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 4974
    },
    {
      "epoch": 1.1637235411063034,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 4975
    },
    {
      "epoch": 1.1639574318793124,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 4976
    },
    {
      "epoch": 1.1641913226523213,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 4977
    },
    {
      "epoch": 1.1644252134253303,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 4978
    },
    {
      "epoch": 1.1646591041983394,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 4979
    },
    {
      "epoch": 1.1648929949713485,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6094,
      "step": 4980
    },
    {
      "epoch": 1.1651268857443573,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 4981
    },
    {
      "epoch": 1.1653607765173664,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 4982
    },
    {
      "epoch": 1.1655946672903754,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 4983
    },
    {
      "epoch": 1.1658285580633845,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 4984
    },
    {
      "epoch": 1.1660624488363933,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5617,
      "step": 4985
    },
    {
      "epoch": 1.1662963396094024,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 4986
    },
    {
      "epoch": 1.1665302303824114,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 4987
    },
    {
      "epoch": 1.1667641211554205,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 4988
    },
    {
      "epoch": 1.1669980119284293,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 4989
    },
    {
      "epoch": 1.1672319027014384,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 4990
    },
    {
      "epoch": 1.1674657934744475,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4677,
      "step": 4991
    },
    {
      "epoch": 1.1676996842474565,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 4992
    },
    {
      "epoch": 1.1679335750204654,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 4993
    },
    {
      "epoch": 1.1681674657934744,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 4994
    },
    {
      "epoch": 1.1684013565664835,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 4995
    },
    {
      "epoch": 1.1686352473394925,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 4996
    },
    {
      "epoch": 1.1688691381125014,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 4997
    },
    {
      "epoch": 1.1691030288855104,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 4998
    },
    {
      "epoch": 1.1693369196585195,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2307,
      "step": 4999
    },
    {
      "epoch": 1.1695708104315286,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 5000
    },
    {
      "epoch": 1.1695708104315286,
      "eval_runtime": 4.6102,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 5000
    },
    {
      "epoch": 1.1698047012045374,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 5001
    },
    {
      "epoch": 1.1700385919775464,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 5002
    },
    {
      "epoch": 1.1702724827505555,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 5003
    },
    {
      "epoch": 1.1705063735235646,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 5004
    },
    {
      "epoch": 1.1707402642965734,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 5005
    },
    {
      "epoch": 1.1709741550695825,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 5006
    },
    {
      "epoch": 1.1712080458425915,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 5007
    },
    {
      "epoch": 1.1714419366156006,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 5008
    },
    {
      "epoch": 1.1716758273886094,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 5009
    },
    {
      "epoch": 1.1719097181616185,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7731,
      "step": 5010
    },
    {
      "epoch": 1.1721436089346275,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 5011
    },
    {
      "epoch": 1.1723774997076366,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 5012
    },
    {
      "epoch": 1.1726113904806454,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9657,
      "step": 5013
    },
    {
      "epoch": 1.1728452812536545,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 5014
    },
    {
      "epoch": 1.1730791720266636,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9152,
      "step": 5015
    },
    {
      "epoch": 1.1733130627996726,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 5016
    },
    {
      "epoch": 1.1735469535726815,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 5017
    },
    {
      "epoch": 1.1737808443456905,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 5018
    },
    {
      "epoch": 1.1740147351186996,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0391,
      "step": 5019
    },
    {
      "epoch": 1.1742486258917086,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8187,
      "step": 5020
    },
    {
      "epoch": 1.1744825166647175,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 5021
    },
    {
      "epoch": 1.1747164074377265,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.565,
      "step": 5022
    },
    {
      "epoch": 1.1749502982107356,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 5023
    },
    {
      "epoch": 1.1751841889837447,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 5024
    },
    {
      "epoch": 1.1754180797567537,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 5025
    },
    {
      "epoch": 1.1756519705297626,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 5026
    },
    {
      "epoch": 1.1758858613027716,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 5027
    },
    {
      "epoch": 1.1761197520757807,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5904,
      "step": 5028
    },
    {
      "epoch": 1.1763536428487895,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 5029
    },
    {
      "epoch": 1.1765875336217986,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 5030
    },
    {
      "epoch": 1.1768214243948076,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 5031
    },
    {
      "epoch": 1.1770553151678167,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 5032
    },
    {
      "epoch": 1.1772892059408258,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6343,
      "step": 5033
    },
    {
      "epoch": 1.1775230967138346,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 5034
    },
    {
      "epoch": 1.1777569874868437,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 5035
    },
    {
      "epoch": 1.1779908782598527,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 5036
    },
    {
      "epoch": 1.1782247690328616,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 5037
    },
    {
      "epoch": 1.1784586598058706,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 5038
    },
    {
      "epoch": 1.1786925505788797,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 5039
    },
    {
      "epoch": 1.1789264413518887,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 5040
    },
    {
      "epoch": 1.1791603321248978,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 5041
    },
    {
      "epoch": 1.1793942228979066,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7729,
      "step": 5042
    },
    {
      "epoch": 1.1796281136709157,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 5043
    },
    {
      "epoch": 1.1798620044439248,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 5044
    },
    {
      "epoch": 1.1800958952169336,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8473,
      "step": 5045
    },
    {
      "epoch": 1.1803297859899426,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 5046
    },
    {
      "epoch": 1.1805636767629517,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 5047
    },
    {
      "epoch": 1.1807975675359608,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 5048
    },
    {
      "epoch": 1.1810314583089698,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 5049
    },
    {
      "epoch": 1.1812653490819787,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 5050
    },
    {
      "epoch": 1.1814992398549877,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 5051
    },
    {
      "epoch": 1.1817331306279968,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 5052
    },
    {
      "epoch": 1.1819670214010056,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7333,
      "step": 5053
    },
    {
      "epoch": 1.1822009121740147,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 5054
    },
    {
      "epoch": 1.1824348029470237,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 5055
    },
    {
      "epoch": 1.1826686937200328,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 5056
    },
    {
      "epoch": 1.1829025844930419,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 5057
    },
    {
      "epoch": 1.1831364752660507,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 5058
    },
    {
      "epoch": 1.1833703660390598,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 5059
    },
    {
      "epoch": 1.1836042568120688,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 5060
    },
    {
      "epoch": 1.1838381475850777,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 5061
    },
    {
      "epoch": 1.1840720383580867,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 5062
    },
    {
      "epoch": 1.1843059291310958,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 5063
    },
    {
      "epoch": 1.1845398199041048,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 5064
    },
    {
      "epoch": 1.184773710677114,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 5065
    },
    {
      "epoch": 1.1850076014501227,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 5066
    },
    {
      "epoch": 1.1852414922231318,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5593,
      "step": 5067
    },
    {
      "epoch": 1.1854753829961409,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 5068
    },
    {
      "epoch": 1.1857092737691497,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 5069
    },
    {
      "epoch": 1.1859431645421588,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 5070
    },
    {
      "epoch": 1.1861770553151678,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 5071
    },
    {
      "epoch": 1.1864109460881769,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3826,
      "step": 5072
    },
    {
      "epoch": 1.186644836861186,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 5073
    },
    {
      "epoch": 1.1868787276341948,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.687,
      "step": 5074
    },
    {
      "epoch": 1.1871126184072038,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 5075
    },
    {
      "epoch": 1.187346509180213,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 5076
    },
    {
      "epoch": 1.1875803999532217,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 5077
    },
    {
      "epoch": 1.1878142907262308,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 5078
    },
    {
      "epoch": 1.1880481814992399,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6116,
      "step": 5079
    },
    {
      "epoch": 1.188282072272249,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 5080
    },
    {
      "epoch": 1.188515963045258,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 5081
    },
    {
      "epoch": 1.1887498538182668,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 5082
    },
    {
      "epoch": 1.1889837445912759,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 5083
    },
    {
      "epoch": 1.189217635364285,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 5084
    },
    {
      "epoch": 1.1894515261372938,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 5085
    },
    {
      "epoch": 1.1896854169103028,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 5086
    },
    {
      "epoch": 1.1899193076833119,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 5087
    },
    {
      "epoch": 1.190153198456321,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 5088
    },
    {
      "epoch": 1.19038708922933,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 5089
    },
    {
      "epoch": 1.1906209800023388,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 5090
    },
    {
      "epoch": 1.190854870775348,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 5091
    },
    {
      "epoch": 1.191088761548357,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 5092
    },
    {
      "epoch": 1.1913226523213658,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 5093
    },
    {
      "epoch": 1.1915565430943749,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 5094
    },
    {
      "epoch": 1.191790433867384,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 5095
    },
    {
      "epoch": 1.192024324640393,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9503,
      "step": 5096
    },
    {
      "epoch": 1.192258215413402,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 5097
    },
    {
      "epoch": 1.1924921061864109,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 5098
    },
    {
      "epoch": 1.19272599695942,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 5099
    },
    {
      "epoch": 1.192959887732429,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7906,
      "step": 5100
    },
    {
      "epoch": 1.192959887732429,
      "eval_runtime": 4.6327,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 5100
    },
    {
      "epoch": 1.1931937785054378,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6998,
      "step": 5101
    },
    {
      "epoch": 1.193427669278447,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 5102
    },
    {
      "epoch": 1.193661560051456,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6235,
      "step": 5103
    },
    {
      "epoch": 1.193895450824465,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6652,
      "step": 5104
    },
    {
      "epoch": 1.194129341597474,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 5105
    },
    {
      "epoch": 1.194363232370483,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2681,
      "step": 5106
    },
    {
      "epoch": 1.194597123143492,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2235,
      "step": 5107
    },
    {
      "epoch": 1.194831013916501,
      "grad_norm": 14.6875,
      "learning_rate": 3e-05,
      "loss": 2.0396,
      "step": 5108
    },
    {
      "epoch": 1.1950649046895099,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 5109
    },
    {
      "epoch": 1.195298795462519,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7827,
      "step": 5110
    },
    {
      "epoch": 1.195532686235528,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 5111
    },
    {
      "epoch": 1.195766577008537,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 5112
    },
    {
      "epoch": 1.1960004677815461,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 5113
    },
    {
      "epoch": 1.196234358554555,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 5114
    },
    {
      "epoch": 1.196468249327564,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 5115
    },
    {
      "epoch": 1.196702140100573,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 5116
    },
    {
      "epoch": 1.196936030873582,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 5117
    },
    {
      "epoch": 1.197169921646591,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 5118
    },
    {
      "epoch": 1.1974038124196,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7261,
      "step": 5119
    },
    {
      "epoch": 1.197637703192609,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.6603,
      "step": 5120
    },
    {
      "epoch": 1.1978715939656182,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 5121
    },
    {
      "epoch": 1.198105484738627,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 5122
    },
    {
      "epoch": 1.198339375511636,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 5123
    },
    {
      "epoch": 1.1985732662846451,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 5124
    },
    {
      "epoch": 1.1988071570576542,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 5125
    },
    {
      "epoch": 1.199041047830663,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 5126
    },
    {
      "epoch": 1.199274938603672,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 5127
    },
    {
      "epoch": 1.1995088293766811,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 5128
    },
    {
      "epoch": 1.1997427201496902,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 5129
    },
    {
      "epoch": 1.199976610922699,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 5130
    },
    {
      "epoch": 1.200210501695708,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 5131
    },
    {
      "epoch": 1.2004443924687171,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 5132
    },
    {
      "epoch": 1.2006782832417262,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 5133
    },
    {
      "epoch": 1.200912174014735,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 5134
    },
    {
      "epoch": 1.201146064787744,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.827,
      "step": 5135
    },
    {
      "epoch": 1.2013799555607532,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 5136
    },
    {
      "epoch": 1.2016138463337622,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 5137
    },
    {
      "epoch": 1.201847737106771,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5946,
      "step": 5138
    },
    {
      "epoch": 1.2020816278797801,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 5139
    },
    {
      "epoch": 1.2023155186527892,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 5140
    },
    {
      "epoch": 1.2025494094257982,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 5141
    },
    {
      "epoch": 1.202783300198807,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 5142
    },
    {
      "epoch": 1.2030171909718161,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 5143
    },
    {
      "epoch": 1.2032510817448252,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 5144
    },
    {
      "epoch": 1.2034849725178343,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 5145
    },
    {
      "epoch": 1.203718863290843,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 5146
    },
    {
      "epoch": 1.2039527540638522,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 5147
    },
    {
      "epoch": 1.2041866448368612,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 5148
    },
    {
      "epoch": 1.2044205356098703,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 5149
    },
    {
      "epoch": 1.2046544263828791,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 5150
    },
    {
      "epoch": 1.2048883171558882,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 5151
    },
    {
      "epoch": 1.2051222079288972,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 5152
    },
    {
      "epoch": 1.2053560987019063,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 5153
    },
    {
      "epoch": 1.2055899894749151,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 5154
    },
    {
      "epoch": 1.2058238802479242,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 5155
    },
    {
      "epoch": 1.2060577710209333,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 5156
    },
    {
      "epoch": 1.2062916617939423,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 5157
    },
    {
      "epoch": 1.2065255525669512,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 5158
    },
    {
      "epoch": 1.2067594433399602,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 5159
    },
    {
      "epoch": 1.2069933341129693,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 5160
    },
    {
      "epoch": 1.2072272248859783,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 5161
    },
    {
      "epoch": 1.2074611156589872,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7052,
      "step": 5162
    },
    {
      "epoch": 1.2076950064319962,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 5163
    },
    {
      "epoch": 1.2079288972050053,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4497,
      "step": 5164
    },
    {
      "epoch": 1.2081627879780144,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 5165
    },
    {
      "epoch": 1.2083966787510232,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 5166
    },
    {
      "epoch": 1.2086305695240322,
      "grad_norm": 7.8125,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 5167
    },
    {
      "epoch": 1.2088644602970413,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 5168
    },
    {
      "epoch": 1.2090983510700504,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 5169
    },
    {
      "epoch": 1.2093322418430592,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 5170
    },
    {
      "epoch": 1.2095661326160683,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 5171
    },
    {
      "epoch": 1.2098000233890773,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 5172
    },
    {
      "epoch": 1.2100339141620864,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 5173
    },
    {
      "epoch": 1.2102678049350952,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 5174
    },
    {
      "epoch": 1.2105016957081043,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 5175
    },
    {
      "epoch": 1.2107355864811133,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 5176
    },
    {
      "epoch": 1.2109694772541224,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 5177
    },
    {
      "epoch": 1.2112033680271312,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 5178
    },
    {
      "epoch": 1.2114372588001403,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 5179
    },
    {
      "epoch": 1.2116711495731494,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 5180
    },
    {
      "epoch": 1.2119050403461584,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 5181
    },
    {
      "epoch": 1.2121389311191673,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 5182
    },
    {
      "epoch": 1.2123728218921763,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 5183
    },
    {
      "epoch": 1.2126067126651854,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 5184
    },
    {
      "epoch": 1.2128406034381944,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 5185
    },
    {
      "epoch": 1.2130744942112033,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 5186
    },
    {
      "epoch": 1.2133083849842123,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 5187
    },
    {
      "epoch": 1.2135422757572214,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 5188
    },
    {
      "epoch": 1.2137761665302305,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 5189
    },
    {
      "epoch": 1.2140100573032393,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 5190
    },
    {
      "epoch": 1.2142439480762484,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 5191
    },
    {
      "epoch": 1.2144778388492574,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 5192
    },
    {
      "epoch": 1.2147117296222665,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 5193
    },
    {
      "epoch": 1.2149456203952753,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 5194
    },
    {
      "epoch": 1.2151795111682844,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 5195
    },
    {
      "epoch": 1.2154134019412934,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 5196
    },
    {
      "epoch": 1.2156472927143025,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 5197
    },
    {
      "epoch": 1.2158811834873113,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 5198
    },
    {
      "epoch": 1.2161150742603204,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 5199
    },
    {
      "epoch": 1.2163489650333295,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 5200
    },
    {
      "epoch": 1.2163489650333295,
      "eval_runtime": 4.6227,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 5200
    },
    {
      "epoch": 1.2165828558063385,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 5201
    },
    {
      "epoch": 1.2168167465793474,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 5202
    },
    {
      "epoch": 1.2170506373523564,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 5203
    },
    {
      "epoch": 1.2172845281253655,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 5204
    },
    {
      "epoch": 1.2175184188983745,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 5205
    },
    {
      "epoch": 1.2177523096713834,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 5206
    },
    {
      "epoch": 1.2179862004443924,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 5207
    },
    {
      "epoch": 1.2182200912174015,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 5208
    },
    {
      "epoch": 1.2184539819904106,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 5209
    },
    {
      "epoch": 1.2186878727634194,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 5210
    },
    {
      "epoch": 1.2189217635364284,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 5211
    },
    {
      "epoch": 1.2191556543094375,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 5212
    },
    {
      "epoch": 1.2193895450824466,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 5213
    },
    {
      "epoch": 1.2196234358554554,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 5214
    },
    {
      "epoch": 1.2198573266284645,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6616,
      "step": 5215
    },
    {
      "epoch": 1.2200912174014735,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 5216
    },
    {
      "epoch": 1.2203251081744826,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 5217
    },
    {
      "epoch": 1.2205589989474914,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.755,
      "step": 5218
    },
    {
      "epoch": 1.2207928897205005,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 5219
    },
    {
      "epoch": 1.2210267804935095,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 5220
    },
    {
      "epoch": 1.2212606712665186,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7549,
      "step": 5221
    },
    {
      "epoch": 1.2214945620395277,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.57,
      "step": 5222
    },
    {
      "epoch": 1.2217284528125365,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 5223
    },
    {
      "epoch": 1.2219623435855456,
      "grad_norm": 7.5625,
      "learning_rate": 3e-05,
      "loss": 2.5273,
      "step": 5224
    },
    {
      "epoch": 1.2221962343585546,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 5225
    },
    {
      "epoch": 1.2224301251315635,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 5226
    },
    {
      "epoch": 1.2226640159045725,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 5227
    },
    {
      "epoch": 1.2228979066775816,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 5228
    },
    {
      "epoch": 1.2231317974505906,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2784,
      "step": 5229
    },
    {
      "epoch": 1.2233656882235997,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 5230
    },
    {
      "epoch": 1.2235995789966085,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 5231
    },
    {
      "epoch": 1.2238334697696176,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 5232
    },
    {
      "epoch": 1.2240673605426267,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 5233
    },
    {
      "epoch": 1.2243012513156355,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 5234
    },
    {
      "epoch": 1.2245351420886446,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 5235
    },
    {
      "epoch": 1.2247690328616536,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 5236
    },
    {
      "epoch": 1.2250029236346627,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 5237
    },
    {
      "epoch": 1.2252368144076717,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 5238
    },
    {
      "epoch": 1.2254707051806806,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5901,
      "step": 5239
    },
    {
      "epoch": 1.2257045959536896,
      "grad_norm": 7.65625,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 5240
    },
    {
      "epoch": 1.2259384867266987,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 5241
    },
    {
      "epoch": 1.2261723774997075,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 5242
    },
    {
      "epoch": 1.2264062682727166,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 5243
    },
    {
      "epoch": 1.2266401590457257,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 5244
    },
    {
      "epoch": 1.2268740498187347,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 5245
    },
    {
      "epoch": 1.2271079405917438,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 5246
    },
    {
      "epoch": 1.2273418313647526,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 5247
    },
    {
      "epoch": 1.2275757221377617,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 5248
    },
    {
      "epoch": 1.2278096129107707,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 5249
    },
    {
      "epoch": 1.2280435036837796,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6795,
      "step": 5250
    },
    {
      "epoch": 1.2282773944567886,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6137,
      "step": 5251
    },
    {
      "epoch": 1.2285112852297977,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 5252
    },
    {
      "epoch": 1.2287451760028067,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 5253
    },
    {
      "epoch": 1.2289790667758158,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 5254
    },
    {
      "epoch": 1.2292129575488246,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 5255
    },
    {
      "epoch": 1.2294468483218337,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 5256
    },
    {
      "epoch": 1.2296807390948428,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 5257
    },
    {
      "epoch": 1.2299146298678516,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6419,
      "step": 5258
    },
    {
      "epoch": 1.2301485206408607,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 5259
    },
    {
      "epoch": 1.2303824114138697,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 5260
    },
    {
      "epoch": 1.2306163021868788,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 5261
    },
    {
      "epoch": 1.2308501929598878,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 5262
    },
    {
      "epoch": 1.2310840837328967,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 5263
    },
    {
      "epoch": 1.2313179745059057,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 5264
    },
    {
      "epoch": 1.2315518652789148,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 5265
    },
    {
      "epoch": 1.2317857560519236,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 5266
    },
    {
      "epoch": 1.2320196468249327,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 5267
    },
    {
      "epoch": 1.2322535375979418,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 5268
    },
    {
      "epoch": 1.2324874283709508,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2332,
      "step": 5269
    },
    {
      "epoch": 1.2327213191439599,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 5270
    },
    {
      "epoch": 1.2329552099169687,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 5271
    },
    {
      "epoch": 1.2331891006899778,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 5272
    },
    {
      "epoch": 1.2334229914629868,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 5273
    },
    {
      "epoch": 1.2336568822359957,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 5274
    },
    {
      "epoch": 1.2338907730090047,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9136,
      "step": 5275
    },
    {
      "epoch": 1.2341246637820138,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 5276
    },
    {
      "epoch": 1.2343585545550229,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 5277
    },
    {
      "epoch": 1.234592445328032,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 5278
    },
    {
      "epoch": 1.2348263361010408,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 5279
    },
    {
      "epoch": 1.2350602268740498,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 5280
    },
    {
      "epoch": 1.2352941176470589,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 5281
    },
    {
      "epoch": 1.2355280084200677,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6882,
      "step": 5282
    },
    {
      "epoch": 1.2357618991930768,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 5283
    },
    {
      "epoch": 1.2359957899660858,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 5284
    },
    {
      "epoch": 1.236229680739095,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4338,
      "step": 5285
    },
    {
      "epoch": 1.236463571512104,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 5286
    },
    {
      "epoch": 1.2366974622851128,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 5287
    },
    {
      "epoch": 1.2369313530581219,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 5288
    },
    {
      "epoch": 1.237165243831131,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 5289
    },
    {
      "epoch": 1.2373991346041397,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0601,
      "step": 5290
    },
    {
      "epoch": 1.2376330253771488,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 5291
    },
    {
      "epoch": 1.2378669161501579,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6486,
      "step": 5292
    },
    {
      "epoch": 1.238100806923167,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.549,
      "step": 5293
    },
    {
      "epoch": 1.238334697696176,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 5294
    },
    {
      "epoch": 1.2385685884691848,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 5295
    },
    {
      "epoch": 1.2388024792421939,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 5296
    },
    {
      "epoch": 1.239036370015203,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 5297
    },
    {
      "epoch": 1.2392702607882118,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 5298
    },
    {
      "epoch": 1.2395041515612208,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 5299
    },
    {
      "epoch": 1.23973804233423,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 5300
    },
    {
      "epoch": 1.23973804233423,
      "eval_runtime": 4.6071,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 5300
    },
    {
      "epoch": 1.239971933107239,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2507,
      "step": 5301
    },
    {
      "epoch": 1.240205823880248,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 5302
    },
    {
      "epoch": 1.2404397146532569,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 5303
    },
    {
      "epoch": 1.240673605426266,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 5304
    },
    {
      "epoch": 1.240907496199275,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 5305
    },
    {
      "epoch": 1.2411413869722838,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 5306
    },
    {
      "epoch": 1.2413752777452929,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 5307
    },
    {
      "epoch": 1.241609168518302,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 5308
    },
    {
      "epoch": 1.241843059291311,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4016,
      "step": 5309
    },
    {
      "epoch": 1.24207695006432,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 5310
    },
    {
      "epoch": 1.242310840837329,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 5311
    },
    {
      "epoch": 1.242544731610338,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6304,
      "step": 5312
    },
    {
      "epoch": 1.242778622383347,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 5313
    },
    {
      "epoch": 1.2430125131563559,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 5314
    },
    {
      "epoch": 1.243246403929365,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 5315
    },
    {
      "epoch": 1.243480294702374,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 5316
    },
    {
      "epoch": 1.243714185475383,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 5317
    },
    {
      "epoch": 1.243948076248392,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 5318
    },
    {
      "epoch": 1.244181967021401,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 5319
    },
    {
      "epoch": 1.24441585779441,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 5320
    },
    {
      "epoch": 1.244649748567419,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6729,
      "step": 5321
    },
    {
      "epoch": 1.2448836393404281,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.5613,
      "step": 5322
    },
    {
      "epoch": 1.245117530113437,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 5323
    },
    {
      "epoch": 1.245351420886446,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 5324
    },
    {
      "epoch": 1.245585311659455,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 5325
    },
    {
      "epoch": 1.2458192024324641,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 5326
    },
    {
      "epoch": 1.246053093205473,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 5327
    },
    {
      "epoch": 1.246286983978482,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 5328
    },
    {
      "epoch": 1.246520874751491,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 5329
    },
    {
      "epoch": 1.2467547655245002,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 5330
    },
    {
      "epoch": 1.246988656297509,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 5331
    },
    {
      "epoch": 1.247222547070518,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 5332
    },
    {
      "epoch": 1.247456437843527,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 5333
    },
    {
      "epoch": 1.2476903286165362,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1369,
      "step": 5334
    },
    {
      "epoch": 1.247924219389545,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 5335
    },
    {
      "epoch": 1.248158110162554,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 5336
    },
    {
      "epoch": 1.2483920009355631,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 5337
    },
    {
      "epoch": 1.2486258917085722,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 5338
    },
    {
      "epoch": 1.248859782481581,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 5339
    },
    {
      "epoch": 1.24909367325459,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 5340
    },
    {
      "epoch": 1.2493275640275991,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7349,
      "step": 5341
    },
    {
      "epoch": 1.2495614548006082,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 5342
    },
    {
      "epoch": 1.249795345573617,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2879,
      "step": 5343
    },
    {
      "epoch": 1.250029236346626,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 5344
    },
    {
      "epoch": 1.2502631271196352,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 5345
    },
    {
      "epoch": 1.250497017892644,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 5346
    },
    {
      "epoch": 1.250730908665653,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 5347
    },
    {
      "epoch": 1.2509647994386621,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 5348
    },
    {
      "epoch": 1.2511986902116712,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8234,
      "step": 5349
    },
    {
      "epoch": 1.2514325809846802,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 5350
    },
    {
      "epoch": 1.251666471757689,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 5351
    },
    {
      "epoch": 1.2519003625306981,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 5352
    },
    {
      "epoch": 1.2521342533037072,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 5353
    },
    {
      "epoch": 1.252368144076716,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7507,
      "step": 5354
    },
    {
      "epoch": 1.252602034849725,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 5355
    },
    {
      "epoch": 1.2528359256227342,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 5356
    },
    {
      "epoch": 1.2530698163957432,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 5357
    },
    {
      "epoch": 1.2533037071687523,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 5358
    },
    {
      "epoch": 1.2535375979417611,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.4696,
      "step": 5359
    },
    {
      "epoch": 1.2537714887147702,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3904,
      "step": 5360
    },
    {
      "epoch": 1.2540053794877792,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 5361
    },
    {
      "epoch": 1.254239270260788,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 5362
    },
    {
      "epoch": 1.2544731610337971,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 5363
    },
    {
      "epoch": 1.2547070518068062,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1686,
      "step": 5364
    },
    {
      "epoch": 1.2549409425798153,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 5365
    },
    {
      "epoch": 1.2551748333528243,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 2.3873,
      "step": 5366
    },
    {
      "epoch": 1.2554087241258332,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8011,
      "step": 5367
    },
    {
      "epoch": 1.2556426148988422,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6076,
      "step": 5368
    },
    {
      "epoch": 1.2558765056718513,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7934,
      "step": 5369
    },
    {
      "epoch": 1.2561103964448603,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 5370
    },
    {
      "epoch": 1.2563442872178694,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6383,
      "step": 5371
    },
    {
      "epoch": 1.2565781779908782,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 5372
    },
    {
      "epoch": 1.2568120687638873,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 5373
    },
    {
      "epoch": 1.2570459595368964,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 5374
    },
    {
      "epoch": 1.2572798503099052,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 5375
    },
    {
      "epoch": 1.2575137410829142,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 5376
    },
    {
      "epoch": 1.2577476318559233,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 5377
    },
    {
      "epoch": 1.2579815226289324,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 5378
    },
    {
      "epoch": 1.2582154134019414,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 5379
    },
    {
      "epoch": 1.2584493041749503,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 5380
    },
    {
      "epoch": 1.2586831949479593,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 5381
    },
    {
      "epoch": 1.2589170857209684,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 5382
    },
    {
      "epoch": 1.2591509764939772,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 5383
    },
    {
      "epoch": 1.2593848672669863,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 5384
    },
    {
      "epoch": 1.2596187580399953,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 5385
    },
    {
      "epoch": 1.2598526488130044,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 5386
    },
    {
      "epoch": 1.2600865395860135,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 5387
    },
    {
      "epoch": 1.2603204303590223,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6603,
      "step": 5388
    },
    {
      "epoch": 1.2605543211320314,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 5389
    },
    {
      "epoch": 1.2607882119050404,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 5390
    },
    {
      "epoch": 1.2610221026780493,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 5391
    },
    {
      "epoch": 1.2612559934510583,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 5392
    },
    {
      "epoch": 1.2614898842240674,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 5393
    },
    {
      "epoch": 1.2617237749970764,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 5394
    },
    {
      "epoch": 1.2619576657700855,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 5395
    },
    {
      "epoch": 1.2621915565430943,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 5396
    },
    {
      "epoch": 1.2624254473161034,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 5397
    },
    {
      "epoch": 1.2626593380891125,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 5398
    },
    {
      "epoch": 1.2628932288621213,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 5399
    },
    {
      "epoch": 1.2631271196351304,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 5400
    },
    {
      "epoch": 1.2631271196351304,
      "eval_runtime": 4.5923,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 5400
    },
    {
      "epoch": 1.2633610104081394,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 5401
    },
    {
      "epoch": 1.2635949011811485,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 5402
    },
    {
      "epoch": 1.2638287919541575,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6268,
      "step": 5403
    },
    {
      "epoch": 1.2640626827271664,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 5404
    },
    {
      "epoch": 1.2642965735001754,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2143,
      "step": 5405
    },
    {
      "epoch": 1.2645304642731845,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 5406
    },
    {
      "epoch": 1.2647643550461933,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6463,
      "step": 5407
    },
    {
      "epoch": 1.2649982458192024,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 5408
    },
    {
      "epoch": 1.2652321365922115,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 5409
    },
    {
      "epoch": 1.2654660273652205,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3063,
      "step": 5410
    },
    {
      "epoch": 1.2656999181382296,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 5411
    },
    {
      "epoch": 1.2659338089112384,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 5412
    },
    {
      "epoch": 1.2661676996842475,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.768,
      "step": 5413
    },
    {
      "epoch": 1.2664015904572565,
      "grad_norm": 7.875,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 5414
    },
    {
      "epoch": 1.2666354812302654,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5863,
      "step": 5415
    },
    {
      "epoch": 1.2668693720032744,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 5416
    },
    {
      "epoch": 1.2671032627762835,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 5417
    },
    {
      "epoch": 1.2673371535492925,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6107,
      "step": 5418
    },
    {
      "epoch": 1.2675710443223016,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 5419
    },
    {
      "epoch": 1.2678049350953104,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 5420
    },
    {
      "epoch": 1.2680388258683195,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 5421
    },
    {
      "epoch": 1.2682727166413286,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 5422
    },
    {
      "epoch": 1.2685066074143374,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 5423
    },
    {
      "epoch": 1.2687404981873465,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 5424
    },
    {
      "epoch": 1.2689743889603555,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5316,
      "step": 5425
    },
    {
      "epoch": 1.2692082797333646,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7908,
      "step": 5426
    },
    {
      "epoch": 1.2694421705063736,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5987,
      "step": 5427
    },
    {
      "epoch": 1.2696760612793825,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 5428
    },
    {
      "epoch": 1.2699099520523915,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 5429
    },
    {
      "epoch": 1.2701438428254006,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5972,
      "step": 5430
    },
    {
      "epoch": 1.2703777335984094,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 5431
    },
    {
      "epoch": 1.2706116243714185,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 5432
    },
    {
      "epoch": 1.2708455151444276,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 5433
    },
    {
      "epoch": 1.2710794059174366,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7372,
      "step": 5434
    },
    {
      "epoch": 1.2713132966904457,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3194,
      "step": 5435
    },
    {
      "epoch": 1.2715471874634545,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 5436
    },
    {
      "epoch": 1.2717810782364636,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 5437
    },
    {
      "epoch": 1.2720149690094726,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 5438
    },
    {
      "epoch": 1.2722488597824815,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 5439
    },
    {
      "epoch": 1.2724827505554905,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 5440
    },
    {
      "epoch": 1.2727166413284996,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 5441
    },
    {
      "epoch": 1.2729505321015087,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 5442
    },
    {
      "epoch": 1.2731844228745177,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 5443
    },
    {
      "epoch": 1.2734183136475266,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 5444
    },
    {
      "epoch": 1.2736522044205356,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 5445
    },
    {
      "epoch": 1.2738860951935447,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 5446
    },
    {
      "epoch": 1.2741199859665535,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 5447
    },
    {
      "epoch": 1.2743538767395626,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 5448
    },
    {
      "epoch": 1.2745877675125716,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 5449
    },
    {
      "epoch": 1.2748216582855807,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 5450
    },
    {
      "epoch": 1.2750555490585898,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 5451
    },
    {
      "epoch": 1.2752894398315986,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 5452
    },
    {
      "epoch": 1.2755233306046077,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 5453
    },
    {
      "epoch": 1.2757572213776167,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 5454
    },
    {
      "epoch": 1.2759911121506256,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 5455
    },
    {
      "epoch": 1.2762250029236346,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 5456
    },
    {
      "epoch": 1.2764588936966437,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 5457
    },
    {
      "epoch": 1.2766927844696527,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 5458
    },
    {
      "epoch": 1.2769266752426618,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 5459
    },
    {
      "epoch": 1.2771605660156706,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 5460
    },
    {
      "epoch": 1.2773944567886797,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 5461
    },
    {
      "epoch": 1.2776283475616887,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.784,
      "step": 5462
    },
    {
      "epoch": 1.2778622383346976,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6239,
      "step": 5463
    },
    {
      "epoch": 1.2780961291077066,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 5464
    },
    {
      "epoch": 1.2783300198807157,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 5465
    },
    {
      "epoch": 1.2785639106537248,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 5466
    },
    {
      "epoch": 1.2787978014267338,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 5467
    },
    {
      "epoch": 1.2790316921997427,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 5468
    },
    {
      "epoch": 1.2792655829727517,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7996,
      "step": 5469
    },
    {
      "epoch": 1.2794994737457608,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 5470
    },
    {
      "epoch": 1.2797333645187696,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 5471
    },
    {
      "epoch": 1.2799672552917787,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 5472
    },
    {
      "epoch": 1.2802011460647877,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 5473
    },
    {
      "epoch": 1.2804350368377968,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 5474
    },
    {
      "epoch": 1.2806689276108059,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 5475
    },
    {
      "epoch": 1.2809028183838147,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 5476
    },
    {
      "epoch": 1.2811367091568238,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 5477
    },
    {
      "epoch": 1.2813705999298328,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 5478
    },
    {
      "epoch": 1.2816044907028417,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 5479
    },
    {
      "epoch": 1.2818383814758507,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 5480
    },
    {
      "epoch": 1.2820722722488598,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 5481
    },
    {
      "epoch": 1.2823061630218688,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.547,
      "step": 5482
    },
    {
      "epoch": 1.282540053794878,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 5483
    },
    {
      "epoch": 1.2827739445678867,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.4498,
      "step": 5484
    },
    {
      "epoch": 1.2830078353408958,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 5485
    },
    {
      "epoch": 1.2832417261139049,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.717,
      "step": 5486
    },
    {
      "epoch": 1.2834756168869137,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2711,
      "step": 5487
    },
    {
      "epoch": 1.2837095076599228,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 5488
    },
    {
      "epoch": 1.2839433984329318,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 5489
    },
    {
      "epoch": 1.2841772892059409,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 5490
    },
    {
      "epoch": 1.28441117997895,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 5491
    },
    {
      "epoch": 1.2846450707519588,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 5492
    },
    {
      "epoch": 1.2848789615249678,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 5493
    },
    {
      "epoch": 1.285112852297977,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 5494
    },
    {
      "epoch": 1.2853467430709857,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8131,
      "step": 5495
    },
    {
      "epoch": 1.2855806338439948,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 5496
    },
    {
      "epoch": 1.2858145246170039,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 5497
    },
    {
      "epoch": 1.286048415390013,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6299,
      "step": 5498
    },
    {
      "epoch": 1.286282306163022,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 5499
    },
    {
      "epoch": 1.2865161969360308,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 5500
    },
    {
      "epoch": 1.2865161969360308,
      "eval_runtime": 4.6015,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 5500
    },
    {
      "epoch": 1.2867500877090399,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 5501
    },
    {
      "epoch": 1.286983978482049,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 5502
    },
    {
      "epoch": 1.2872178692550578,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 5503
    },
    {
      "epoch": 1.2874517600280668,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 5504
    },
    {
      "epoch": 1.2876856508010759,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 5505
    },
    {
      "epoch": 1.287919541574085,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 5506
    },
    {
      "epoch": 1.288153432347094,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5751,
      "step": 5507
    },
    {
      "epoch": 1.2883873231201028,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 5508
    },
    {
      "epoch": 1.288621213893112,
      "grad_norm": 17.25,
      "learning_rate": 3e-05,
      "loss": 2.4563,
      "step": 5509
    },
    {
      "epoch": 1.288855104666121,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.719,
      "step": 5510
    },
    {
      "epoch": 1.2890889954391298,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7123,
      "step": 5511
    },
    {
      "epoch": 1.2893228862121389,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 5512
    },
    {
      "epoch": 1.289556776985148,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 5513
    },
    {
      "epoch": 1.289790667758157,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8007,
      "step": 5514
    },
    {
      "epoch": 1.290024558531166,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 5515
    },
    {
      "epoch": 1.2902584493041749,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 5516
    },
    {
      "epoch": 1.290492340077184,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3109,
      "step": 5517
    },
    {
      "epoch": 1.290726230850193,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 5518
    },
    {
      "epoch": 1.2909601216232018,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7656,
      "step": 5519
    },
    {
      "epoch": 1.291194012396211,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2513,
      "step": 5520
    },
    {
      "epoch": 1.29142790316922,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 5521
    },
    {
      "epoch": 1.291661793942229,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 5522
    },
    {
      "epoch": 1.291895684715238,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 5523
    },
    {
      "epoch": 1.292129575488247,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4047,
      "step": 5524
    },
    {
      "epoch": 1.292363466261256,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 5525
    },
    {
      "epoch": 1.292597357034265,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7732,
      "step": 5526
    },
    {
      "epoch": 1.2928312478072739,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 5527
    },
    {
      "epoch": 1.293065138580283,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6173,
      "step": 5528
    },
    {
      "epoch": 1.293299029353292,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 5529
    },
    {
      "epoch": 1.293532920126301,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 5530
    },
    {
      "epoch": 1.2937668108993101,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 5531
    },
    {
      "epoch": 1.294000701672319,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 5532
    },
    {
      "epoch": 1.294234592445328,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6552,
      "step": 5533
    },
    {
      "epoch": 1.294468483218337,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 5534
    },
    {
      "epoch": 1.294702373991346,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 5535
    },
    {
      "epoch": 1.294936264764355,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0913,
      "step": 5536
    },
    {
      "epoch": 1.295170155537364,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 5537
    },
    {
      "epoch": 1.295404046310373,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 5538
    },
    {
      "epoch": 1.2956379370833822,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 5539
    },
    {
      "epoch": 1.295871827856391,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 5540
    },
    {
      "epoch": 1.2961057186294,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 5541
    },
    {
      "epoch": 1.296339609402409,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 5542
    },
    {
      "epoch": 1.296573500175418,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6625,
      "step": 5543
    },
    {
      "epoch": 1.296807390948427,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 5544
    },
    {
      "epoch": 1.297041281721436,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 5545
    },
    {
      "epoch": 1.2972751724944451,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7073,
      "step": 5546
    },
    {
      "epoch": 1.2975090632674542,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 5547
    },
    {
      "epoch": 1.297742954040463,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 5548
    },
    {
      "epoch": 1.297976844813472,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 5549
    },
    {
      "epoch": 1.2982107355864811,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7932,
      "step": 5550
    },
    {
      "epoch": 1.29844462635949,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7121,
      "step": 5551
    },
    {
      "epoch": 1.298678517132499,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 5552
    },
    {
      "epoch": 1.298912407905508,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 5553
    },
    {
      "epoch": 1.2991462986785172,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 5554
    },
    {
      "epoch": 1.2993801894515262,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 5555
    },
    {
      "epoch": 1.299614080224535,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 5556
    },
    {
      "epoch": 1.2998479709975441,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 5557
    },
    {
      "epoch": 1.3000818617705532,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 5558
    },
    {
      "epoch": 1.300315752543562,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6918,
      "step": 5559
    },
    {
      "epoch": 1.300549643316571,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 5560
    },
    {
      "epoch": 1.3007835340895801,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 5561
    },
    {
      "epoch": 1.3010174248625892,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 5562
    },
    {
      "epoch": 1.3012513156355983,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 5563
    },
    {
      "epoch": 1.301485206408607,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 5564
    },
    {
      "epoch": 1.3017190971816162,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 5565
    },
    {
      "epoch": 1.3019529879546252,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 5566
    },
    {
      "epoch": 1.302186878727634,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 5567
    },
    {
      "epoch": 1.3024207695006433,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 5568
    },
    {
      "epoch": 1.3026546602736522,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 5569
    },
    {
      "epoch": 1.3028885510466612,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 5570
    },
    {
      "epoch": 1.3031224418196703,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 5571
    },
    {
      "epoch": 1.3033563325926791,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 5572
    },
    {
      "epoch": 1.3035902233656882,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6793,
      "step": 5573
    },
    {
      "epoch": 1.3038241141386973,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 5574
    },
    {
      "epoch": 1.3040580049117063,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 5575
    },
    {
      "epoch": 1.3042918956847154,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.437,
      "step": 5576
    },
    {
      "epoch": 1.3045257864577242,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6406,
      "step": 5577
    },
    {
      "epoch": 1.3047596772307333,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 5578
    },
    {
      "epoch": 1.3049935680037423,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6741,
      "step": 5579
    },
    {
      "epoch": 1.3052274587767512,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 5580
    },
    {
      "epoch": 1.3054613495497602,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 5581
    },
    {
      "epoch": 1.3056952403227693,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 5582
    },
    {
      "epoch": 1.3059291310957784,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 5583
    },
    {
      "epoch": 1.3061630218687874,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 5584
    },
    {
      "epoch": 1.3063969126417962,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7366,
      "step": 5585
    },
    {
      "epoch": 1.3066308034148053,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 5586
    },
    {
      "epoch": 1.3068646941878144,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 5587
    },
    {
      "epoch": 1.3070985849608232,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 5588
    },
    {
      "epoch": 1.3073324757338323,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 5589
    },
    {
      "epoch": 1.3075663665068413,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 5590
    },
    {
      "epoch": 1.3078002572798504,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5503,
      "step": 5591
    },
    {
      "epoch": 1.3080341480528594,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5898,
      "step": 5592
    },
    {
      "epoch": 1.3082680388258683,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 5593
    },
    {
      "epoch": 1.3085019295988773,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1122,
      "step": 5594
    },
    {
      "epoch": 1.3087358203718864,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8751,
      "step": 5595
    },
    {
      "epoch": 1.3089697111448952,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 5596
    },
    {
      "epoch": 1.3092036019179043,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 5597
    },
    {
      "epoch": 1.3094374926909134,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 5598
    },
    {
      "epoch": 1.3096713834639224,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 5599
    },
    {
      "epoch": 1.3099052742369315,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 5600
    },
    {
      "epoch": 1.3099052742369315,
      "eval_runtime": 4.6095,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 5600
    },
    {
      "epoch": 1.3101391650099403,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 5601
    },
    {
      "epoch": 1.3103730557829494,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 5602
    },
    {
      "epoch": 1.3106069465559584,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4786,
      "step": 5603
    },
    {
      "epoch": 1.3108408373289673,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 5604
    },
    {
      "epoch": 1.3110747281019763,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 5605
    },
    {
      "epoch": 1.3113086188749854,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7609,
      "step": 5606
    },
    {
      "epoch": 1.3115425096479945,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 5607
    },
    {
      "epoch": 1.3117764004210035,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 5608
    },
    {
      "epoch": 1.3120102911940124,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8903,
      "step": 5609
    },
    {
      "epoch": 1.3122441819670214,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 5610
    },
    {
      "epoch": 1.3124780727400305,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 5611
    },
    {
      "epoch": 1.3127119635130393,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3615,
      "step": 5612
    },
    {
      "epoch": 1.3129458542860484,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 5613
    },
    {
      "epoch": 1.3131797450590574,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6782,
      "step": 5614
    },
    {
      "epoch": 1.3134136358320665,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 5615
    },
    {
      "epoch": 1.3136475266050756,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 5616
    },
    {
      "epoch": 1.3138814173780844,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 5617
    },
    {
      "epoch": 1.3141153081510935,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 5618
    },
    {
      "epoch": 1.3143491989241025,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 5619
    },
    {
      "epoch": 1.3145830896971114,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 5620
    },
    {
      "epoch": 1.3148169804701204,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 5621
    },
    {
      "epoch": 1.3150508712431295,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 5622
    },
    {
      "epoch": 1.3152847620161385,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 5623
    },
    {
      "epoch": 1.3155186527891476,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 5624
    },
    {
      "epoch": 1.3157525435621564,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 5625
    },
    {
      "epoch": 1.3159864343351655,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 5626
    },
    {
      "epoch": 1.3162203251081745,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 5627
    },
    {
      "epoch": 1.3164542158811834,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7076,
      "step": 5628
    },
    {
      "epoch": 1.3166881066541924,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 5629
    },
    {
      "epoch": 1.3169219974272015,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 5630
    },
    {
      "epoch": 1.3171558882002106,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 5631
    },
    {
      "epoch": 1.3173897789732196,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 5632
    },
    {
      "epoch": 1.3176236697462285,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 5633
    },
    {
      "epoch": 1.3178575605192375,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 5634
    },
    {
      "epoch": 1.3180914512922466,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 5635
    },
    {
      "epoch": 1.3183253420652554,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.916,
      "step": 5636
    },
    {
      "epoch": 1.3185592328382645,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2664,
      "step": 5637
    },
    {
      "epoch": 1.3187931236112735,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 5638
    },
    {
      "epoch": 1.3190270143842826,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 5639
    },
    {
      "epoch": 1.3192609051572917,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7872,
      "step": 5640
    },
    {
      "epoch": 1.3194947959303005,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 5641
    },
    {
      "epoch": 1.3197286867033096,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 5642
    },
    {
      "epoch": 1.3199625774763186,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 5643
    },
    {
      "epoch": 1.3201964682493275,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 5644
    },
    {
      "epoch": 1.3204303590223365,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 5645
    },
    {
      "epoch": 1.3206642497953456,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 5646
    },
    {
      "epoch": 1.3208981405683546,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 5647
    },
    {
      "epoch": 1.3211320313413637,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 5648
    },
    {
      "epoch": 1.3213659221143725,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 5649
    },
    {
      "epoch": 1.3215998128873816,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 5650
    },
    {
      "epoch": 1.3218337036603907,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 5651
    },
    {
      "epoch": 1.3220675944333995,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 5652
    },
    {
      "epoch": 1.3223014852064086,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 5653
    },
    {
      "epoch": 1.3225353759794176,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 5654
    },
    {
      "epoch": 1.3227692667524267,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 5655
    },
    {
      "epoch": 1.3230031575254357,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 5656
    },
    {
      "epoch": 1.3232370482984446,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9017,
      "step": 5657
    },
    {
      "epoch": 1.3234709390714536,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 5658
    },
    {
      "epoch": 1.3237048298444627,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 5659
    },
    {
      "epoch": 1.3239387206174715,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 5660
    },
    {
      "epoch": 1.3241726113904806,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 5661
    },
    {
      "epoch": 1.3244065021634897,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1761,
      "step": 5662
    },
    {
      "epoch": 1.3246403929364987,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 5663
    },
    {
      "epoch": 1.3248742837095078,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 5664
    },
    {
      "epoch": 1.3251081744825166,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 5665
    },
    {
      "epoch": 1.3253420652555257,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 5666
    },
    {
      "epoch": 1.3255759560285347,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 5667
    },
    {
      "epoch": 1.3258098468015436,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 5668
    },
    {
      "epoch": 1.3260437375745526,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 5669
    },
    {
      "epoch": 1.3262776283475617,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1556,
      "step": 5670
    },
    {
      "epoch": 1.3265115191205707,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 5671
    },
    {
      "epoch": 1.3267454098935798,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 5672
    },
    {
      "epoch": 1.3269793006665886,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8251,
      "step": 5673
    },
    {
      "epoch": 1.3272131914395977,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 5674
    },
    {
      "epoch": 1.3274470822126068,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 5675
    },
    {
      "epoch": 1.3276809729856156,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 5676
    },
    {
      "epoch": 1.3279148637586247,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 5677
    },
    {
      "epoch": 1.3281487545316337,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 5678
    },
    {
      "epoch": 1.3283826453046428,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 5679
    },
    {
      "epoch": 1.3286165360776518,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 5680
    },
    {
      "epoch": 1.3288504268506607,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 5681
    },
    {
      "epoch": 1.3290843176236697,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 5682
    },
    {
      "epoch": 1.3293182083966788,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 5683
    },
    {
      "epoch": 1.3295520991696876,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 5684
    },
    {
      "epoch": 1.3297859899426967,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 5685
    },
    {
      "epoch": 1.3300198807157058,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2758,
      "step": 5686
    },
    {
      "epoch": 1.3302537714887148,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6192,
      "step": 5687
    },
    {
      "epoch": 1.3304876622617239,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 5688
    },
    {
      "epoch": 1.3307215530347327,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 5689
    },
    {
      "epoch": 1.3309554438077418,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 5690
    },
    {
      "epoch": 1.3311893345807508,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 5691
    },
    {
      "epoch": 1.3314232253537597,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 5692
    },
    {
      "epoch": 1.3316571161267687,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 5693
    },
    {
      "epoch": 1.3318910068997778,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 5694
    },
    {
      "epoch": 1.3321248976727869,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 5695
    },
    {
      "epoch": 1.332358788445796,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8568,
      "step": 5696
    },
    {
      "epoch": 1.3325926792188048,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 5697
    },
    {
      "epoch": 1.3328265699918138,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2587,
      "step": 5698
    },
    {
      "epoch": 1.3330604607648229,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 5699
    },
    {
      "epoch": 1.3332943515378317,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 5700
    },
    {
      "epoch": 1.3332943515378317,
      "eval_runtime": 4.6686,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 5700
    },
    {
      "epoch": 1.3335282423108408,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 5701
    },
    {
      "epoch": 1.3337621330838498,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 5702
    },
    {
      "epoch": 1.333996023856859,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 5703
    },
    {
      "epoch": 1.334229914629868,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 5704
    },
    {
      "epoch": 1.3344638054028768,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 5705
    },
    {
      "epoch": 1.3346976961758859,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.206,
      "step": 5706
    },
    {
      "epoch": 1.334931586948895,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6539,
      "step": 5707
    },
    {
      "epoch": 1.3351654777219037,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 5708
    },
    {
      "epoch": 1.3353993684949128,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 5709
    },
    {
      "epoch": 1.3356332592679219,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6466,
      "step": 5710
    },
    {
      "epoch": 1.335867150040931,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 5711
    },
    {
      "epoch": 1.33610104081394,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 5712
    },
    {
      "epoch": 1.3363349315869488,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 5713
    },
    {
      "epoch": 1.3365688223599579,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 5714
    },
    {
      "epoch": 1.336802713132967,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 5715
    },
    {
      "epoch": 1.3370366039059758,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 5716
    },
    {
      "epoch": 1.3372704946789848,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 5717
    },
    {
      "epoch": 1.337504385451994,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 5718
    },
    {
      "epoch": 1.337738276225003,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 5719
    },
    {
      "epoch": 1.337972166998012,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 5720
    },
    {
      "epoch": 1.3382060577710209,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6361,
      "step": 5721
    },
    {
      "epoch": 1.33843994854403,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 5722
    },
    {
      "epoch": 1.338673839317039,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 5723
    },
    {
      "epoch": 1.3389077300900478,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 5724
    },
    {
      "epoch": 1.3391416208630569,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 5725
    },
    {
      "epoch": 1.339375511636066,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 5726
    },
    {
      "epoch": 1.339609402409075,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 5727
    },
    {
      "epoch": 1.339843293182084,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 5728
    },
    {
      "epoch": 1.340077183955093,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 5729
    },
    {
      "epoch": 1.340311074728102,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6355,
      "step": 5730
    },
    {
      "epoch": 1.340544965501111,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 5731
    },
    {
      "epoch": 1.3407788562741199,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 5732
    },
    {
      "epoch": 1.341012747047129,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 5733
    },
    {
      "epoch": 1.341246637820138,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 5734
    },
    {
      "epoch": 1.341480528593147,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 5735
    },
    {
      "epoch": 1.341714419366156,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 5736
    },
    {
      "epoch": 1.341948310139165,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.403,
      "step": 5737
    },
    {
      "epoch": 1.342182200912174,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 5738
    },
    {
      "epoch": 1.342416091685183,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 5739
    },
    {
      "epoch": 1.342649982458192,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 5740
    },
    {
      "epoch": 1.342883873231201,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 5741
    },
    {
      "epoch": 1.34311776400421,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 5742
    },
    {
      "epoch": 1.343351654777219,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 5743
    },
    {
      "epoch": 1.3435855455502281,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 5744
    },
    {
      "epoch": 1.343819436323237,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 5745
    },
    {
      "epoch": 1.344053327096246,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 5746
    },
    {
      "epoch": 1.344287217869255,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 5747
    },
    {
      "epoch": 1.344521108642264,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 5748
    },
    {
      "epoch": 1.344754999415273,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 5749
    },
    {
      "epoch": 1.344988890188282,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.6135,
      "step": 5750
    },
    {
      "epoch": 1.345222780961291,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 5751
    },
    {
      "epoch": 1.3454566717343002,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 5752
    },
    {
      "epoch": 1.345690562507309,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 5753
    },
    {
      "epoch": 1.345924453280318,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 5754
    },
    {
      "epoch": 1.3461583440533271,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 5755
    },
    {
      "epoch": 1.346392234826336,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 5756
    },
    {
      "epoch": 1.346626125599345,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8043,
      "step": 5757
    },
    {
      "epoch": 1.346860016372354,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 5758
    },
    {
      "epoch": 1.3470939071453631,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 5759
    },
    {
      "epoch": 1.3473277979183722,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 5760
    },
    {
      "epoch": 1.347561688691381,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 5761
    },
    {
      "epoch": 1.34779557946439,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 5762
    },
    {
      "epoch": 1.3480294702373992,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 5763
    },
    {
      "epoch": 1.348263361010408,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 5764
    },
    {
      "epoch": 1.348497251783417,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 5765
    },
    {
      "epoch": 1.3487311425564261,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 5766
    },
    {
      "epoch": 1.3489650333294352,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 5767
    },
    {
      "epoch": 1.3491989241024442,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 5768
    },
    {
      "epoch": 1.349432814875453,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 5769
    },
    {
      "epoch": 1.3496667056484621,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 5770
    },
    {
      "epoch": 1.3499005964214712,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 5771
    },
    {
      "epoch": 1.3501344871944803,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 5772
    },
    {
      "epoch": 1.3503683779674893,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.658,
      "step": 5773
    },
    {
      "epoch": 1.3506022687404982,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 5774
    },
    {
      "epoch": 1.3508361595135072,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 5775
    },
    {
      "epoch": 1.3510700502865163,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 5776
    },
    {
      "epoch": 1.3513039410595251,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 5777
    },
    {
      "epoch": 1.3515378318325342,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 5778
    },
    {
      "epoch": 1.3517717226055432,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 5779
    },
    {
      "epoch": 1.3520056133785523,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7008,
      "step": 5780
    },
    {
      "epoch": 1.3522395041515614,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 5781
    },
    {
      "epoch": 1.3524733949245702,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 5782
    },
    {
      "epoch": 1.3527072856975793,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 5783
    },
    {
      "epoch": 1.3529411764705883,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 5784
    },
    {
      "epoch": 1.3531750672435972,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 5785
    },
    {
      "epoch": 1.3534089580166062,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 5786
    },
    {
      "epoch": 1.3536428487896153,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7083,
      "step": 5787
    },
    {
      "epoch": 1.3538767395626243,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.3537,
      "step": 5788
    },
    {
      "epoch": 1.3541106303356334,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 5789
    },
    {
      "epoch": 1.3543445211086422,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 5790
    },
    {
      "epoch": 1.3545784118816513,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 5791
    },
    {
      "epoch": 1.3548123026546603,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8476,
      "step": 5792
    },
    {
      "epoch": 1.3550461934276692,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 5793
    },
    {
      "epoch": 1.3552800842006782,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 5794
    },
    {
      "epoch": 1.3555139749736873,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 5795
    },
    {
      "epoch": 1.3557478657466964,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 5796
    },
    {
      "epoch": 1.3559817565197054,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 5797
    },
    {
      "epoch": 1.3562156472927143,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 5798
    },
    {
      "epoch": 1.3564495380657233,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 5799
    },
    {
      "epoch": 1.3566834288387324,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 5800
    },
    {
      "epoch": 1.3566834288387324,
      "eval_runtime": 4.6546,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 5800
    },
    {
      "epoch": 1.3569173196117412,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 5801
    },
    {
      "epoch": 1.3571512103847503,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 5802
    },
    {
      "epoch": 1.3573851011577593,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 5803
    },
    {
      "epoch": 1.3576189919307684,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 5804
    },
    {
      "epoch": 1.3578528827037775,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 5805
    },
    {
      "epoch": 1.3580867734767863,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 5806
    },
    {
      "epoch": 1.3583206642497954,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 5807
    },
    {
      "epoch": 1.3585545550228044,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 5808
    },
    {
      "epoch": 1.3587884457958133,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3923,
      "step": 5809
    },
    {
      "epoch": 1.3590223365688223,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 5810
    },
    {
      "epoch": 1.3592562273418314,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7283,
      "step": 5811
    },
    {
      "epoch": 1.3594901181148404,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 5812
    },
    {
      "epoch": 1.3597240088878495,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 5813
    },
    {
      "epoch": 1.3599578996608583,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 5814
    },
    {
      "epoch": 1.3601917904338674,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 5815
    },
    {
      "epoch": 1.3604256812068765,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 5816
    },
    {
      "epoch": 1.3606595719798853,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 5817
    },
    {
      "epoch": 1.3608934627528944,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 5818
    },
    {
      "epoch": 1.3611273535259034,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 5819
    },
    {
      "epoch": 1.3613612442989125,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 5820
    },
    {
      "epoch": 1.3615951350719215,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9094,
      "step": 5821
    },
    {
      "epoch": 1.3618290258449304,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 5822
    },
    {
      "epoch": 1.3620629166179394,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 5823
    },
    {
      "epoch": 1.3622968073909485,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 5824
    },
    {
      "epoch": 1.3625306981639573,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 5825
    },
    {
      "epoch": 1.3627645889369664,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 5826
    },
    {
      "epoch": 1.3629984797099755,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7664,
      "step": 5827
    },
    {
      "epoch": 1.3632323704829845,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.708,
      "step": 5828
    },
    {
      "epoch": 1.3634662612559936,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 5829
    },
    {
      "epoch": 1.3637001520290024,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 5830
    },
    {
      "epoch": 1.3639340428020115,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 5831
    },
    {
      "epoch": 1.3641679335750205,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.78,
      "step": 5832
    },
    {
      "epoch": 1.3644018243480294,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 5833
    },
    {
      "epoch": 1.3646357151210384,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 5834
    },
    {
      "epoch": 1.3648696058940475,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6534,
      "step": 5835
    },
    {
      "epoch": 1.3651034966670565,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 5836
    },
    {
      "epoch": 1.3653373874400656,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7426,
      "step": 5837
    },
    {
      "epoch": 1.3655712782130744,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 5838
    },
    {
      "epoch": 1.3658051689860835,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 5839
    },
    {
      "epoch": 1.3660390597590926,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 5840
    },
    {
      "epoch": 1.3662729505321014,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 5841
    },
    {
      "epoch": 1.3665068413051105,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 5842
    },
    {
      "epoch": 1.3667407320781195,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 5843
    },
    {
      "epoch": 1.3669746228511286,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 5844
    },
    {
      "epoch": 1.3672085136241376,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 5845
    },
    {
      "epoch": 1.3674424043971465,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 5846
    },
    {
      "epoch": 1.3676762951701555,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 5847
    },
    {
      "epoch": 1.3679101859431646,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 5848
    },
    {
      "epoch": 1.3681440767161734,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 5849
    },
    {
      "epoch": 1.3683779674891825,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 5850
    },
    {
      "epoch": 1.3686118582621916,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 5851
    },
    {
      "epoch": 1.3688457490352006,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 5852
    },
    {
      "epoch": 1.3690796398082097,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 5853
    },
    {
      "epoch": 1.3693135305812185,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 5854
    },
    {
      "epoch": 1.3695474213542276,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 5855
    },
    {
      "epoch": 1.3697813121272366,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2903,
      "step": 5856
    },
    {
      "epoch": 1.3700152029002455,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 5857
    },
    {
      "epoch": 1.3702490936732545,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 5858
    },
    {
      "epoch": 1.3704829844462636,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 5859
    },
    {
      "epoch": 1.3707168752192727,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 5860
    },
    {
      "epoch": 1.3709507659922817,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 5861
    },
    {
      "epoch": 1.3711846567652906,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 5862
    },
    {
      "epoch": 1.3714185475382996,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6498,
      "step": 5863
    },
    {
      "epoch": 1.3716524383113087,
      "grad_norm": 10.3125,
      "learning_rate": 3e-05,
      "loss": 2.5569,
      "step": 5864
    },
    {
      "epoch": 1.3718863290843175,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 5865
    },
    {
      "epoch": 1.3721202198573266,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 5866
    },
    {
      "epoch": 1.3723541106303356,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 5867
    },
    {
      "epoch": 1.3725880014033447,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 5868
    },
    {
      "epoch": 1.3728218921763538,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 5869
    },
    {
      "epoch": 1.3730557829493626,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8306,
      "step": 5870
    },
    {
      "epoch": 1.3732896737223717,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 5871
    },
    {
      "epoch": 1.3735235644953807,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 5872
    },
    {
      "epoch": 1.3737574552683895,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 5873
    },
    {
      "epoch": 1.3739913460413986,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 5874
    },
    {
      "epoch": 1.3742252368144077,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 5875
    },
    {
      "epoch": 1.3744591275874167,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 5876
    },
    {
      "epoch": 1.3746930183604258,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 5877
    },
    {
      "epoch": 1.3749269091334346,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.844,
      "step": 5878
    },
    {
      "epoch": 1.3751607999064437,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 5879
    },
    {
      "epoch": 1.3753946906794527,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7273,
      "step": 5880
    },
    {
      "epoch": 1.3756285814524616,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 5881
    },
    {
      "epoch": 1.3758624722254706,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6894,
      "step": 5882
    },
    {
      "epoch": 1.3760963629984797,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 5883
    },
    {
      "epoch": 1.3763302537714888,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 5884
    },
    {
      "epoch": 1.3765641445444978,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2911,
      "step": 5885
    },
    {
      "epoch": 1.3767980353175067,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 5886
    },
    {
      "epoch": 1.3770319260905157,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 5887
    },
    {
      "epoch": 1.3772658168635248,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 5888
    },
    {
      "epoch": 1.3774997076365336,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 5889
    },
    {
      "epoch": 1.3777335984095427,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 5890
    },
    {
      "epoch": 1.3779674891825517,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 5891
    },
    {
      "epoch": 1.3782013799555608,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 5892
    },
    {
      "epoch": 1.3784352707285699,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 5893
    },
    {
      "epoch": 1.3786691615015787,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6609,
      "step": 5894
    },
    {
      "epoch": 1.3789030522745878,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 5895
    },
    {
      "epoch": 1.3791369430475968,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3508,
      "step": 5896
    },
    {
      "epoch": 1.3793708338206057,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 5897
    },
    {
      "epoch": 1.3796047245936147,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 5898
    },
    {
      "epoch": 1.3798386153666238,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 5899
    },
    {
      "epoch": 1.3800725061396328,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 5900
    },
    {
      "epoch": 1.3800725061396328,
      "eval_runtime": 4.5884,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 5900
    },
    {
      "epoch": 1.380306396912642,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 5901
    },
    {
      "epoch": 1.3805402876856507,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.7447,
      "step": 5902
    },
    {
      "epoch": 1.3807741784586598,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 5903
    },
    {
      "epoch": 1.3810080692316689,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 5904
    },
    {
      "epoch": 1.3812419600046777,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 5905
    },
    {
      "epoch": 1.3814758507776868,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 5906
    },
    {
      "epoch": 1.3817097415506958,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 5907
    },
    {
      "epoch": 1.3819436323237049,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 5908
    },
    {
      "epoch": 1.382177523096714,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 5909
    },
    {
      "epoch": 1.3824114138697228,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 5910
    },
    {
      "epoch": 1.3826453046427318,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8203,
      "step": 5911
    },
    {
      "epoch": 1.382879195415741,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 5912
    },
    {
      "epoch": 1.3831130861887497,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 5913
    },
    {
      "epoch": 1.3833469769617588,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 5914
    },
    {
      "epoch": 1.3835808677347678,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 5915
    },
    {
      "epoch": 1.383814758507777,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 5916
    },
    {
      "epoch": 1.384048649280786,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.448,
      "step": 5917
    },
    {
      "epoch": 1.3842825400537948,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 5918
    },
    {
      "epoch": 1.3845164308268039,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.71,
      "step": 5919
    },
    {
      "epoch": 1.384750321599813,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 5920
    },
    {
      "epoch": 1.3849842123728218,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 5921
    },
    {
      "epoch": 1.3852181031458308,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 5922
    },
    {
      "epoch": 1.3854519939188399,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8183,
      "step": 5923
    },
    {
      "epoch": 1.385685884691849,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 5924
    },
    {
      "epoch": 1.385919775464858,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 5925
    },
    {
      "epoch": 1.3861536662378668,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 5926
    },
    {
      "epoch": 1.386387557010876,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 5927
    },
    {
      "epoch": 1.386621447783885,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6087,
      "step": 5928
    },
    {
      "epoch": 1.3868553385568938,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 5929
    },
    {
      "epoch": 1.3870892293299029,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6243,
      "step": 5930
    },
    {
      "epoch": 1.387323120102912,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7258,
      "step": 5931
    },
    {
      "epoch": 1.387557010875921,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7758,
      "step": 5932
    },
    {
      "epoch": 1.38779090164893,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 5933
    },
    {
      "epoch": 1.3880247924219389,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3042,
      "step": 5934
    },
    {
      "epoch": 1.388258683194948,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 5935
    },
    {
      "epoch": 1.388492573967957,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 5936
    },
    {
      "epoch": 1.3887264647409658,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 5937
    },
    {
      "epoch": 1.388960355513975,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1317,
      "step": 5938
    },
    {
      "epoch": 1.389194246286984,
      "grad_norm": 8.5,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 5939
    },
    {
      "epoch": 1.389428137059993,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 5940
    },
    {
      "epoch": 1.389662027833002,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 5941
    },
    {
      "epoch": 1.389895918606011,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 5942
    },
    {
      "epoch": 1.39012980937902,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9605,
      "step": 5943
    },
    {
      "epoch": 1.390363700152029,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.663,
      "step": 5944
    },
    {
      "epoch": 1.3905975909250379,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 5945
    },
    {
      "epoch": 1.390831481698047,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8363,
      "step": 5946
    },
    {
      "epoch": 1.391065372471056,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 5947
    },
    {
      "epoch": 1.391299263244065,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 5948
    },
    {
      "epoch": 1.3915331540170741,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 5949
    },
    {
      "epoch": 1.391767044790083,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 5950
    },
    {
      "epoch": 1.392000935563092,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 5951
    },
    {
      "epoch": 1.392234826336101,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 5952
    },
    {
      "epoch": 1.39246871710911,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7426,
      "step": 5953
    },
    {
      "epoch": 1.392702607882119,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7141,
      "step": 5954
    },
    {
      "epoch": 1.392936498655128,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 5955
    },
    {
      "epoch": 1.393170389428137,
      "grad_norm": 14.0,
      "learning_rate": 3e-05,
      "loss": 2.2911,
      "step": 5956
    },
    {
      "epoch": 1.3934042802011462,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 5957
    },
    {
      "epoch": 1.393638170974155,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 5958
    },
    {
      "epoch": 1.393872061747164,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 5959
    },
    {
      "epoch": 1.394105952520173,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 5960
    },
    {
      "epoch": 1.394339843293182,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8156,
      "step": 5961
    },
    {
      "epoch": 1.394573734066191,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 5962
    },
    {
      "epoch": 1.3948076248392,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 5963
    },
    {
      "epoch": 1.3950415156122091,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7067,
      "step": 5964
    },
    {
      "epoch": 1.3952754063852182,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2531,
      "step": 5965
    },
    {
      "epoch": 1.395509297158227,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6997,
      "step": 5966
    },
    {
      "epoch": 1.395743187931236,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 5967
    },
    {
      "epoch": 1.3959770787042451,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 5968
    },
    {
      "epoch": 1.3962109694772542,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 5969
    },
    {
      "epoch": 1.3964448602502633,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 5970
    },
    {
      "epoch": 1.396678751023272,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7482,
      "step": 5971
    },
    {
      "epoch": 1.3969126417962812,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 5972
    },
    {
      "epoch": 1.3971465325692902,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 5973
    },
    {
      "epoch": 1.397380423342299,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 5974
    },
    {
      "epoch": 1.3976143141153081,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 5975
    },
    {
      "epoch": 1.3978482048883172,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 5976
    },
    {
      "epoch": 1.3980820956613262,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 5977
    },
    {
      "epoch": 1.3983159864343353,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1038,
      "step": 5978
    },
    {
      "epoch": 1.3985498772073441,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 5979
    },
    {
      "epoch": 1.3987837679803532,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 5980
    },
    {
      "epoch": 1.3990176587533623,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 5981
    },
    {
      "epoch": 1.399251549526371,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7656,
      "step": 5982
    },
    {
      "epoch": 1.3994854402993802,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 5983
    },
    {
      "epoch": 1.3997193310723892,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8359,
      "step": 5984
    },
    {
      "epoch": 1.3999532218453983,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 5985
    },
    {
      "epoch": 1.4001871126184073,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 5986
    },
    {
      "epoch": 1.4004210033914162,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 5987
    },
    {
      "epoch": 1.4006548941644252,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 5988
    },
    {
      "epoch": 1.4008887849374343,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2047,
      "step": 5989
    },
    {
      "epoch": 1.4011226757104431,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6299,
      "step": 5990
    },
    {
      "epoch": 1.4013565664834522,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 5991
    },
    {
      "epoch": 1.4015904572564613,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 5992
    },
    {
      "epoch": 1.4018243480294703,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6218,
      "step": 5993
    },
    {
      "epoch": 1.4020582388024794,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 5994
    },
    {
      "epoch": 1.4022921295754882,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 5995
    },
    {
      "epoch": 1.4025260203484973,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 5996
    },
    {
      "epoch": 1.4027599111215063,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 5997
    },
    {
      "epoch": 1.4029938018945152,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 5998
    },
    {
      "epoch": 1.4032276926675242,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 5999
    },
    {
      "epoch": 1.4034615834405333,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7254,
      "step": 6000
    },
    {
      "epoch": 1.4034615834405333,
      "eval_runtime": 4.6162,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 6000
    },
    {
      "epoch": 1.4036954742135423,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 6001
    },
    {
      "epoch": 1.4039293649865514,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 6002
    },
    {
      "epoch": 1.4041632557595602,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 6003
    },
    {
      "epoch": 1.4043971465325693,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 6004
    },
    {
      "epoch": 1.4046310373055784,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 6005
    },
    {
      "epoch": 1.4048649280785872,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8815,
      "step": 6006
    },
    {
      "epoch": 1.4050988188515963,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 6007
    },
    {
      "epoch": 1.4053327096246053,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 6008
    },
    {
      "epoch": 1.4055666003976144,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 6009
    },
    {
      "epoch": 1.4058004911706234,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 6010
    },
    {
      "epoch": 1.4060343819436323,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 6011
    },
    {
      "epoch": 1.4062682727166413,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2691,
      "step": 6012
    },
    {
      "epoch": 1.4065021634896504,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 6013
    },
    {
      "epoch": 1.4067360542626592,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 6014
    },
    {
      "epoch": 1.4069699450356683,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 6015
    },
    {
      "epoch": 1.4072038358086774,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 6016
    },
    {
      "epoch": 1.4074377265816864,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7003,
      "step": 6017
    },
    {
      "epoch": 1.4076716173546955,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 6018
    },
    {
      "epoch": 1.4079055081277043,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 6019
    },
    {
      "epoch": 1.4081393989007134,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 6020
    },
    {
      "epoch": 1.4083732896737224,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7495,
      "step": 6021
    },
    {
      "epoch": 1.4086071804467313,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 6022
    },
    {
      "epoch": 1.4088410712197403,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 6023
    },
    {
      "epoch": 1.4090749619927494,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 6024
    },
    {
      "epoch": 1.4093088527657585,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 6025
    },
    {
      "epoch": 1.4095427435387675,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 6026
    },
    {
      "epoch": 1.4097766343117764,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 6027
    },
    {
      "epoch": 1.4100105250847854,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 6028
    },
    {
      "epoch": 1.4102444158577945,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 6029
    },
    {
      "epoch": 1.4104783066308033,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 6030
    },
    {
      "epoch": 1.4107121974038124,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7604,
      "step": 6031
    },
    {
      "epoch": 1.4109460881768214,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 6032
    },
    {
      "epoch": 1.4111799789498305,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 6033
    },
    {
      "epoch": 1.4114138697228396,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 6034
    },
    {
      "epoch": 1.4116477604958484,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 6035
    },
    {
      "epoch": 1.4118816512688575,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 6036
    },
    {
      "epoch": 1.4121155420418665,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 6037
    },
    {
      "epoch": 1.4123494328148753,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 6038
    },
    {
      "epoch": 1.4125833235878844,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 6039
    },
    {
      "epoch": 1.4128172143608935,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 6040
    },
    {
      "epoch": 1.4130511051339025,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 6041
    },
    {
      "epoch": 1.4132849959069116,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 6042
    },
    {
      "epoch": 1.4135188866799204,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 6043
    },
    {
      "epoch": 1.4137527774529295,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6818,
      "step": 6044
    },
    {
      "epoch": 1.4139866682259385,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 6045
    },
    {
      "epoch": 1.4142205589989474,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 6046
    },
    {
      "epoch": 1.4144544497719564,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 6047
    },
    {
      "epoch": 1.4146883405449655,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 6048
    },
    {
      "epoch": 1.4149222313179746,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 6049
    },
    {
      "epoch": 1.4151561220909836,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 6050
    },
    {
      "epoch": 1.4153900128639925,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 6051
    },
    {
      "epoch": 1.4156239036370015,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6716,
      "step": 6052
    },
    {
      "epoch": 1.4158577944100106,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 6053
    },
    {
      "epoch": 1.4160916851830194,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 6054
    },
    {
      "epoch": 1.4163255759560285,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 6055
    },
    {
      "epoch": 1.4165594667290375,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 6056
    },
    {
      "epoch": 1.4167933575020466,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 6057
    },
    {
      "epoch": 1.4170272482750557,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5044,
      "step": 6058
    },
    {
      "epoch": 1.4172611390480645,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 6059
    },
    {
      "epoch": 1.4174950298210736,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 6060
    },
    {
      "epoch": 1.4177289205940826,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 6061
    },
    {
      "epoch": 1.4179628113670915,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 6062
    },
    {
      "epoch": 1.4181967021401005,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 6063
    },
    {
      "epoch": 1.4184305929131096,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2331,
      "step": 6064
    },
    {
      "epoch": 1.4186644836861186,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 6065
    },
    {
      "epoch": 1.4188983744591277,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 6066
    },
    {
      "epoch": 1.4191322652321365,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 6067
    },
    {
      "epoch": 1.4193661560051456,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.3029,
      "step": 6068
    },
    {
      "epoch": 1.4196000467781547,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 6069
    },
    {
      "epoch": 1.4198339375511635,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8361,
      "step": 6070
    },
    {
      "epoch": 1.4200678283241726,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 6071
    },
    {
      "epoch": 1.4203017190971816,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 6072
    },
    {
      "epoch": 1.4205356098701907,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 6073
    },
    {
      "epoch": 1.4207695006431997,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 6074
    },
    {
      "epoch": 1.4210033914162086,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8089,
      "step": 6075
    },
    {
      "epoch": 1.4212372821892176,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 6076
    },
    {
      "epoch": 1.4214711729622267,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6214,
      "step": 6077
    },
    {
      "epoch": 1.4217050637352355,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 6078
    },
    {
      "epoch": 1.4219389545082446,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 6079
    },
    {
      "epoch": 1.4221728452812537,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 6080
    },
    {
      "epoch": 1.4224067360542627,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 6081
    },
    {
      "epoch": 1.4226406268272718,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 6082
    },
    {
      "epoch": 1.4228745176002806,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 6083
    },
    {
      "epoch": 1.4231084083732897,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 6084
    },
    {
      "epoch": 1.4233422991462987,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 6085
    },
    {
      "epoch": 1.4235761899193076,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 6086
    },
    {
      "epoch": 1.4238100806923166,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 6087
    },
    {
      "epoch": 1.4240439714653257,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 6088
    },
    {
      "epoch": 1.4242778622383347,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 6089
    },
    {
      "epoch": 1.4245117530113438,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 6090
    },
    {
      "epoch": 1.4247456437843526,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 6091
    },
    {
      "epoch": 1.4249795345573617,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3447,
      "step": 6092
    },
    {
      "epoch": 1.4252134253303708,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 6093
    },
    {
      "epoch": 1.4254473161033796,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 6094
    },
    {
      "epoch": 1.4256812068763887,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 6095
    },
    {
      "epoch": 1.4259150976493977,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 6096
    },
    {
      "epoch": 1.4261489884224068,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 6097
    },
    {
      "epoch": 1.4263828791954158,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 6098
    },
    {
      "epoch": 1.4266167699684247,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 6099
    },
    {
      "epoch": 1.4268506607414337,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 6100
    },
    {
      "epoch": 1.4268506607414337,
      "eval_runtime": 4.7122,
      "eval_samples_per_second": 0.212,
      "eval_steps_per_second": 0.212,
      "step": 6100
    },
    {
      "epoch": 1.4270845515144428,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 6101
    },
    {
      "epoch": 1.4273184422874516,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 6102
    },
    {
      "epoch": 1.4275523330604607,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.731,
      "step": 6103
    },
    {
      "epoch": 1.4277862238334698,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8377,
      "step": 6104
    },
    {
      "epoch": 1.4280201146064788,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 6105
    },
    {
      "epoch": 1.4282540053794879,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 6106
    },
    {
      "epoch": 1.4284878961524967,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 6107
    },
    {
      "epoch": 1.4287217869255058,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 6108
    },
    {
      "epoch": 1.4289556776985148,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 6109
    },
    {
      "epoch": 1.4291895684715237,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 6110
    },
    {
      "epoch": 1.4294234592445327,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7699,
      "step": 6111
    },
    {
      "epoch": 1.4296573500175418,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 6112
    },
    {
      "epoch": 1.4298912407905509,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 6113
    },
    {
      "epoch": 1.43012513156356,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 6114
    },
    {
      "epoch": 1.4303590223365688,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 6115
    },
    {
      "epoch": 1.4305929131095778,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 6116
    },
    {
      "epoch": 1.4308268038825869,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6862,
      "step": 6117
    },
    {
      "epoch": 1.4310606946555957,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 6118
    },
    {
      "epoch": 1.4312945854286048,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 6119
    },
    {
      "epoch": 1.4315284762016138,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.154,
      "step": 6120
    },
    {
      "epoch": 1.431762366974623,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 6121
    },
    {
      "epoch": 1.431996257747632,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 6122
    },
    {
      "epoch": 1.4322301485206408,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 6123
    },
    {
      "epoch": 1.4324640392936498,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 6124
    },
    {
      "epoch": 1.432697930066659,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 6125
    },
    {
      "epoch": 1.4329318208396677,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 6126
    },
    {
      "epoch": 1.4331657116126768,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 6127
    },
    {
      "epoch": 1.4333996023856859,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 6128
    },
    {
      "epoch": 1.433633493158695,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4308,
      "step": 6129
    },
    {
      "epoch": 1.433867383931704,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 6130
    },
    {
      "epoch": 1.4341012747047128,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7461,
      "step": 6131
    },
    {
      "epoch": 1.4343351654777219,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 6132
    },
    {
      "epoch": 1.434569056250731,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 6133
    },
    {
      "epoch": 1.4348029470237398,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2156,
      "step": 6134
    },
    {
      "epoch": 1.4350368377967488,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 6135
    },
    {
      "epoch": 1.435270728569758,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 6136
    },
    {
      "epoch": 1.435504619342767,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 6137
    },
    {
      "epoch": 1.435738510115776,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.4993,
      "step": 6138
    },
    {
      "epoch": 1.4359724008887849,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 6139
    },
    {
      "epoch": 1.436206291661794,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 6140
    },
    {
      "epoch": 1.436440182434803,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 6141
    },
    {
      "epoch": 1.4366740732078118,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 6142
    },
    {
      "epoch": 1.4369079639808209,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 6143
    },
    {
      "epoch": 1.43714185475383,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 6144
    },
    {
      "epoch": 1.437375745526839,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7526,
      "step": 6145
    },
    {
      "epoch": 1.437609636299848,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 6146
    },
    {
      "epoch": 1.437843527072857,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 6147
    },
    {
      "epoch": 1.438077417845866,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 6148
    },
    {
      "epoch": 1.438311308618875,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 6149
    },
    {
      "epoch": 1.4385451993918839,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 6150
    },
    {
      "epoch": 1.438779090164893,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7028,
      "step": 6151
    },
    {
      "epoch": 1.439012980937902,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 6152
    },
    {
      "epoch": 1.439246871710911,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 6153
    },
    {
      "epoch": 1.43948076248392,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 6154
    },
    {
      "epoch": 1.439714653256929,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 6155
    },
    {
      "epoch": 1.439948544029938,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9196,
      "step": 6156
    },
    {
      "epoch": 1.440182434802947,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2839,
      "step": 6157
    },
    {
      "epoch": 1.440416325575956,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3129,
      "step": 6158
    },
    {
      "epoch": 1.440650216348965,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7019,
      "step": 6159
    },
    {
      "epoch": 1.440884107121974,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7671,
      "step": 6160
    },
    {
      "epoch": 1.441117997894983,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 6161
    },
    {
      "epoch": 1.4413518886679921,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5917,
      "step": 6162
    },
    {
      "epoch": 1.441585779441001,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7252,
      "step": 6163
    },
    {
      "epoch": 1.44181967021401,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 6164
    },
    {
      "epoch": 1.442053560987019,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 6165
    },
    {
      "epoch": 1.4422874517600281,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 6166
    },
    {
      "epoch": 1.4425213425330372,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 6167
    },
    {
      "epoch": 1.442755233306046,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 6168
    },
    {
      "epoch": 1.442989124079055,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 6169
    },
    {
      "epoch": 1.4432230148520642,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 6170
    },
    {
      "epoch": 1.443456905625073,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 6171
    },
    {
      "epoch": 1.443690796398082,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 6172
    },
    {
      "epoch": 1.4439246871710911,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 6173
    },
    {
      "epoch": 1.4441585779441002,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 6174
    },
    {
      "epoch": 1.4443924687171092,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 6175
    },
    {
      "epoch": 1.444626359490118,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 6176
    },
    {
      "epoch": 1.4448602502631271,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 6177
    },
    {
      "epoch": 1.4450941410361362,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6479,
      "step": 6178
    },
    {
      "epoch": 1.445328031809145,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 6179
    },
    {
      "epoch": 1.445561922582154,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1819,
      "step": 6180
    },
    {
      "epoch": 1.4457958133551632,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 6181
    },
    {
      "epoch": 1.4460297041281722,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 6182
    },
    {
      "epoch": 1.4462635949011813,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 6183
    },
    {
      "epoch": 1.4464974856741901,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5345,
      "step": 6184
    },
    {
      "epoch": 1.4467313764471992,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 6185
    },
    {
      "epoch": 1.4469652672202082,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6918,
      "step": 6186
    },
    {
      "epoch": 1.447199157993217,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 6187
    },
    {
      "epoch": 1.4474330487662261,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7368,
      "step": 6188
    },
    {
      "epoch": 1.4476669395392352,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8141,
      "step": 6189
    },
    {
      "epoch": 1.4479008303122443,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 6190
    },
    {
      "epoch": 1.4481347210852533,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 6191
    },
    {
      "epoch": 1.4483686118582622,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 6192
    },
    {
      "epoch": 1.4486025026312712,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 6193
    },
    {
      "epoch": 1.4488363934042803,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1283,
      "step": 6194
    },
    {
      "epoch": 1.4490702841772891,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4148,
      "step": 6195
    },
    {
      "epoch": 1.4493041749502982,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 6196
    },
    {
      "epoch": 1.4495380657233072,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 6197
    },
    {
      "epoch": 1.4497719564963163,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 6198
    },
    {
      "epoch": 1.4500058472693254,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 6199
    },
    {
      "epoch": 1.4502397380423342,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 6200
    },
    {
      "epoch": 1.4502397380423342,
      "eval_runtime": 4.5915,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 6200
    },
    {
      "epoch": 1.4504736288153433,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 6201
    },
    {
      "epoch": 1.4507075195883523,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8122,
      "step": 6202
    },
    {
      "epoch": 1.4509414103613612,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 6203
    },
    {
      "epoch": 1.4511753011343702,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 6204
    },
    {
      "epoch": 1.4514091919073793,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 6205
    },
    {
      "epoch": 1.4516430826803883,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3365,
      "step": 6206
    },
    {
      "epoch": 1.4518769734533974,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 6207
    },
    {
      "epoch": 1.4521108642264062,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 6208
    },
    {
      "epoch": 1.4523447549994153,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 6209
    },
    {
      "epoch": 1.4525786457724243,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 6210
    },
    {
      "epoch": 1.4528125365454332,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 6211
    },
    {
      "epoch": 1.4530464273184422,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 6212
    },
    {
      "epoch": 1.4532803180914513,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 6213
    },
    {
      "epoch": 1.4535142088644604,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 6214
    },
    {
      "epoch": 1.4537480996374694,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 6215
    },
    {
      "epoch": 1.4539819904104783,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 6216
    },
    {
      "epoch": 1.4542158811834873,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 6217
    },
    {
      "epoch": 1.4544497719564964,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 6218
    },
    {
      "epoch": 1.4546836627295052,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6961,
      "step": 6219
    },
    {
      "epoch": 1.4549175535025143,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 6220
    },
    {
      "epoch": 1.4551514442755233,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 6221
    },
    {
      "epoch": 1.4553853350485324,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 6222
    },
    {
      "epoch": 1.4556192258215415,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 6223
    },
    {
      "epoch": 1.4558531165945503,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 6224
    },
    {
      "epoch": 1.4560870073675594,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 6225
    },
    {
      "epoch": 1.4563208981405684,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 6226
    },
    {
      "epoch": 1.4565547889135773,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9322,
      "step": 6227
    },
    {
      "epoch": 1.4567886796865863,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6635,
      "step": 6228
    },
    {
      "epoch": 1.4570225704595954,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 6229
    },
    {
      "epoch": 1.4572564612326044,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 6230
    },
    {
      "epoch": 1.4574903520056135,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 6231
    },
    {
      "epoch": 1.4577242427786223,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 6232
    },
    {
      "epoch": 1.4579581335516314,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7405,
      "step": 6233
    },
    {
      "epoch": 1.4581920243246405,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 6234
    },
    {
      "epoch": 1.4584259150976493,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 6235
    },
    {
      "epoch": 1.4586598058706584,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 6236
    },
    {
      "epoch": 1.4588936966436674,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2561,
      "step": 6237
    },
    {
      "epoch": 1.4591275874166765,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7228,
      "step": 6238
    },
    {
      "epoch": 1.4593614781896855,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 6239
    },
    {
      "epoch": 1.4595953689626944,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 6240
    },
    {
      "epoch": 1.4598292597357034,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 6241
    },
    {
      "epoch": 1.4600631505087125,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 6242
    },
    {
      "epoch": 1.4602970412817213,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 6243
    },
    {
      "epoch": 1.4605309320547304,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 6244
    },
    {
      "epoch": 1.4607648228277395,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 6245
    },
    {
      "epoch": 1.4609987136007485,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 6246
    },
    {
      "epoch": 1.4612326043737576,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 6247
    },
    {
      "epoch": 1.4614664951467664,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 6248
    },
    {
      "epoch": 1.4617003859197755,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 6249
    },
    {
      "epoch": 1.4619342766927845,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 6250
    },
    {
      "epoch": 1.4621681674657934,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 6251
    },
    {
      "epoch": 1.4624020582388024,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 6252
    },
    {
      "epoch": 1.4626359490118115,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 6253
    },
    {
      "epoch": 1.4628698397848205,
      "grad_norm": 7.90625,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 6254
    },
    {
      "epoch": 1.4631037305578296,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 6255
    },
    {
      "epoch": 1.4633376213308384,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 6256
    },
    {
      "epoch": 1.4635715121038475,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 6257
    },
    {
      "epoch": 1.4638054028768566,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 6258
    },
    {
      "epoch": 1.4640392936498654,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 6259
    },
    {
      "epoch": 1.4642731844228745,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 6260
    },
    {
      "epoch": 1.4645070751958835,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 6261
    },
    {
      "epoch": 1.4647409659688926,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 6262
    },
    {
      "epoch": 1.4649748567419016,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 6263
    },
    {
      "epoch": 1.4652087475149105,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 6264
    },
    {
      "epoch": 1.4654426382879195,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 6265
    },
    {
      "epoch": 1.4656765290609286,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7252,
      "step": 6266
    },
    {
      "epoch": 1.4659104198339374,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 6267
    },
    {
      "epoch": 1.4661443106069465,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 6268
    },
    {
      "epoch": 1.4663782013799556,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6828,
      "step": 6269
    },
    {
      "epoch": 1.4666120921529646,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 6270
    },
    {
      "epoch": 1.4668459829259737,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6886,
      "step": 6271
    },
    {
      "epoch": 1.4670798736989825,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 6272
    },
    {
      "epoch": 1.4673137644719916,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8318,
      "step": 6273
    },
    {
      "epoch": 1.4675476552450006,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 6274
    },
    {
      "epoch": 1.4677815460180095,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 6275
    },
    {
      "epoch": 1.4680154367910185,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 6276
    },
    {
      "epoch": 1.4682493275640276,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 6277
    },
    {
      "epoch": 1.4684832183370367,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3645,
      "step": 6278
    },
    {
      "epoch": 1.4687171091100457,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 6279
    },
    {
      "epoch": 1.4689509998830546,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 6280
    },
    {
      "epoch": 1.4691848906560636,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 6281
    },
    {
      "epoch": 1.4694187814290727,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 6282
    },
    {
      "epoch": 1.4696526722020815,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 6283
    },
    {
      "epoch": 1.4698865629750906,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 6284
    },
    {
      "epoch": 1.4701204537480996,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 6285
    },
    {
      "epoch": 1.4703543445211087,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 6286
    },
    {
      "epoch": 1.4705882352941178,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7061,
      "step": 6287
    },
    {
      "epoch": 1.4708221260671266,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 6288
    },
    {
      "epoch": 1.4710560168401356,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7384,
      "step": 6289
    },
    {
      "epoch": 1.4712899076131447,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 6290
    },
    {
      "epoch": 1.4715237983861535,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 6291
    },
    {
      "epoch": 1.4717576891591626,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 6292
    },
    {
      "epoch": 1.4719915799321717,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 6293
    },
    {
      "epoch": 1.4722254707051807,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 6294
    },
    {
      "epoch": 1.4724593614781898,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 6295
    },
    {
      "epoch": 1.4726932522511986,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 6296
    },
    {
      "epoch": 1.4729271430242077,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 6297
    },
    {
      "epoch": 1.4731610337972167,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 6298
    },
    {
      "epoch": 1.4733949245702256,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 6299
    },
    {
      "epoch": 1.4736288153432346,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 6300
    },
    {
      "epoch": 1.4736288153432346,
      "eval_runtime": 4.6029,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 6300
    },
    {
      "epoch": 1.4738627061162437,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 6301
    },
    {
      "epoch": 1.4740965968892528,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 6302
    },
    {
      "epoch": 1.4743304876622618,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 6303
    },
    {
      "epoch": 1.4745643784352707,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 6304
    },
    {
      "epoch": 1.4747982692082797,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 6305
    },
    {
      "epoch": 1.4750321599812888,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 6306
    },
    {
      "epoch": 1.4752660507542976,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6908,
      "step": 6307
    },
    {
      "epoch": 1.4754999415273067,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 6308
    },
    {
      "epoch": 1.4757338323003157,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 6309
    },
    {
      "epoch": 1.4759677230733248,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 6310
    },
    {
      "epoch": 1.4762016138463339,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6885,
      "step": 6311
    },
    {
      "epoch": 1.4764355046193427,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 6312
    },
    {
      "epoch": 1.4766693953923518,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 6313
    },
    {
      "epoch": 1.4769032861653608,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 6314
    },
    {
      "epoch": 1.4771371769383697,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 6315
    },
    {
      "epoch": 1.4773710677113787,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7666,
      "step": 6316
    },
    {
      "epoch": 1.4776049584843878,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 6317
    },
    {
      "epoch": 1.4778388492573968,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 6318
    },
    {
      "epoch": 1.478072740030406,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8047,
      "step": 6319
    },
    {
      "epoch": 1.4783066308034147,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 6320
    },
    {
      "epoch": 1.4785405215764238,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 6321
    },
    {
      "epoch": 1.4787744123494329,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 6322
    },
    {
      "epoch": 1.4790083031224417,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 6323
    },
    {
      "epoch": 1.4792421938954508,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6355,
      "step": 6324
    },
    {
      "epoch": 1.4794760846684598,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 6325
    },
    {
      "epoch": 1.4797099754414689,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.5552,
      "step": 6326
    },
    {
      "epoch": 1.479943866214478,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 6327
    },
    {
      "epoch": 1.4801777569874868,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3434,
      "step": 6328
    },
    {
      "epoch": 1.4804116477604958,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2097,
      "step": 6329
    },
    {
      "epoch": 1.480645538533505,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 6330
    },
    {
      "epoch": 1.4808794293065137,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 6331
    },
    {
      "epoch": 1.4811133200795228,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 6332
    },
    {
      "epoch": 1.4813472108525318,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7631,
      "step": 6333
    },
    {
      "epoch": 1.481581101625541,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 6334
    },
    {
      "epoch": 1.48181499239855,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 6335
    },
    {
      "epoch": 1.4820488831715588,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 6336
    },
    {
      "epoch": 1.4822827739445679,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 6337
    },
    {
      "epoch": 1.482516664717577,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 6338
    },
    {
      "epoch": 1.4827505554905858,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 6339
    },
    {
      "epoch": 1.4829844462635948,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 6340
    },
    {
      "epoch": 1.4832183370366039,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 6341
    },
    {
      "epoch": 1.483452227809613,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 6342
    },
    {
      "epoch": 1.483686118582622,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 6343
    },
    {
      "epoch": 1.4839200093556308,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 6344
    },
    {
      "epoch": 1.48415390012864,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 6345
    },
    {
      "epoch": 1.484387790901649,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 6346
    },
    {
      "epoch": 1.4846216816746578,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3546,
      "step": 6347
    },
    {
      "epoch": 1.4848555724476669,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7212,
      "step": 6348
    },
    {
      "epoch": 1.485089463220676,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 6349
    },
    {
      "epoch": 1.485323353993685,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 6350
    },
    {
      "epoch": 1.485557244766694,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 6351
    },
    {
      "epoch": 1.4857911355397029,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 6352
    },
    {
      "epoch": 1.486025026312712,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 6353
    },
    {
      "epoch": 1.486258917085721,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 6354
    },
    {
      "epoch": 1.4864928078587298,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 6355
    },
    {
      "epoch": 1.486726698631739,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 6356
    },
    {
      "epoch": 1.486960589404748,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 6357
    },
    {
      "epoch": 1.487194480177757,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 6358
    },
    {
      "epoch": 1.487428370950766,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 6359
    },
    {
      "epoch": 1.487662261723775,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6256,
      "step": 6360
    },
    {
      "epoch": 1.487896152496784,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 6361
    },
    {
      "epoch": 1.488130043269793,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 6362
    },
    {
      "epoch": 1.4883639340428019,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7667,
      "step": 6363
    },
    {
      "epoch": 1.4885978248158112,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 6364
    },
    {
      "epoch": 1.48883171558882,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 6365
    },
    {
      "epoch": 1.489065606361829,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 6366
    },
    {
      "epoch": 1.4892994971348381,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2737,
      "step": 6367
    },
    {
      "epoch": 1.489533387907847,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 6368
    },
    {
      "epoch": 1.489767278680856,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6986,
      "step": 6369
    },
    {
      "epoch": 1.490001169453865,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 6370
    },
    {
      "epoch": 1.4902350602268741,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 6371
    },
    {
      "epoch": 1.4904689509998832,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 6372
    },
    {
      "epoch": 1.490702841772892,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 6373
    },
    {
      "epoch": 1.490936732545901,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 6374
    },
    {
      "epoch": 1.4911706233189101,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 6375
    },
    {
      "epoch": 1.491404514091919,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 6376
    },
    {
      "epoch": 1.491638404864928,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 6377
    },
    {
      "epoch": 1.491872295637937,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 6378
    },
    {
      "epoch": 1.4921061864109462,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2469,
      "step": 6379
    },
    {
      "epoch": 1.4923400771839552,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2502,
      "step": 6380
    },
    {
      "epoch": 1.492573967956964,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 6381
    },
    {
      "epoch": 1.4928078587299731,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 6382
    },
    {
      "epoch": 1.4930417495029822,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 6383
    },
    {
      "epoch": 1.493275640275991,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 6384
    },
    {
      "epoch": 1.493509531049,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 6385
    },
    {
      "epoch": 1.4937434218220091,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 6386
    },
    {
      "epoch": 1.4939773125950182,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 6387
    },
    {
      "epoch": 1.4942112033680273,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 6388
    },
    {
      "epoch": 1.494445094141036,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 6389
    },
    {
      "epoch": 1.4946789849140452,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 6390
    },
    {
      "epoch": 1.4949128756870542,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2832,
      "step": 6391
    },
    {
      "epoch": 1.495146766460063,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 6392
    },
    {
      "epoch": 1.4953806572330721,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 6393
    },
    {
      "epoch": 1.4956145480060812,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 6394
    },
    {
      "epoch": 1.4958484387790902,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 6395
    },
    {
      "epoch": 1.4960823295520993,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 6396
    },
    {
      "epoch": 1.4963162203251081,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 6397
    },
    {
      "epoch": 1.4965501110981172,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 6398
    },
    {
      "epoch": 1.4967840018711263,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7883,
      "step": 6399
    },
    {
      "epoch": 1.497017892644135,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 6400
    },
    {
      "epoch": 1.497017892644135,
      "eval_runtime": 4.5805,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 6400
    },
    {
      "epoch": 1.4972517834171442,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7729,
      "step": 6401
    },
    {
      "epoch": 1.4974856741901532,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 6402
    },
    {
      "epoch": 1.4977195649631623,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 6403
    },
    {
      "epoch": 1.4979534557361713,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4001,
      "step": 6404
    },
    {
      "epoch": 1.4981873465091802,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 6405
    },
    {
      "epoch": 1.4984212372821892,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 6406
    },
    {
      "epoch": 1.4986551280551983,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 6407
    },
    {
      "epoch": 1.4988890188282071,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 6408
    },
    {
      "epoch": 1.4991229096012162,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6234,
      "step": 6409
    },
    {
      "epoch": 1.4993568003742253,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 6410
    },
    {
      "epoch": 1.4995906911472343,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7485,
      "step": 6411
    },
    {
      "epoch": 1.4998245819202434,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 6412
    },
    {
      "epoch": 1.5000584726932522,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 6413
    },
    {
      "epoch": 1.5002923634662613,
      "grad_norm": 9.3125,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 6414
    },
    {
      "epoch": 1.5005262542392703,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 6415
    },
    {
      "epoch": 1.5007601450122792,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 6416
    },
    {
      "epoch": 1.5009940357852882,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 6417
    },
    {
      "epoch": 1.5012279265582973,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8549,
      "step": 6418
    },
    {
      "epoch": 1.5014618173313061,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 6419
    },
    {
      "epoch": 1.5016957081043154,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 6420
    },
    {
      "epoch": 1.5019295988773242,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 6421
    },
    {
      "epoch": 1.5021634896503333,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7908,
      "step": 6422
    },
    {
      "epoch": 1.5023973804233424,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 6423
    },
    {
      "epoch": 1.5026312711963512,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 6424
    },
    {
      "epoch": 1.5028651619693603,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 6425
    },
    {
      "epoch": 1.5030990527423693,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 6426
    },
    {
      "epoch": 1.5033329435153782,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 6427
    },
    {
      "epoch": 1.5035668342883874,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 6428
    },
    {
      "epoch": 1.5038007250613963,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 6429
    },
    {
      "epoch": 1.5040346158344053,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 6430
    },
    {
      "epoch": 1.5042685066074144,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 6431
    },
    {
      "epoch": 1.5045023973804232,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 6432
    },
    {
      "epoch": 1.5047362881534323,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 6433
    },
    {
      "epoch": 1.5049701789264414,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 6434
    },
    {
      "epoch": 1.5052040696994502,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 6435
    },
    {
      "epoch": 1.5054379604724595,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 6436
    },
    {
      "epoch": 1.5056718512454683,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 6437
    },
    {
      "epoch": 1.5059057420184774,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2188,
      "step": 6438
    },
    {
      "epoch": 1.5061396327914864,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9811,
      "step": 6439
    },
    {
      "epoch": 1.5063735235644953,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 6440
    },
    {
      "epoch": 1.5066074143375043,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 6441
    },
    {
      "epoch": 1.5068413051105134,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 6442
    },
    {
      "epoch": 1.5070751958835222,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7503,
      "step": 6443
    },
    {
      "epoch": 1.5073090866565315,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9758,
      "step": 6444
    },
    {
      "epoch": 1.5075429774295404,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 6445
    },
    {
      "epoch": 1.5077768682025494,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 6446
    },
    {
      "epoch": 1.5080107589755585,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 6447
    },
    {
      "epoch": 1.5082446497485673,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 6448
    },
    {
      "epoch": 1.5084785405215764,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 6449
    },
    {
      "epoch": 1.5087124312945854,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 6450
    },
    {
      "epoch": 1.5089463220675943,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7345,
      "step": 6451
    },
    {
      "epoch": 1.5091802128406036,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 6452
    },
    {
      "epoch": 1.5094141036136124,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 6453
    },
    {
      "epoch": 1.5096479943866215,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 6454
    },
    {
      "epoch": 1.5098818851596305,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 6455
    },
    {
      "epoch": 1.5101157759326393,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 6456
    },
    {
      "epoch": 1.5103496667056484,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 6457
    },
    {
      "epoch": 1.5105835574786575,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7377,
      "step": 6458
    },
    {
      "epoch": 1.5108174482516663,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2074,
      "step": 6459
    },
    {
      "epoch": 1.5110513390246756,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 6460
    },
    {
      "epoch": 1.5112852297976844,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 6461
    },
    {
      "epoch": 1.5115191205706935,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7642,
      "step": 6462
    },
    {
      "epoch": 1.5117530113437025,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 6463
    },
    {
      "epoch": 1.5119869021167114,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 6464
    },
    {
      "epoch": 1.5122207928897207,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 6465
    },
    {
      "epoch": 1.5124546836627295,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 6466
    },
    {
      "epoch": 1.5126885744357386,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 6467
    },
    {
      "epoch": 1.5129224652087476,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6318,
      "step": 6468
    },
    {
      "epoch": 1.5131563559817565,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 6469
    },
    {
      "epoch": 1.5133902467547655,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.6117,
      "step": 6470
    },
    {
      "epoch": 1.5136241375277746,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7576,
      "step": 6471
    },
    {
      "epoch": 1.5138580283007834,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 6472
    },
    {
      "epoch": 1.5140919190737927,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 6473
    },
    {
      "epoch": 1.5143258098468015,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 6474
    },
    {
      "epoch": 1.5145597006198106,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 6475
    },
    {
      "epoch": 1.5147935913928197,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 6476
    },
    {
      "epoch": 1.5150274821658285,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 6477
    },
    {
      "epoch": 1.5152613729388376,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 6478
    },
    {
      "epoch": 1.5154952637118466,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 6479
    },
    {
      "epoch": 1.5157291544848555,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 6480
    },
    {
      "epoch": 1.5159630452578647,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 6481
    },
    {
      "epoch": 1.5161969360308736,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 6482
    },
    {
      "epoch": 1.5164308268038826,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1864,
      "step": 6483
    },
    {
      "epoch": 1.5166647175768917,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 6484
    },
    {
      "epoch": 1.5168986083499005,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 6485
    },
    {
      "epoch": 1.5171324991229096,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 6486
    },
    {
      "epoch": 1.5173663898959187,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 6487
    },
    {
      "epoch": 1.5176002806689275,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 6488
    },
    {
      "epoch": 1.5178341714419368,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 6489
    },
    {
      "epoch": 1.5180680622149456,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6962,
      "step": 6490
    },
    {
      "epoch": 1.5183019529879547,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 6491
    },
    {
      "epoch": 1.5185358437609637,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 6492
    },
    {
      "epoch": 1.5187697345339726,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 6493
    },
    {
      "epoch": 1.5190036253069816,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2224,
      "step": 6494
    },
    {
      "epoch": 1.5192375160799907,
      "grad_norm": 9.4375,
      "learning_rate": 3e-05,
      "loss": 2.4841,
      "step": 6495
    },
    {
      "epoch": 1.5194714068529995,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 6496
    },
    {
      "epoch": 1.5197052976260088,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 6497
    },
    {
      "epoch": 1.5199391883990176,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6082,
      "step": 6498
    },
    {
      "epoch": 1.5201730791720267,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 6499
    },
    {
      "epoch": 1.5204069699450358,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 6500
    },
    {
      "epoch": 1.5204069699450358,
      "eval_runtime": 4.5975,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 6500
    },
    {
      "epoch": 1.5206408607180446,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 6501
    },
    {
      "epoch": 1.5208747514910537,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 6502
    },
    {
      "epoch": 1.5211086422640627,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2897,
      "step": 6503
    },
    {
      "epoch": 1.5213425330370716,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7093,
      "step": 6504
    },
    {
      "epoch": 1.5215764238100808,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1954,
      "step": 6505
    },
    {
      "epoch": 1.5218103145830897,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 6506
    },
    {
      "epoch": 1.5220442053560987,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 6507
    },
    {
      "epoch": 1.5222780961291078,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 6508
    },
    {
      "epoch": 1.5225119869021166,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 6509
    },
    {
      "epoch": 1.5227458776751257,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 6510
    },
    {
      "epoch": 1.5229797684481348,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 6511
    },
    {
      "epoch": 1.5232136592211436,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5654,
      "step": 6512
    },
    {
      "epoch": 1.5234475499941529,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 6513
    },
    {
      "epoch": 1.5236814407671617,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 6514
    },
    {
      "epoch": 1.5239153315401708,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 6515
    },
    {
      "epoch": 1.5241492223131798,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9017,
      "step": 6516
    },
    {
      "epoch": 1.5243831130861887,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 6517
    },
    {
      "epoch": 1.5246170038591977,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 6518
    },
    {
      "epoch": 1.5248508946322068,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6902,
      "step": 6519
    },
    {
      "epoch": 1.5250847854052156,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 6520
    },
    {
      "epoch": 1.525318676178225,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 6521
    },
    {
      "epoch": 1.5255525669512338,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7238,
      "step": 6522
    },
    {
      "epoch": 1.5257864577242428,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 6523
    },
    {
      "epoch": 1.5260203484972519,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 6524
    },
    {
      "epoch": 1.5262542392702607,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 6525
    },
    {
      "epoch": 1.5264881300432698,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 6526
    },
    {
      "epoch": 1.5267220208162788,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 6527
    },
    {
      "epoch": 1.5269559115892877,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 6528
    },
    {
      "epoch": 1.527189802362297,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 6529
    },
    {
      "epoch": 1.5274236931353058,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 6530
    },
    {
      "epoch": 1.5276575839083149,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 6531
    },
    {
      "epoch": 1.527891474681324,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 6532
    },
    {
      "epoch": 1.5281253654543328,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 6533
    },
    {
      "epoch": 1.5283592562273418,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 6534
    },
    {
      "epoch": 1.5285931470003509,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 6535
    },
    {
      "epoch": 1.5288270377733597,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.788,
      "step": 6536
    },
    {
      "epoch": 1.529060928546369,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 6537
    },
    {
      "epoch": 1.5292948193193778,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 6538
    },
    {
      "epoch": 1.529528710092387,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 6539
    },
    {
      "epoch": 1.529762600865396,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 6540
    },
    {
      "epoch": 1.5299964916384048,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5884,
      "step": 6541
    },
    {
      "epoch": 1.5302303824114138,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 6542
    },
    {
      "epoch": 1.530464273184423,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 6543
    },
    {
      "epoch": 1.5306981639574317,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9244,
      "step": 6544
    },
    {
      "epoch": 1.530932054730441,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 6545
    },
    {
      "epoch": 1.5311659455034499,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 6546
    },
    {
      "epoch": 1.531399836276459,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 6547
    },
    {
      "epoch": 1.531633727049468,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 6548
    },
    {
      "epoch": 1.5318676178224768,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2485,
      "step": 6549
    },
    {
      "epoch": 1.5321015085954859,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 6550
    },
    {
      "epoch": 1.532335399368495,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 6551
    },
    {
      "epoch": 1.5325692901415038,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 6552
    },
    {
      "epoch": 1.532803180914513,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 6553
    },
    {
      "epoch": 1.533037071687522,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 6554
    },
    {
      "epoch": 1.533270962460531,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 6555
    },
    {
      "epoch": 1.53350485323354,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 6556
    },
    {
      "epoch": 1.5337387440065489,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 6557
    },
    {
      "epoch": 1.533972634779558,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 6558
    },
    {
      "epoch": 1.534206525552567,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 6559
    },
    {
      "epoch": 1.5344404163255758,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 6560
    },
    {
      "epoch": 1.534674307098585,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 6561
    },
    {
      "epoch": 1.534908197871594,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 6562
    },
    {
      "epoch": 1.535142088644603,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 6563
    },
    {
      "epoch": 1.535375979417612,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6818,
      "step": 6564
    },
    {
      "epoch": 1.535609870190621,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1765,
      "step": 6565
    },
    {
      "epoch": 1.53584376096363,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 6566
    },
    {
      "epoch": 1.536077651736639,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 6567
    },
    {
      "epoch": 1.5363115425096479,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 6568
    },
    {
      "epoch": 1.5365454332826571,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 6569
    },
    {
      "epoch": 1.536779324055666,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 6570
    },
    {
      "epoch": 1.537013214828675,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 6571
    },
    {
      "epoch": 1.537247105601684,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 6572
    },
    {
      "epoch": 1.537480996374693,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 6573
    },
    {
      "epoch": 1.537714887147702,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 6574
    },
    {
      "epoch": 1.537948777920711,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6643,
      "step": 6575
    },
    {
      "epoch": 1.53818266869372,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 6576
    },
    {
      "epoch": 1.5384165594667292,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 6577
    },
    {
      "epoch": 1.538650450239738,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2123,
      "step": 6578
    },
    {
      "epoch": 1.538884341012747,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 6579
    },
    {
      "epoch": 1.5391182317857561,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6157,
      "step": 6580
    },
    {
      "epoch": 1.539352122558765,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6285,
      "step": 6581
    },
    {
      "epoch": 1.539586013331774,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 6582
    },
    {
      "epoch": 1.539819904104783,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 6583
    },
    {
      "epoch": 1.540053794877792,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 6584
    },
    {
      "epoch": 1.5402876856508012,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7284,
      "step": 6585
    },
    {
      "epoch": 1.54052157642381,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 6586
    },
    {
      "epoch": 1.540755467196819,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 6587
    },
    {
      "epoch": 1.5409893579698282,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 6588
    },
    {
      "epoch": 1.541223248742837,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 6589
    },
    {
      "epoch": 1.541457139515846,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 6590
    },
    {
      "epoch": 1.5416910302888551,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 6591
    },
    {
      "epoch": 1.541924921061864,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 6592
    },
    {
      "epoch": 1.5421588118348732,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 6593
    },
    {
      "epoch": 1.542392702607882,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 6594
    },
    {
      "epoch": 1.5426265933808911,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 6595
    },
    {
      "epoch": 1.5428604841539002,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 6596
    },
    {
      "epoch": 1.543094374926909,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 6597
    },
    {
      "epoch": 1.543328265699918,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7543,
      "step": 6598
    },
    {
      "epoch": 1.5435621564729272,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 6599
    },
    {
      "epoch": 1.543796047245936,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 6600
    },
    {
      "epoch": 1.543796047245936,
      "eval_runtime": 4.6262,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 6600
    },
    {
      "epoch": 1.5440299380189453,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 6601
    },
    {
      "epoch": 1.5442638287919541,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 6602
    },
    {
      "epoch": 1.5444977195649632,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 6603
    },
    {
      "epoch": 1.5447316103379722,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2029,
      "step": 6604
    },
    {
      "epoch": 1.544965501110981,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 6605
    },
    {
      "epoch": 1.5451993918839901,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6844,
      "step": 6606
    },
    {
      "epoch": 1.5454332826569992,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 6607
    },
    {
      "epoch": 1.545667173430008,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 6608
    },
    {
      "epoch": 1.5459010642030173,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 6609
    },
    {
      "epoch": 1.5461349549760262,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 6610
    },
    {
      "epoch": 1.5463688457490352,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 6611
    },
    {
      "epoch": 1.5466027365220443,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 6612
    },
    {
      "epoch": 1.5468366272950531,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2451,
      "step": 6613
    },
    {
      "epoch": 1.5470705180680622,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8038,
      "step": 6614
    },
    {
      "epoch": 1.5473044088410712,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 6615
    },
    {
      "epoch": 1.54753829961408,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.779,
      "step": 6616
    },
    {
      "epoch": 1.5477721903870894,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 6617
    },
    {
      "epoch": 1.5480060811600982,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7328,
      "step": 6618
    },
    {
      "epoch": 1.5482399719331073,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 6619
    },
    {
      "epoch": 1.5484738627061163,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 6620
    },
    {
      "epoch": 1.5487077534791251,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0799,
      "step": 6621
    },
    {
      "epoch": 1.5489416442521342,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 6622
    },
    {
      "epoch": 1.5491755350251433,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 6623
    },
    {
      "epoch": 1.549409425798152,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3687,
      "step": 6624
    },
    {
      "epoch": 1.5496433165711614,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7998,
      "step": 6625
    },
    {
      "epoch": 1.5498772073441702,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 6626
    },
    {
      "epoch": 1.5501110981171793,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 6627
    },
    {
      "epoch": 1.5503449888901883,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 6628
    },
    {
      "epoch": 1.5505788796631972,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 6629
    },
    {
      "epoch": 1.5508127704362062,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 6630
    },
    {
      "epoch": 1.5510466612092153,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 6631
    },
    {
      "epoch": 1.5512805519822241,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2334,
      "step": 6632
    },
    {
      "epoch": 1.5515144427552334,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 6633
    },
    {
      "epoch": 1.5517483335282423,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 6634
    },
    {
      "epoch": 1.5519822243012513,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 6635
    },
    {
      "epoch": 1.5522161150742604,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8377,
      "step": 6636
    },
    {
      "epoch": 1.5524500058472692,
      "grad_norm": 15.5625,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 6637
    },
    {
      "epoch": 1.5526838966202783,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 6638
    },
    {
      "epoch": 1.5529177873932873,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 6639
    },
    {
      "epoch": 1.5531516781662962,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3734,
      "step": 6640
    },
    {
      "epoch": 1.5533855689393055,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 6641
    },
    {
      "epoch": 1.5536194597123143,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 6642
    },
    {
      "epoch": 1.5538533504853234,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 6643
    },
    {
      "epoch": 1.5540872412583324,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 6644
    },
    {
      "epoch": 1.5543211320313413,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 6645
    },
    {
      "epoch": 1.5545550228043503,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7269,
      "step": 6646
    },
    {
      "epoch": 1.5547889135773594,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 6647
    },
    {
      "epoch": 1.5550228043503682,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 6648
    },
    {
      "epoch": 1.5552566951233775,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8838,
      "step": 6649
    },
    {
      "epoch": 1.5554905858963863,
      "grad_norm": 9.1875,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 6650
    },
    {
      "epoch": 1.5557244766693954,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6745,
      "step": 6651
    },
    {
      "epoch": 1.5559583674424045,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 6652
    },
    {
      "epoch": 1.5561922582154133,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 6653
    },
    {
      "epoch": 1.5564261489884224,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 6654
    },
    {
      "epoch": 1.5566600397614314,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7214,
      "step": 6655
    },
    {
      "epoch": 1.5568939305344403,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 6656
    },
    {
      "epoch": 1.5571278213074495,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 6657
    },
    {
      "epoch": 1.5573617120804584,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5116,
      "step": 6658
    },
    {
      "epoch": 1.5575956028534674,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1483,
      "step": 6659
    },
    {
      "epoch": 1.5578294936264765,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 6660
    },
    {
      "epoch": 1.5580633843994853,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7317,
      "step": 6661
    },
    {
      "epoch": 1.5582972751724946,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 6662
    },
    {
      "epoch": 1.5585311659455034,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 6663
    },
    {
      "epoch": 1.5587650567185125,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 6664
    },
    {
      "epoch": 1.5589989474915216,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 6665
    },
    {
      "epoch": 1.5592328382645304,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 6666
    },
    {
      "epoch": 1.5594667290375395,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 6667
    },
    {
      "epoch": 1.5597006198105485,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 6668
    },
    {
      "epoch": 1.5599345105835574,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 6669
    },
    {
      "epoch": 1.5601684013565666,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 6670
    },
    {
      "epoch": 1.5604022921295755,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 6671
    },
    {
      "epoch": 1.5606361829025845,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 6672
    },
    {
      "epoch": 1.5608700736755936,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 6673
    },
    {
      "epoch": 1.5611039644486024,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8315,
      "step": 6674
    },
    {
      "epoch": 1.5613378552216115,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 6675
    },
    {
      "epoch": 1.5615717459946206,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 6676
    },
    {
      "epoch": 1.5618056367676294,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 6677
    },
    {
      "epoch": 1.5620395275406387,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 6678
    },
    {
      "epoch": 1.5622734183136475,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5691,
      "step": 6679
    },
    {
      "epoch": 1.5625073090866566,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 6680
    },
    {
      "epoch": 1.5627411998596656,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 6681
    },
    {
      "epoch": 1.5629750906326745,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 6682
    },
    {
      "epoch": 1.5632089814056835,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 6683
    },
    {
      "epoch": 1.5634428721786926,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 6684
    },
    {
      "epoch": 1.5636767629517014,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 6685
    },
    {
      "epoch": 1.5639106537247107,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 6686
    },
    {
      "epoch": 1.5641445444977196,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 6687
    },
    {
      "epoch": 1.5643784352707286,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 6688
    },
    {
      "epoch": 1.5646123260437377,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 6689
    },
    {
      "epoch": 1.5648462168167465,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 6690
    },
    {
      "epoch": 1.5650801075897556,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 6691
    },
    {
      "epoch": 1.5653139983627646,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.5983,
      "step": 6692
    },
    {
      "epoch": 1.5655478891357735,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.65,
      "step": 6693
    },
    {
      "epoch": 1.5657817799087828,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 6694
    },
    {
      "epoch": 1.5660156706817916,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.69,
      "step": 6695
    },
    {
      "epoch": 1.5662495614548007,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 6696
    },
    {
      "epoch": 1.5664834522278097,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 6697
    },
    {
      "epoch": 1.5667173430008186,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 6698
    },
    {
      "epoch": 1.5669512337738276,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 6699
    },
    {
      "epoch": 1.5671851245468367,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 6700
    },
    {
      "epoch": 1.5671851245468367,
      "eval_runtime": 4.6793,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 6700
    },
    {
      "epoch": 1.5674190153198455,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 6701
    },
    {
      "epoch": 1.5676529060928548,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 6702
    },
    {
      "epoch": 1.5678867968658636,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 6703
    },
    {
      "epoch": 1.5681206876388727,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 6704
    },
    {
      "epoch": 1.5683545784118818,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7764,
      "step": 6705
    },
    {
      "epoch": 1.5685884691848906,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 6706
    },
    {
      "epoch": 1.5688223599578996,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 6707
    },
    {
      "epoch": 1.5690562507309087,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 6708
    },
    {
      "epoch": 1.5692901415039175,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 6709
    },
    {
      "epoch": 1.5695240322769268,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 6710
    },
    {
      "epoch": 1.5697579230499357,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 6711
    },
    {
      "epoch": 1.5699918138229447,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2845,
      "step": 6712
    },
    {
      "epoch": 1.5702257045959538,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 6713
    },
    {
      "epoch": 1.5704595953689626,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 6714
    },
    {
      "epoch": 1.5706934861419717,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 6715
    },
    {
      "epoch": 1.5709273769149807,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 6716
    },
    {
      "epoch": 1.5711612676879896,
      "grad_norm": 9.1875,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 6717
    },
    {
      "epoch": 1.5713951584609989,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 6718
    },
    {
      "epoch": 1.5716290492340077,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 6719
    },
    {
      "epoch": 1.5718629400070168,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 6720
    },
    {
      "epoch": 1.5720968307800258,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 6721
    },
    {
      "epoch": 1.5723307215530347,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 6722
    },
    {
      "epoch": 1.5725646123260437,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 6723
    },
    {
      "epoch": 1.5727985030990528,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 6724
    },
    {
      "epoch": 1.5730323938720616,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 6725
    },
    {
      "epoch": 1.573266284645071,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 6726
    },
    {
      "epoch": 1.5735001754180797,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 6727
    },
    {
      "epoch": 1.5737340661910888,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 6728
    },
    {
      "epoch": 1.5739679569640979,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 6729
    },
    {
      "epoch": 1.5742018477371067,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 6730
    },
    {
      "epoch": 1.5744357385101158,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 6731
    },
    {
      "epoch": 1.5746696292831248,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 6732
    },
    {
      "epoch": 1.5749035200561337,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 6733
    },
    {
      "epoch": 1.575137410829143,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 6734
    },
    {
      "epoch": 1.5753713016021518,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 6735
    },
    {
      "epoch": 1.5756051923751608,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 6736
    },
    {
      "epoch": 1.57583908314817,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7616,
      "step": 6737
    },
    {
      "epoch": 1.5760729739211787,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 6738
    },
    {
      "epoch": 1.5763068646941878,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 6739
    },
    {
      "epoch": 1.5765407554671969,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 6740
    },
    {
      "epoch": 1.5767746462402057,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 6741
    },
    {
      "epoch": 1.577008537013215,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 6742
    },
    {
      "epoch": 1.5772424277862238,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6211,
      "step": 6743
    },
    {
      "epoch": 1.5774763185592329,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 6744
    },
    {
      "epoch": 1.577710209332242,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 6745
    },
    {
      "epoch": 1.5779441001052508,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 6746
    },
    {
      "epoch": 1.5781779908782598,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 6747
    },
    {
      "epoch": 1.5784118816512689,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.063,
      "step": 6748
    },
    {
      "epoch": 1.5786457724242777,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 6749
    },
    {
      "epoch": 1.578879663197287,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 6750
    },
    {
      "epoch": 1.5791135539702958,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 6751
    },
    {
      "epoch": 1.579347444743305,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 6752
    },
    {
      "epoch": 1.579581335516314,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.2676,
      "step": 6753
    },
    {
      "epoch": 1.5798152262893228,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 6754
    },
    {
      "epoch": 1.5800491170623319,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 6755
    },
    {
      "epoch": 1.580283007835341,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 6756
    },
    {
      "epoch": 1.5805168986083498,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 6757
    },
    {
      "epoch": 1.580750789381359,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.517,
      "step": 6758
    },
    {
      "epoch": 1.5809846801543679,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 6759
    },
    {
      "epoch": 1.581218570927377,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 6760
    },
    {
      "epoch": 1.581452461700386,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 6761
    },
    {
      "epoch": 1.5816863524733948,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6731,
      "step": 6762
    },
    {
      "epoch": 1.581920243246404,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 6763
    },
    {
      "epoch": 1.582154134019413,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 6764
    },
    {
      "epoch": 1.5823880247924218,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 6765
    },
    {
      "epoch": 1.582621915565431,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 6766
    },
    {
      "epoch": 1.58285580633844,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 6767
    },
    {
      "epoch": 1.583089697111449,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.563,
      "step": 6768
    },
    {
      "epoch": 1.583323587884458,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 6769
    },
    {
      "epoch": 1.5835574786574669,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.5934,
      "step": 6770
    },
    {
      "epoch": 1.583791369430476,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 6771
    },
    {
      "epoch": 1.584025260203485,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 6772
    },
    {
      "epoch": 1.5842591509764938,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 6773
    },
    {
      "epoch": 1.5844930417495031,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4055,
      "step": 6774
    },
    {
      "epoch": 1.584726932522512,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 6775
    },
    {
      "epoch": 1.584960823295521,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 6776
    },
    {
      "epoch": 1.58519471406853,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 6777
    },
    {
      "epoch": 1.585428604841539,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 6778
    },
    {
      "epoch": 1.585662495614548,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 6779
    },
    {
      "epoch": 1.585896386387557,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 6780
    },
    {
      "epoch": 1.5861302771605659,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 6781
    },
    {
      "epoch": 1.5863641679335752,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7554,
      "step": 6782
    },
    {
      "epoch": 1.586598058706584,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 6783
    },
    {
      "epoch": 1.586831949479593,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.445,
      "step": 6784
    },
    {
      "epoch": 1.5870658402526021,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 6785
    },
    {
      "epoch": 1.587299731025611,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7802,
      "step": 6786
    },
    {
      "epoch": 1.58753362179862,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 6787
    },
    {
      "epoch": 1.587767512571629,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 6788
    },
    {
      "epoch": 1.588001403344638,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 6789
    },
    {
      "epoch": 1.5882352941176472,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.953,
      "step": 6790
    },
    {
      "epoch": 1.588469184890656,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 6791
    },
    {
      "epoch": 1.588703075663665,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8466,
      "step": 6792
    },
    {
      "epoch": 1.5889369664366741,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 6793
    },
    {
      "epoch": 1.589170857209683,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 6794
    },
    {
      "epoch": 1.589404747982692,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 6795
    },
    {
      "epoch": 1.589638638755701,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 6796
    },
    {
      "epoch": 1.58987252952871,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 6797
    },
    {
      "epoch": 1.5901064203017192,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 6798
    },
    {
      "epoch": 1.590340311074728,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 6799
    },
    {
      "epoch": 1.5905742018477371,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 6800
    },
    {
      "epoch": 1.5905742018477371,
      "eval_runtime": 4.5988,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 6800
    },
    {
      "epoch": 1.5908080926207462,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 6801
    },
    {
      "epoch": 1.591041983393755,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 6802
    },
    {
      "epoch": 1.591275874166764,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 6803
    },
    {
      "epoch": 1.5915097649397731,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 6804
    },
    {
      "epoch": 1.591743655712782,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 6805
    },
    {
      "epoch": 1.5919775464857913,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 6806
    },
    {
      "epoch": 1.5922114372588,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 6807
    },
    {
      "epoch": 1.5924453280318092,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7489,
      "step": 6808
    },
    {
      "epoch": 1.5926792188048182,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 6809
    },
    {
      "epoch": 1.592913109577827,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6312,
      "step": 6810
    },
    {
      "epoch": 1.5931470003508361,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 6811
    },
    {
      "epoch": 1.5933808911238452,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2757,
      "step": 6812
    },
    {
      "epoch": 1.593614781896854,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5282,
      "step": 6813
    },
    {
      "epoch": 1.5938486726698633,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 6814
    },
    {
      "epoch": 1.5940825634428721,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 6815
    },
    {
      "epoch": 1.5943164542158812,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 6816
    },
    {
      "epoch": 1.5945503449888903,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 6817
    },
    {
      "epoch": 1.594784235761899,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 6818
    },
    {
      "epoch": 1.5950181265349082,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.23,
      "step": 6819
    },
    {
      "epoch": 1.5952520173079172,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 6820
    },
    {
      "epoch": 1.595485908080926,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 6821
    },
    {
      "epoch": 1.5957197988539353,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 6822
    },
    {
      "epoch": 1.5959536896269442,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 6823
    },
    {
      "epoch": 1.5961875803999532,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 6824
    },
    {
      "epoch": 1.5964214711729623,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6491,
      "step": 6825
    },
    {
      "epoch": 1.5966553619459711,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.4336,
      "step": 6826
    },
    {
      "epoch": 1.5968892527189802,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 6827
    },
    {
      "epoch": 1.5971231434919893,
      "grad_norm": 7.3125,
      "learning_rate": 3e-05,
      "loss": 2.2867,
      "step": 6828
    },
    {
      "epoch": 1.597357034264998,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 6829
    },
    {
      "epoch": 1.5975909250380074,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 6830
    },
    {
      "epoch": 1.5978248158110162,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 6831
    },
    {
      "epoch": 1.5980587065840253,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 6832
    },
    {
      "epoch": 1.5982925973570343,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 6833
    },
    {
      "epoch": 1.5985264881300432,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3055,
      "step": 6834
    },
    {
      "epoch": 1.5987603789030522,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 6835
    },
    {
      "epoch": 1.5989942696760613,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 6836
    },
    {
      "epoch": 1.5992281604490701,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 6837
    },
    {
      "epoch": 1.5994620512220794,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 6838
    },
    {
      "epoch": 1.5996959419950882,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 6839
    },
    {
      "epoch": 1.5999298327680973,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 6840
    },
    {
      "epoch": 1.6001637235411064,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 6841
    },
    {
      "epoch": 1.6003976143141152,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 6842
    },
    {
      "epoch": 1.6006315050871243,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 6843
    },
    {
      "epoch": 1.6008653958601333,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5104,
      "step": 6844
    },
    {
      "epoch": 1.6010992866331422,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 6845
    },
    {
      "epoch": 1.6013331774061514,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 6846
    },
    {
      "epoch": 1.6015670681791603,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 6847
    },
    {
      "epoch": 1.6018009589521693,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 6848
    },
    {
      "epoch": 1.6020348497251784,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 6849
    },
    {
      "epoch": 1.6022687404981872,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 6850
    },
    {
      "epoch": 1.6025026312711963,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 6851
    },
    {
      "epoch": 1.6027365220442054,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 6852
    },
    {
      "epoch": 1.6029704128172142,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 6853
    },
    {
      "epoch": 1.6032043035902235,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 6854
    },
    {
      "epoch": 1.6034381943632323,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 6855
    },
    {
      "epoch": 1.6036720851362414,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 6856
    },
    {
      "epoch": 1.6039059759092504,
      "grad_norm": 9.8125,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 6857
    },
    {
      "epoch": 1.6041398666822593,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2376,
      "step": 6858
    },
    {
      "epoch": 1.6043737574552683,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 6859
    },
    {
      "epoch": 1.6046076482282774,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 6860
    },
    {
      "epoch": 1.6048415390012865,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 6861
    },
    {
      "epoch": 1.6050754297742955,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 6862
    },
    {
      "epoch": 1.6053093205473044,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 6863
    },
    {
      "epoch": 1.6055432113203134,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7587,
      "step": 6864
    },
    {
      "epoch": 1.6057771020933225,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 6865
    },
    {
      "epoch": 1.6060109928663313,
      "grad_norm": 2.46875,
      "learning_rate": 3e-05,
      "loss": 1.5914,
      "step": 6866
    },
    {
      "epoch": 1.6062448836393406,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 6867
    },
    {
      "epoch": 1.6064787744123494,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 6868
    },
    {
      "epoch": 1.6067126651853585,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 6869
    },
    {
      "epoch": 1.6069465559583676,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 6870
    },
    {
      "epoch": 1.6071804467313764,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7148,
      "step": 6871
    },
    {
      "epoch": 1.6074143375043854,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 6872
    },
    {
      "epoch": 1.6076482282773945,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.6741,
      "step": 6873
    },
    {
      "epoch": 1.6078821190504033,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 6874
    },
    {
      "epoch": 1.6081160098234126,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 6875
    },
    {
      "epoch": 1.6083499005964215,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 6876
    },
    {
      "epoch": 1.6085837913694305,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4236,
      "step": 6877
    },
    {
      "epoch": 1.6088176821424396,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4376,
      "step": 6878
    },
    {
      "epoch": 1.6090515729154484,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 6879
    },
    {
      "epoch": 1.6092854636884575,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 6880
    },
    {
      "epoch": 1.6095193544614665,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4663,
      "step": 6881
    },
    {
      "epoch": 1.6097532452344754,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5434,
      "step": 6882
    },
    {
      "epoch": 1.6099871360074847,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 6883
    },
    {
      "epoch": 1.6102210267804935,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 6884
    },
    {
      "epoch": 1.6104549175535026,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 6885
    },
    {
      "epoch": 1.6106888083265116,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 6886
    },
    {
      "epoch": 1.6109226990995205,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 6887
    },
    {
      "epoch": 1.6111565898725295,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 6888
    },
    {
      "epoch": 1.6113904806455386,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 6889
    },
    {
      "epoch": 1.6116243714185474,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 6890
    },
    {
      "epoch": 1.6118582621915567,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 6891
    },
    {
      "epoch": 1.6120921529645655,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 6892
    },
    {
      "epoch": 1.6123260437375746,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 6893
    },
    {
      "epoch": 1.6125599345105837,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 6894
    },
    {
      "epoch": 1.6127938252835925,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 6895
    },
    {
      "epoch": 1.6130277160566016,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 6896
    },
    {
      "epoch": 1.6132616068296106,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 6897
    },
    {
      "epoch": 1.6134954976026195,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6777,
      "step": 6898
    },
    {
      "epoch": 1.6137293883756287,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 6899
    },
    {
      "epoch": 1.6139632791486376,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 6900
    },
    {
      "epoch": 1.6139632791486376,
      "eval_runtime": 4.6862,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 6900
    },
    {
      "epoch": 1.6141971699216466,
      "grad_norm": 9.1875,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 6901
    },
    {
      "epoch": 1.6144310606946557,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6973,
      "step": 6902
    },
    {
      "epoch": 1.6146649514676645,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 6903
    },
    {
      "epoch": 1.6148988422406736,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 6904
    },
    {
      "epoch": 1.6151327330136827,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6035,
      "step": 6905
    },
    {
      "epoch": 1.6153666237866915,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 6906
    },
    {
      "epoch": 1.6156005145597008,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 6907
    },
    {
      "epoch": 1.6158344053327096,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 6908
    },
    {
      "epoch": 1.6160682961057187,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 6909
    },
    {
      "epoch": 1.6163021868787277,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 6910
    },
    {
      "epoch": 1.6165360776517366,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 6911
    },
    {
      "epoch": 1.6167699684247456,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 6912
    },
    {
      "epoch": 1.6170038591977547,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 6913
    },
    {
      "epoch": 1.6172377499707635,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 6914
    },
    {
      "epoch": 1.6174716407437728,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 6915
    },
    {
      "epoch": 1.6177055315167816,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 6916
    },
    {
      "epoch": 1.6179394222897907,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 6917
    },
    {
      "epoch": 1.6181733130627998,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 6918
    },
    {
      "epoch": 1.6184072038358086,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2033,
      "step": 6919
    },
    {
      "epoch": 1.6186410946088177,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 6920
    },
    {
      "epoch": 1.6188749853818267,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6644,
      "step": 6921
    },
    {
      "epoch": 1.6191088761548356,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 6922
    },
    {
      "epoch": 1.6193427669278448,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 6923
    },
    {
      "epoch": 1.6195766577008537,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 6924
    },
    {
      "epoch": 1.6198105484738627,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 6925
    },
    {
      "epoch": 1.6200444392468718,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 6926
    },
    {
      "epoch": 1.6202783300198806,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 6927
    },
    {
      "epoch": 1.6205122207928897,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 6928
    },
    {
      "epoch": 1.6207461115658988,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 6929
    },
    {
      "epoch": 1.6209800023389076,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 6930
    },
    {
      "epoch": 1.6212138931119169,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 6931
    },
    {
      "epoch": 1.6214477838849257,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 6932
    },
    {
      "epoch": 1.6216816746579348,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 6933
    },
    {
      "epoch": 1.6219155654309438,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 6934
    },
    {
      "epoch": 1.6221494562039527,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 6935
    },
    {
      "epoch": 1.6223833469769617,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 6936
    },
    {
      "epoch": 1.6226172377499708,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6918,
      "step": 6937
    },
    {
      "epoch": 1.6228511285229796,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 6938
    },
    {
      "epoch": 1.623085019295989,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 6939
    },
    {
      "epoch": 1.6233189100689978,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8714,
      "step": 6940
    },
    {
      "epoch": 1.6235528008420068,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 6941
    },
    {
      "epoch": 1.6237866916150159,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2548,
      "step": 6942
    },
    {
      "epoch": 1.6240205823880247,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 6943
    },
    {
      "epoch": 1.6242544731610338,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 6944
    },
    {
      "epoch": 1.6244883639340428,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 6945
    },
    {
      "epoch": 1.6247222547070517,
      "grad_norm": 8.75,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 6946
    },
    {
      "epoch": 1.624956145480061,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 6947
    },
    {
      "epoch": 1.6251900362530698,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 6948
    },
    {
      "epoch": 1.6254239270260789,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 6949
    },
    {
      "epoch": 1.625657817799088,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7395,
      "step": 6950
    },
    {
      "epoch": 1.6258917085720967,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.5227,
      "step": 6951
    },
    {
      "epoch": 1.6261255993451058,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 6952
    },
    {
      "epoch": 1.6263594901181149,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 6953
    },
    {
      "epoch": 1.6265933808911237,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 6954
    },
    {
      "epoch": 1.626827271664133,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 6955
    },
    {
      "epoch": 1.6270611624371418,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 6956
    },
    {
      "epoch": 1.6272950532101509,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 6957
    },
    {
      "epoch": 1.62752894398316,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 6958
    },
    {
      "epoch": 1.6277628347561688,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 6959
    },
    {
      "epoch": 1.6279967255291778,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 6960
    },
    {
      "epoch": 1.628230616302187,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 6961
    },
    {
      "epoch": 1.6284645070751957,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8058,
      "step": 6962
    },
    {
      "epoch": 1.628698397848205,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 6963
    },
    {
      "epoch": 1.6289322886212139,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7255,
      "step": 6964
    },
    {
      "epoch": 1.629166179394223,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 6965
    },
    {
      "epoch": 1.629400070167232,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 6966
    },
    {
      "epoch": 1.6296339609402408,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 6967
    },
    {
      "epoch": 1.6298678517132499,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 6968
    },
    {
      "epoch": 1.630101742486259,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 6969
    },
    {
      "epoch": 1.6303356332592678,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 6970
    },
    {
      "epoch": 1.630569524032277,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.668,
      "step": 6971
    },
    {
      "epoch": 1.630803414805286,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 6972
    },
    {
      "epoch": 1.631037305578295,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2237,
      "step": 6973
    },
    {
      "epoch": 1.631271196351304,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 6974
    },
    {
      "epoch": 1.6315050871243129,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.6283,
      "step": 6975
    },
    {
      "epoch": 1.631738977897322,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 6976
    },
    {
      "epoch": 1.631972868670331,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 6977
    },
    {
      "epoch": 1.6322067594433398,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 6978
    },
    {
      "epoch": 1.632440650216349,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7923,
      "step": 6979
    },
    {
      "epoch": 1.632674540989358,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4528,
      "step": 6980
    },
    {
      "epoch": 1.632908431762367,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 6981
    },
    {
      "epoch": 1.633142322535376,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 6982
    },
    {
      "epoch": 1.633376213308385,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 6983
    },
    {
      "epoch": 1.633610104081394,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6324,
      "step": 6984
    },
    {
      "epoch": 1.633843994854403,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 6985
    },
    {
      "epoch": 1.6340778856274119,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 6986
    },
    {
      "epoch": 1.6343117764004211,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 6987
    },
    {
      "epoch": 1.63454566717343,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 6988
    },
    {
      "epoch": 1.634779557946439,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 6989
    },
    {
      "epoch": 1.635013448719448,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 6990
    },
    {
      "epoch": 1.635247339492457,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8701,
      "step": 6991
    },
    {
      "epoch": 1.635481230265466,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 6992
    },
    {
      "epoch": 1.635715121038475,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2736,
      "step": 6993
    },
    {
      "epoch": 1.6359490118114839,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 6994
    },
    {
      "epoch": 1.6361829025844932,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 6995
    },
    {
      "epoch": 1.636416793357502,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 6996
    },
    {
      "epoch": 1.636650684130511,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 6997
    },
    {
      "epoch": 1.6368845749035201,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 6998
    },
    {
      "epoch": 1.637118465676529,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6746,
      "step": 6999
    },
    {
      "epoch": 1.637352356449538,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 7000
    },
    {
      "epoch": 1.637352356449538,
      "eval_runtime": 4.6025,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 7000
    },
    {
      "epoch": 1.637586247222547,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 7001
    },
    {
      "epoch": 1.637820137995556,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 7002
    },
    {
      "epoch": 1.6380540287685652,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 7003
    },
    {
      "epoch": 1.638287919541574,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.901,
      "step": 7004
    },
    {
      "epoch": 1.638521810314583,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 7005
    },
    {
      "epoch": 1.6387557010875922,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 7006
    },
    {
      "epoch": 1.638989591860601,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6271,
      "step": 7007
    },
    {
      "epoch": 1.63922348263361,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 7008
    },
    {
      "epoch": 1.6394573734066191,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 7009
    },
    {
      "epoch": 1.639691264179628,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 7010
    },
    {
      "epoch": 1.6399251549526372,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.549,
      "step": 7011
    },
    {
      "epoch": 1.640159045725646,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 7012
    },
    {
      "epoch": 1.6403929364986551,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6425,
      "step": 7013
    },
    {
      "epoch": 1.6406268272716642,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 7014
    },
    {
      "epoch": 1.640860718044673,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 7015
    },
    {
      "epoch": 1.641094608817682,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 7016
    },
    {
      "epoch": 1.6413284995906912,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7501,
      "step": 7017
    },
    {
      "epoch": 1.6415623903637,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 7018
    },
    {
      "epoch": 1.6417962811367093,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 7019
    },
    {
      "epoch": 1.6420301719097181,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 7020
    },
    {
      "epoch": 1.6422640626827272,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 7021
    },
    {
      "epoch": 1.6424979534557362,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 7022
    },
    {
      "epoch": 1.642731844228745,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 7023
    },
    {
      "epoch": 1.6429657350017541,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8346,
      "step": 7024
    },
    {
      "epoch": 1.6431996257747632,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 7025
    },
    {
      "epoch": 1.643433516547772,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 7026
    },
    {
      "epoch": 1.6436674073207813,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 7027
    },
    {
      "epoch": 1.6439012980937902,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2942,
      "step": 7028
    },
    {
      "epoch": 1.6441351888667992,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 7029
    },
    {
      "epoch": 1.6443690796398083,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 7030
    },
    {
      "epoch": 1.6446029704128171,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 7031
    },
    {
      "epoch": 1.6448368611858262,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 7032
    },
    {
      "epoch": 1.6450707519588352,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 7033
    },
    {
      "epoch": 1.645304642731844,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3251,
      "step": 7034
    },
    {
      "epoch": 1.6455385335048534,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2877,
      "step": 7035
    },
    {
      "epoch": 1.6457724242778622,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 7036
    },
    {
      "epoch": 1.6460063150508712,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 7037
    },
    {
      "epoch": 1.6462402058238803,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 7038
    },
    {
      "epoch": 1.6464740965968891,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.359,
      "step": 7039
    },
    {
      "epoch": 1.6467079873698982,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 7040
    },
    {
      "epoch": 1.6469418781429073,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 7041
    },
    {
      "epoch": 1.647175768915916,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 7042
    },
    {
      "epoch": 1.6474096596889254,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 7043
    },
    {
      "epoch": 1.6476435504619342,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 7044
    },
    {
      "epoch": 1.6478774412349433,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 7045
    },
    {
      "epoch": 1.6481113320079523,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 7046
    },
    {
      "epoch": 1.6483452227809612,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 7047
    },
    {
      "epoch": 1.6485791135539702,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 7048
    },
    {
      "epoch": 1.6488130043269793,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 7049
    },
    {
      "epoch": 1.6490468950999881,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6936,
      "step": 7050
    },
    {
      "epoch": 1.6492807858729974,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6245,
      "step": 7051
    },
    {
      "epoch": 1.6495146766460063,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 7052
    },
    {
      "epoch": 1.6497485674190153,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 7053
    },
    {
      "epoch": 1.6499824581920244,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 7054
    },
    {
      "epoch": 1.6502163489650332,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 7055
    },
    {
      "epoch": 1.6504502397380423,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 7056
    },
    {
      "epoch": 1.6506841305110513,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 7057
    },
    {
      "epoch": 1.6509180212840602,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 7058
    },
    {
      "epoch": 1.6511519120570695,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 7059
    },
    {
      "epoch": 1.6513858028300783,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5736,
      "step": 7060
    },
    {
      "epoch": 1.6516196936030874,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 7061
    },
    {
      "epoch": 1.6518535843760964,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3263,
      "step": 7062
    },
    {
      "epoch": 1.6520874751491053,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 7063
    },
    {
      "epoch": 1.6523213659221145,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8276,
      "step": 7064
    },
    {
      "epoch": 1.6525552566951234,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 7065
    },
    {
      "epoch": 1.6527891474681324,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5728,
      "step": 7066
    },
    {
      "epoch": 1.6530230382411415,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 7067
    },
    {
      "epoch": 1.6532569290141503,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2425,
      "step": 7068
    },
    {
      "epoch": 1.6534908197871594,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 7069
    },
    {
      "epoch": 1.6537247105601685,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 7070
    },
    {
      "epoch": 1.6539586013331773,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5802,
      "step": 7071
    },
    {
      "epoch": 1.6541924921061866,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 7072
    },
    {
      "epoch": 1.6544263828791954,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 7073
    },
    {
      "epoch": 1.6546602736522045,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 7074
    },
    {
      "epoch": 1.6548941644252135,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 7075
    },
    {
      "epoch": 1.6551280551982224,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 7076
    },
    {
      "epoch": 1.6553619459712314,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 7077
    },
    {
      "epoch": 1.6555958367442405,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.617,
      "step": 7078
    },
    {
      "epoch": 1.6558297275172493,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 7079
    },
    {
      "epoch": 1.6560636182902586,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 7080
    },
    {
      "epoch": 1.6562975090632674,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 7081
    },
    {
      "epoch": 1.6565313998362765,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9733,
      "step": 7082
    },
    {
      "epoch": 1.6567652906092856,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 7083
    },
    {
      "epoch": 1.6569991813822944,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 7084
    },
    {
      "epoch": 1.6572330721553035,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 7085
    },
    {
      "epoch": 1.6574669629283125,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8306,
      "step": 7086
    },
    {
      "epoch": 1.6577008537013214,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 7087
    },
    {
      "epoch": 1.6579347444743306,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 7088
    },
    {
      "epoch": 1.6581686352473395,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 7089
    },
    {
      "epoch": 1.6584025260203485,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 7090
    },
    {
      "epoch": 1.6586364167933576,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7494,
      "step": 7091
    },
    {
      "epoch": 1.6588703075663664,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 7092
    },
    {
      "epoch": 1.6591041983393755,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5938,
      "step": 7093
    },
    {
      "epoch": 1.6593380891123846,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 7094
    },
    {
      "epoch": 1.6595719798853934,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.5964,
      "step": 7095
    },
    {
      "epoch": 1.6598058706584027,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 7096
    },
    {
      "epoch": 1.6600397614314115,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 7097
    },
    {
      "epoch": 1.6602736522044206,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 7098
    },
    {
      "epoch": 1.6605075429774296,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 7099
    },
    {
      "epoch": 1.6607414337504385,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 7100
    },
    {
      "epoch": 1.6607414337504385,
      "eval_runtime": 4.5971,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 7100
    },
    {
      "epoch": 1.6609753245234475,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 7101
    },
    {
      "epoch": 1.6612092152964566,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 7102
    },
    {
      "epoch": 1.6614431060694654,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 7103
    },
    {
      "epoch": 1.6616769968424747,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 7104
    },
    {
      "epoch": 1.6619108876154836,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 7105
    },
    {
      "epoch": 1.6621447783884926,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 7106
    },
    {
      "epoch": 1.6623786691615017,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 7107
    },
    {
      "epoch": 1.6626125599345105,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 7108
    },
    {
      "epoch": 1.6628464507075196,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 7109
    },
    {
      "epoch": 1.6630803414805286,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 7110
    },
    {
      "epoch": 1.6633142322535375,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 7111
    },
    {
      "epoch": 1.6635481230265468,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 7112
    },
    {
      "epoch": 1.6637820137995556,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 7113
    },
    {
      "epoch": 1.6640159045725647,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 7114
    },
    {
      "epoch": 1.6642497953455737,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 7115
    },
    {
      "epoch": 1.6644836861185826,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5751,
      "step": 7116
    },
    {
      "epoch": 1.6647175768915916,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 7117
    },
    {
      "epoch": 1.6649514676646007,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5392,
      "step": 7118
    },
    {
      "epoch": 1.6651853584376095,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.395,
      "step": 7119
    },
    {
      "epoch": 1.6654192492106188,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 7120
    },
    {
      "epoch": 1.6656531399836276,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 7121
    },
    {
      "epoch": 1.6658870307566367,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 7122
    },
    {
      "epoch": 1.6661209215296457,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 7123
    },
    {
      "epoch": 1.6663548123026546,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 7124
    },
    {
      "epoch": 1.6665887030756636,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6972,
      "step": 7125
    },
    {
      "epoch": 1.6668225938486727,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 7126
    },
    {
      "epoch": 1.6670564846216815,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 7127
    },
    {
      "epoch": 1.6672903753946908,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 7128
    },
    {
      "epoch": 1.6675242661676997,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 7129
    },
    {
      "epoch": 1.6677581569407087,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 7130
    },
    {
      "epoch": 1.6679920477137178,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 7131
    },
    {
      "epoch": 1.6682259384867266,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5775,
      "step": 7132
    },
    {
      "epoch": 1.6684598292597357,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 7133
    },
    {
      "epoch": 1.6686937200327447,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 7134
    },
    {
      "epoch": 1.6689276108057536,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 7135
    },
    {
      "epoch": 1.6691615015787629,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 7136
    },
    {
      "epoch": 1.6693953923517717,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 7137
    },
    {
      "epoch": 1.6696292831247808,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 7138
    },
    {
      "epoch": 1.6698631738977898,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 7139
    },
    {
      "epoch": 1.6700970646707987,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 7140
    },
    {
      "epoch": 1.6703309554438077,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 7141
    },
    {
      "epoch": 1.6705648462168168,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 7142
    },
    {
      "epoch": 1.6707987369898256,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 7143
    },
    {
      "epoch": 1.671032627762835,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 7144
    },
    {
      "epoch": 1.6712665185358437,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 7145
    },
    {
      "epoch": 1.6715004093088528,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 7146
    },
    {
      "epoch": 1.6717343000818619,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 7147
    },
    {
      "epoch": 1.6719681908548707,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 7148
    },
    {
      "epoch": 1.6722020816278798,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 7149
    },
    {
      "epoch": 1.6724359724008888,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 7150
    },
    {
      "epoch": 1.6726698631738977,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 7151
    },
    {
      "epoch": 1.672903753946907,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 7152
    },
    {
      "epoch": 1.6731376447199158,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 7153
    },
    {
      "epoch": 1.6733715354929248,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.685,
      "step": 7154
    },
    {
      "epoch": 1.673605426265934,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 7155
    },
    {
      "epoch": 1.6738393170389427,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 7156
    },
    {
      "epoch": 1.6740732078119518,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7261,
      "step": 7157
    },
    {
      "epoch": 1.6743070985849609,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 7158
    },
    {
      "epoch": 1.6745409893579697,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 7159
    },
    {
      "epoch": 1.674774880130979,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 7160
    },
    {
      "epoch": 1.6750087709039878,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 7161
    },
    {
      "epoch": 1.6752426616769969,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 7162
    },
    {
      "epoch": 1.675476552450006,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.696,
      "step": 7163
    },
    {
      "epoch": 1.6757104432230148,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0503,
      "step": 7164
    },
    {
      "epoch": 1.6759443339960238,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 7165
    },
    {
      "epoch": 1.6761782247690329,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 7166
    },
    {
      "epoch": 1.6764121155420417,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 7167
    },
    {
      "epoch": 1.676646006315051,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9759,
      "step": 7168
    },
    {
      "epoch": 1.6768798970880598,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 7169
    },
    {
      "epoch": 1.677113787861069,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1575,
      "step": 7170
    },
    {
      "epoch": 1.677347678634078,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1506,
      "step": 7171
    },
    {
      "epoch": 1.6775815694070868,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 7172
    },
    {
      "epoch": 1.6778154601800959,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 7173
    },
    {
      "epoch": 1.678049350953105,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 7174
    },
    {
      "epoch": 1.6782832417261138,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 7175
    },
    {
      "epoch": 1.678517132499123,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 7176
    },
    {
      "epoch": 1.6787510232721319,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 7177
    },
    {
      "epoch": 1.678984914045141,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 7178
    },
    {
      "epoch": 1.67921880481815,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 7179
    },
    {
      "epoch": 1.6794526955911588,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 7180
    },
    {
      "epoch": 1.679686586364168,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 7181
    },
    {
      "epoch": 1.679920477137177,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 7182
    },
    {
      "epoch": 1.6801543679101858,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 7183
    },
    {
      "epoch": 1.680388258683195,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 7184
    },
    {
      "epoch": 1.680622149456204,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6414,
      "step": 7185
    },
    {
      "epoch": 1.680856040229213,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 7186
    },
    {
      "epoch": 1.681089931002222,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 7187
    },
    {
      "epoch": 1.6813238217752309,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 7188
    },
    {
      "epoch": 1.68155771254824,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 7189
    },
    {
      "epoch": 1.681791603321249,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 7190
    },
    {
      "epoch": 1.6820254940942578,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 7191
    },
    {
      "epoch": 1.6822593848672671,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 7192
    },
    {
      "epoch": 1.682493275640276,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 7193
    },
    {
      "epoch": 1.682727166413285,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5916,
      "step": 7194
    },
    {
      "epoch": 1.682961057186294,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 7195
    },
    {
      "epoch": 1.683194947959303,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 7196
    },
    {
      "epoch": 1.683428838732312,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6795,
      "step": 7197
    },
    {
      "epoch": 1.683662729505321,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7295,
      "step": 7198
    },
    {
      "epoch": 1.6838966202783299,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 7199
    },
    {
      "epoch": 1.6841305110513392,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 7200
    },
    {
      "epoch": 1.6841305110513392,
      "eval_runtime": 4.6822,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 7200
    },
    {
      "epoch": 1.684364401824348,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 7201
    },
    {
      "epoch": 1.684598292597357,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7749,
      "step": 7202
    },
    {
      "epoch": 1.684832183370366,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8468,
      "step": 7203
    },
    {
      "epoch": 1.685066074143375,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 7204
    },
    {
      "epoch": 1.685299964916384,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 7205
    },
    {
      "epoch": 1.685533855689393,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 7206
    },
    {
      "epoch": 1.685767746462402,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 7207
    },
    {
      "epoch": 1.6860016372354112,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 7208
    },
    {
      "epoch": 1.68623552800842,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 7209
    },
    {
      "epoch": 1.686469418781429,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 7210
    },
    {
      "epoch": 1.6867033095544381,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.4638,
      "step": 7211
    },
    {
      "epoch": 1.686937200327447,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 7212
    },
    {
      "epoch": 1.687171091100456,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 7213
    },
    {
      "epoch": 1.687404981873465,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6226,
      "step": 7214
    },
    {
      "epoch": 1.687638872646474,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 7215
    },
    {
      "epoch": 1.6878727634194832,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 7216
    },
    {
      "epoch": 1.688106654192492,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5313,
      "step": 7217
    },
    {
      "epoch": 1.6883405449655011,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 7218
    },
    {
      "epoch": 1.6885744357385102,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.736,
      "step": 7219
    },
    {
      "epoch": 1.688808326511519,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 7220
    },
    {
      "epoch": 1.689042217284528,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 7221
    },
    {
      "epoch": 1.6892761080575371,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 7222
    },
    {
      "epoch": 1.689509998830546,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 7223
    },
    {
      "epoch": 1.6897438896035553,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 7224
    },
    {
      "epoch": 1.689977780376564,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 7225
    },
    {
      "epoch": 1.6902116711495732,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 7226
    },
    {
      "epoch": 1.6904455619225822,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 7227
    },
    {
      "epoch": 1.690679452695591,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8451,
      "step": 7228
    },
    {
      "epoch": 1.6909133434686001,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 7229
    },
    {
      "epoch": 1.6911472342416092,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 7230
    },
    {
      "epoch": 1.691381125014618,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 7231
    },
    {
      "epoch": 1.6916150157876273,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 7232
    },
    {
      "epoch": 1.6918489065606361,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 7233
    },
    {
      "epoch": 1.6920827973336452,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 7234
    },
    {
      "epoch": 1.6923166881066543,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 7235
    },
    {
      "epoch": 1.692550578879663,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6123,
      "step": 7236
    },
    {
      "epoch": 1.6927844696526722,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3895,
      "step": 7237
    },
    {
      "epoch": 1.6930183604256812,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6949,
      "step": 7238
    },
    {
      "epoch": 1.69325225119869,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 7239
    },
    {
      "epoch": 1.6934861419716993,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6237,
      "step": 7240
    },
    {
      "epoch": 1.6937200327447082,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 7241
    },
    {
      "epoch": 1.6939539235177172,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6091,
      "step": 7242
    },
    {
      "epoch": 1.6941878142907263,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 7243
    },
    {
      "epoch": 1.6944217050637351,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 7244
    },
    {
      "epoch": 1.6946555958367442,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7619,
      "step": 7245
    },
    {
      "epoch": 1.6948894866097532,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 7246
    },
    {
      "epoch": 1.695123377382762,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 7247
    },
    {
      "epoch": 1.6953572681557714,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 7248
    },
    {
      "epoch": 1.6955911589287802,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 7249
    },
    {
      "epoch": 1.6958250497017893,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 7250
    },
    {
      "epoch": 1.6960589404747983,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 7251
    },
    {
      "epoch": 1.6962928312478072,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 7252
    },
    {
      "epoch": 1.6965267220208162,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 7253
    },
    {
      "epoch": 1.6967606127938253,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 7254
    },
    {
      "epoch": 1.6969945035668341,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 7255
    },
    {
      "epoch": 1.6972283943398434,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 7256
    },
    {
      "epoch": 1.6974622851128522,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 7257
    },
    {
      "epoch": 1.6976961758858613,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 7258
    },
    {
      "epoch": 1.6979300666588704,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 7259
    },
    {
      "epoch": 1.6981639574318792,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7088,
      "step": 7260
    },
    {
      "epoch": 1.6983978482048885,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 7261
    },
    {
      "epoch": 1.6986317389778973,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 7262
    },
    {
      "epoch": 1.6988656297509064,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 7263
    },
    {
      "epoch": 1.6990995205239154,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 7264
    },
    {
      "epoch": 1.6993334112969243,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7204,
      "step": 7265
    },
    {
      "epoch": 1.6995673020699333,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4046,
      "step": 7266
    },
    {
      "epoch": 1.6998011928429424,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 7267
    },
    {
      "epoch": 1.7000350836159512,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 7268
    },
    {
      "epoch": 1.7002689743889605,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 7269
    },
    {
      "epoch": 1.7005028651619694,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 7270
    },
    {
      "epoch": 1.7007367559349784,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6834,
      "step": 7271
    },
    {
      "epoch": 1.7009706467079875,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 7272
    },
    {
      "epoch": 1.7012045374809963,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 7273
    },
    {
      "epoch": 1.7014384282540054,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6225,
      "step": 7274
    },
    {
      "epoch": 1.7016723190270144,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 7275
    },
    {
      "epoch": 1.7019062098000233,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.26,
      "step": 7276
    },
    {
      "epoch": 1.7021401005730326,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8929,
      "step": 7277
    },
    {
      "epoch": 1.7023739913460414,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 7278
    },
    {
      "epoch": 1.7026078821190505,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6879,
      "step": 7279
    },
    {
      "epoch": 1.7028417728920595,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9758,
      "step": 7280
    },
    {
      "epoch": 1.7030756636650684,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 7281
    },
    {
      "epoch": 1.7033095544380774,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 7282
    },
    {
      "epoch": 1.7035434452110865,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2614,
      "step": 7283
    },
    {
      "epoch": 1.7037773359840953,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 7284
    },
    {
      "epoch": 1.7040112267571046,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 7285
    },
    {
      "epoch": 1.7042451175301134,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8026,
      "step": 7286
    },
    {
      "epoch": 1.7044790083031225,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 7287
    },
    {
      "epoch": 1.7047128990761315,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 7288
    },
    {
      "epoch": 1.7049467898491404,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 7289
    },
    {
      "epoch": 1.7051806806221494,
      "grad_norm": 14.625,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 7290
    },
    {
      "epoch": 1.7054145713951585,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 7291
    },
    {
      "epoch": 1.7056484621681673,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 7292
    },
    {
      "epoch": 1.7058823529411766,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 7293
    },
    {
      "epoch": 1.7061162437141855,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 7294
    },
    {
      "epoch": 1.7063501344871945,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 7295
    },
    {
      "epoch": 1.7065840252602036,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 7296
    },
    {
      "epoch": 1.7068179160332124,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 7297
    },
    {
      "epoch": 1.7070518068062215,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 7298
    },
    {
      "epoch": 1.7072856975792305,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 7299
    },
    {
      "epoch": 1.7075195883522394,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 7300
    },
    {
      "epoch": 1.7075195883522394,
      "eval_runtime": 4.6064,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 7300
    },
    {
      "epoch": 1.7077534791252487,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 7301
    },
    {
      "epoch": 1.7079873698982575,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 7302
    },
    {
      "epoch": 1.7082212606712666,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 7303
    },
    {
      "epoch": 1.7084551514442756,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 7304
    },
    {
      "epoch": 1.7086890422172845,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1573,
      "step": 7305
    },
    {
      "epoch": 1.7089229329902935,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5157,
      "step": 7306
    },
    {
      "epoch": 1.7091568237633026,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 7307
    },
    {
      "epoch": 1.7093907145363114,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 7308
    },
    {
      "epoch": 1.7096246053093207,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 7309
    },
    {
      "epoch": 1.7098584960823295,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 7310
    },
    {
      "epoch": 1.7100923868553386,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 7311
    },
    {
      "epoch": 1.7103262776283477,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 7312
    },
    {
      "epoch": 1.7105601684013565,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 7313
    },
    {
      "epoch": 1.7107940591743656,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 7314
    },
    {
      "epoch": 1.7110279499473746,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 7315
    },
    {
      "epoch": 1.7112618407203835,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6829,
      "step": 7316
    },
    {
      "epoch": 1.7114957314933927,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 7317
    },
    {
      "epoch": 1.7117296222664016,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 7318
    },
    {
      "epoch": 1.7119635130394106,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5794,
      "step": 7319
    },
    {
      "epoch": 1.7121974038124197,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 7320
    },
    {
      "epoch": 1.7124312945854285,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 7321
    },
    {
      "epoch": 1.7126651853584376,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 7322
    },
    {
      "epoch": 1.7128990761314467,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 7323
    },
    {
      "epoch": 1.7131329669044555,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 7324
    },
    {
      "epoch": 1.7133668576774648,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2086,
      "step": 7325
    },
    {
      "epoch": 1.7136007484504736,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3042,
      "step": 7326
    },
    {
      "epoch": 1.7138346392234827,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 7327
    },
    {
      "epoch": 1.7140685299964917,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 7328
    },
    {
      "epoch": 1.7143024207695006,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 7329
    },
    {
      "epoch": 1.7145363115425096,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8385,
      "step": 7330
    },
    {
      "epoch": 1.7147702023155187,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 7331
    },
    {
      "epoch": 1.7150040930885275,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 7332
    },
    {
      "epoch": 1.7152379838615368,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 7333
    },
    {
      "epoch": 1.7154718746345456,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 7334
    },
    {
      "epoch": 1.7157057654075547,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 7335
    },
    {
      "epoch": 1.7159396561805638,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 7336
    },
    {
      "epoch": 1.7161735469535726,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8027,
      "step": 7337
    },
    {
      "epoch": 1.7164074377265817,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 7338
    },
    {
      "epoch": 1.7166413284995907,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 7339
    },
    {
      "epoch": 1.7168752192725996,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6263,
      "step": 7340
    },
    {
      "epoch": 1.7171091100456088,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 7341
    },
    {
      "epoch": 1.7173430008186177,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5621,
      "step": 7342
    },
    {
      "epoch": 1.7175768915916267,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 7343
    },
    {
      "epoch": 1.7178107823646358,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3522,
      "step": 7344
    },
    {
      "epoch": 1.7180446731376446,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5556,
      "step": 7345
    },
    {
      "epoch": 1.7182785639106537,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 7346
    },
    {
      "epoch": 1.7185124546836628,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6165,
      "step": 7347
    },
    {
      "epoch": 1.7187463454566716,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.885,
      "step": 7348
    },
    {
      "epoch": 1.7189802362296809,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5698,
      "step": 7349
    },
    {
      "epoch": 1.7192141270026897,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3878,
      "step": 7350
    },
    {
      "epoch": 1.7194480177756988,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 7351
    },
    {
      "epoch": 1.7196819085487078,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 7352
    },
    {
      "epoch": 1.7199157993217167,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 7353
    },
    {
      "epoch": 1.7201496900947257,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 7354
    },
    {
      "epoch": 1.7203835808677348,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 7355
    },
    {
      "epoch": 1.7206174716407436,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 7356
    },
    {
      "epoch": 1.720851362413753,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6686,
      "step": 7357
    },
    {
      "epoch": 1.7210852531867618,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 7358
    },
    {
      "epoch": 1.7213191439597708,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 7359
    },
    {
      "epoch": 1.7215530347327799,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 7360
    },
    {
      "epoch": 1.7217869255057887,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 7361
    },
    {
      "epoch": 1.7220208162787978,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 7362
    },
    {
      "epoch": 1.7222547070518068,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 7363
    },
    {
      "epoch": 1.7224885978248157,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 7364
    },
    {
      "epoch": 1.722722488597825,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 7365
    },
    {
      "epoch": 1.7229563793708338,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 7366
    },
    {
      "epoch": 1.7231902701438429,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 7367
    },
    {
      "epoch": 1.723424160916852,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 7368
    },
    {
      "epoch": 1.7236580516898607,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 7369
    },
    {
      "epoch": 1.7238919424628698,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 7370
    },
    {
      "epoch": 1.7241258332358789,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 7371
    },
    {
      "epoch": 1.7243597240088877,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 7372
    },
    {
      "epoch": 1.724593614781897,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 7373
    },
    {
      "epoch": 1.7248275055549058,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 7374
    },
    {
      "epoch": 1.7250613963279149,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 7375
    },
    {
      "epoch": 1.725295287100924,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 7376
    },
    {
      "epoch": 1.7255291778739328,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 7377
    },
    {
      "epoch": 1.7257630686469418,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 7378
    },
    {
      "epoch": 1.725996959419951,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 7379
    },
    {
      "epoch": 1.7262308501929597,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 7380
    },
    {
      "epoch": 1.726464740965969,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 7381
    },
    {
      "epoch": 1.7266986317389779,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 7382
    },
    {
      "epoch": 1.726932522511987,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 7383
    },
    {
      "epoch": 1.727166413284996,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 7384
    },
    {
      "epoch": 1.7274003040580048,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 7385
    },
    {
      "epoch": 1.7276341948310139,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 7386
    },
    {
      "epoch": 1.727868085604023,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9152,
      "step": 7387
    },
    {
      "epoch": 1.7281019763770318,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.659,
      "step": 7388
    },
    {
      "epoch": 1.728335867150041,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 7389
    },
    {
      "epoch": 1.72856975792305,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 7390
    },
    {
      "epoch": 1.728803648696059,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 7391
    },
    {
      "epoch": 1.729037539469068,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 7392
    },
    {
      "epoch": 1.7292714302420769,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 7393
    },
    {
      "epoch": 1.729505321015086,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6798,
      "step": 7394
    },
    {
      "epoch": 1.729739211788095,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 7395
    },
    {
      "epoch": 1.7299731025611038,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 7396
    },
    {
      "epoch": 1.730206993334113,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 7397
    },
    {
      "epoch": 1.730440884107122,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 7398
    },
    {
      "epoch": 1.730674774880131,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 7399
    },
    {
      "epoch": 1.73090866565314,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 7400
    },
    {
      "epoch": 1.73090866565314,
      "eval_runtime": 4.6286,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 7400
    },
    {
      "epoch": 1.731142556426149,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.8007,
      "step": 7401
    },
    {
      "epoch": 1.731376447199158,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 7402
    },
    {
      "epoch": 1.731610337972167,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 7403
    },
    {
      "epoch": 1.7318442287451759,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 7404
    },
    {
      "epoch": 1.7320781195181851,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 7405
    },
    {
      "epoch": 1.732312010291194,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 7406
    },
    {
      "epoch": 1.732545901064203,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 7407
    },
    {
      "epoch": 1.732779791837212,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 7408
    },
    {
      "epoch": 1.733013682610221,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 7409
    },
    {
      "epoch": 1.73324757338323,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 7410
    },
    {
      "epoch": 1.733481464156239,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 7411
    },
    {
      "epoch": 1.7337153549292479,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 7412
    },
    {
      "epoch": 1.7339492457022572,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7254,
      "step": 7413
    },
    {
      "epoch": 1.734183136475266,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 7414
    },
    {
      "epoch": 1.734417027248275,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 7415
    },
    {
      "epoch": 1.7346509180212841,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 7416
    },
    {
      "epoch": 1.734884808794293,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6649,
      "step": 7417
    },
    {
      "epoch": 1.735118699567302,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 7418
    },
    {
      "epoch": 1.735352590340311,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 7419
    },
    {
      "epoch": 1.73558648111332,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8815,
      "step": 7420
    },
    {
      "epoch": 1.7358203718863292,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.4477,
      "step": 7421
    },
    {
      "epoch": 1.736054262659338,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6616,
      "step": 7422
    },
    {
      "epoch": 1.736288153432347,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 7423
    },
    {
      "epoch": 1.7365220442053562,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 7424
    },
    {
      "epoch": 1.736755934978365,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 7425
    },
    {
      "epoch": 1.736989825751374,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 7426
    },
    {
      "epoch": 1.7372237165243831,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 7427
    },
    {
      "epoch": 1.737457607297392,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 7428
    },
    {
      "epoch": 1.7376914980704012,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 7429
    },
    {
      "epoch": 1.73792538884341,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 7430
    },
    {
      "epoch": 1.7381592796164191,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 7431
    },
    {
      "epoch": 1.7383931703894282,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 7432
    },
    {
      "epoch": 1.738627061162437,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 7433
    },
    {
      "epoch": 1.738860951935446,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 7434
    },
    {
      "epoch": 1.7390948427084552,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6876,
      "step": 7435
    },
    {
      "epoch": 1.739328733481464,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6286,
      "step": 7436
    },
    {
      "epoch": 1.7395626242544733,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 7437
    },
    {
      "epoch": 1.7397965150274821,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 7438
    },
    {
      "epoch": 1.7400304058004912,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 7439
    },
    {
      "epoch": 1.7402642965735002,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 7440
    },
    {
      "epoch": 1.740498187346509,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 7441
    },
    {
      "epoch": 1.7407320781195181,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 7442
    },
    {
      "epoch": 1.7409659688925272,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 7443
    },
    {
      "epoch": 1.741199859665536,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 7444
    },
    {
      "epoch": 1.7414337504385453,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 7445
    },
    {
      "epoch": 1.7416676412115542,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 7446
    },
    {
      "epoch": 1.7419015319845632,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 7447
    },
    {
      "epoch": 1.7421354227575723,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8058,
      "step": 7448
    },
    {
      "epoch": 1.742369313530581,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 7449
    },
    {
      "epoch": 1.7426032043035902,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9103,
      "step": 7450
    },
    {
      "epoch": 1.7428370950765992,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 7451
    },
    {
      "epoch": 1.743070985849608,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 7452
    },
    {
      "epoch": 1.7433048766226173,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 7453
    },
    {
      "epoch": 1.7435387673956262,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 7454
    },
    {
      "epoch": 1.7437726581686352,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 7455
    },
    {
      "epoch": 1.7440065489416443,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 7456
    },
    {
      "epoch": 1.7442404397146531,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 7457
    },
    {
      "epoch": 1.7444743304876624,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 7458
    },
    {
      "epoch": 1.7447082212606713,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 7459
    },
    {
      "epoch": 1.7449421120336803,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6695,
      "step": 7460
    },
    {
      "epoch": 1.7451760028066894,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 7461
    },
    {
      "epoch": 1.7454098935796982,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 7462
    },
    {
      "epoch": 1.7456437843527073,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 7463
    },
    {
      "epoch": 1.7458776751257163,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 7464
    },
    {
      "epoch": 1.7461115658987252,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.273,
      "step": 7465
    },
    {
      "epoch": 1.7463454566717345,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 7466
    },
    {
      "epoch": 1.7465793474447433,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8347,
      "step": 7467
    },
    {
      "epoch": 1.7468132382177524,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 7468
    },
    {
      "epoch": 1.7470471289907614,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 7469
    },
    {
      "epoch": 1.7472810197637703,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 7470
    },
    {
      "epoch": 1.7475149105367793,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.4358,
      "step": 7471
    },
    {
      "epoch": 1.7477488013097884,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 7472
    },
    {
      "epoch": 1.7479826920827972,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 7473
    },
    {
      "epoch": 1.7482165828558065,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9455,
      "step": 7474
    },
    {
      "epoch": 1.7484504736288153,
      "grad_norm": 9.75,
      "learning_rate": 3e-05,
      "loss": 2.4018,
      "step": 7475
    },
    {
      "epoch": 1.7486843644018244,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 7476
    },
    {
      "epoch": 1.7489182551748335,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 7477
    },
    {
      "epoch": 1.7491521459478423,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 7478
    },
    {
      "epoch": 1.7493860367208514,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 7479
    },
    {
      "epoch": 1.7496199274938604,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 7480
    },
    {
      "epoch": 1.7498538182668693,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 7481
    },
    {
      "epoch": 1.7500877090398785,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 7482
    },
    {
      "epoch": 1.7503215998128874,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 7483
    },
    {
      "epoch": 1.7505554905858964,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 7484
    },
    {
      "epoch": 1.7507893813589055,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3135,
      "step": 7485
    },
    {
      "epoch": 1.7510232721319143,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 7486
    },
    {
      "epoch": 1.7512571629049234,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 7487
    },
    {
      "epoch": 1.7514910536779325,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 7488
    },
    {
      "epoch": 1.7517249444509413,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 7489
    },
    {
      "epoch": 1.7519588352239506,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 7490
    },
    {
      "epoch": 1.7521927259969594,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 7491
    },
    {
      "epoch": 1.7524266167699685,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 7492
    },
    {
      "epoch": 1.7526605075429775,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 7493
    },
    {
      "epoch": 1.7528943983159864,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 7494
    },
    {
      "epoch": 1.7531282890889954,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 7495
    },
    {
      "epoch": 1.7533621798620045,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1233,
      "step": 7496
    },
    {
      "epoch": 1.7535960706350133,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 7497
    },
    {
      "epoch": 1.7538299614080226,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 7498
    },
    {
      "epoch": 1.7540638521810314,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 7499
    },
    {
      "epoch": 1.7542977429540405,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 7500
    },
    {
      "epoch": 1.7542977429540405,
      "eval_runtime": 4.6486,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 7500
    },
    {
      "epoch": 1.7545316337270496,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 7501
    },
    {
      "epoch": 1.7547655245000584,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 7502
    },
    {
      "epoch": 1.7549994152730675,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 7503
    },
    {
      "epoch": 1.7552333060460765,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 7504
    },
    {
      "epoch": 1.7554671968190854,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7608,
      "step": 7505
    },
    {
      "epoch": 1.7557010875920946,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 7506
    },
    {
      "epoch": 1.7559349783651035,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4276,
      "step": 7507
    },
    {
      "epoch": 1.7561688691381125,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 7508
    },
    {
      "epoch": 1.7564027599111216,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 7509
    },
    {
      "epoch": 1.7566366506841304,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 7510
    },
    {
      "epoch": 1.7568705414571395,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.4952,
      "step": 7511
    },
    {
      "epoch": 1.7571044322301486,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 7512
    },
    {
      "epoch": 1.7573383230031574,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 7513
    },
    {
      "epoch": 1.7575722137761667,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9388,
      "step": 7514
    },
    {
      "epoch": 1.7578061045491755,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 7515
    },
    {
      "epoch": 1.7580399953221846,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 7516
    },
    {
      "epoch": 1.7582738860951936,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3397,
      "step": 7517
    },
    {
      "epoch": 1.7585077768682025,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 7518
    },
    {
      "epoch": 1.7587416676412115,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3727,
      "step": 7519
    },
    {
      "epoch": 1.7589755584142206,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 7520
    },
    {
      "epoch": 1.7592094491872294,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 7521
    },
    {
      "epoch": 1.7594433399602387,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 7522
    },
    {
      "epoch": 1.7596772307332476,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2488,
      "step": 7523
    },
    {
      "epoch": 1.7599111215062566,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7998,
      "step": 7524
    },
    {
      "epoch": 1.7601450122792657,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 7525
    },
    {
      "epoch": 1.7603789030522745,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 7526
    },
    {
      "epoch": 1.7606127938252836,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 7527
    },
    {
      "epoch": 1.7608466845982926,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 7528
    },
    {
      "epoch": 1.7610805753713015,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6166,
      "step": 7529
    },
    {
      "epoch": 1.7613144661443108,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 7530
    },
    {
      "epoch": 1.7615483569173196,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 7531
    },
    {
      "epoch": 1.7617822476903287,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 7532
    },
    {
      "epoch": 1.7620161384633377,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 7533
    },
    {
      "epoch": 1.7622500292363465,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 7534
    },
    {
      "epoch": 1.7624839200093556,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.729,
      "step": 7535
    },
    {
      "epoch": 1.7627178107823647,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 7536
    },
    {
      "epoch": 1.7629517015553735,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 7537
    },
    {
      "epoch": 1.7631855923283828,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 7538
    },
    {
      "epoch": 1.7634194831013916,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 7539
    },
    {
      "epoch": 1.7636533738744007,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3401,
      "step": 7540
    },
    {
      "epoch": 1.7638872646474097,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 7541
    },
    {
      "epoch": 1.7641211554204186,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 7542
    },
    {
      "epoch": 1.7643550461934276,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 7543
    },
    {
      "epoch": 1.7645889369664367,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 7544
    },
    {
      "epoch": 1.7648228277394455,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 7545
    },
    {
      "epoch": 1.7650567185124548,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7112,
      "step": 7546
    },
    {
      "epoch": 1.7652906092854637,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2502,
      "step": 7547
    },
    {
      "epoch": 1.7655245000584727,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 7548
    },
    {
      "epoch": 1.7657583908314818,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8801,
      "step": 7549
    },
    {
      "epoch": 1.7659922816044906,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 7550
    },
    {
      "epoch": 1.7662261723774997,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 7551
    },
    {
      "epoch": 1.7664600631505087,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 7552
    },
    {
      "epoch": 1.7666939539235176,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2718,
      "step": 7553
    },
    {
      "epoch": 1.7669278446965269,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 7554
    },
    {
      "epoch": 1.7671617354695357,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 7555
    },
    {
      "epoch": 1.7673956262425448,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 7556
    },
    {
      "epoch": 1.7676295170155538,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 7557
    },
    {
      "epoch": 1.7678634077885627,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 7558
    },
    {
      "epoch": 1.7680972985615717,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 7559
    },
    {
      "epoch": 1.7683311893345808,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 7560
    },
    {
      "epoch": 1.7685650801075896,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 7561
    },
    {
      "epoch": 1.768798970880599,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 7562
    },
    {
      "epoch": 1.7690328616536077,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 7563
    },
    {
      "epoch": 1.7692667524266168,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 7564
    },
    {
      "epoch": 1.7695006431996259,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 7565
    },
    {
      "epoch": 1.7697345339726347,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6042,
      "step": 7566
    },
    {
      "epoch": 1.7699684247456438,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 7567
    },
    {
      "epoch": 1.7702023155186528,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 7568
    },
    {
      "epoch": 1.7704362062916617,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 7569
    },
    {
      "epoch": 1.770670097064671,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 7570
    },
    {
      "epoch": 1.7709039878376798,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 7571
    },
    {
      "epoch": 1.7711378786106888,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 7572
    },
    {
      "epoch": 1.771371769383698,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 7573
    },
    {
      "epoch": 1.7716056601567067,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 7574
    },
    {
      "epoch": 1.7718395509297158,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 7575
    },
    {
      "epoch": 1.7720734417027248,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 7576
    },
    {
      "epoch": 1.7723073324757337,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 7577
    },
    {
      "epoch": 1.772541223248743,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 7578
    },
    {
      "epoch": 1.7727751140217518,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4587,
      "step": 7579
    },
    {
      "epoch": 1.7730090047947609,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 7580
    },
    {
      "epoch": 1.77324289556777,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 7581
    },
    {
      "epoch": 1.7734767863407788,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8359,
      "step": 7582
    },
    {
      "epoch": 1.7737106771137878,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 7583
    },
    {
      "epoch": 1.7739445678867969,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 7584
    },
    {
      "epoch": 1.7741784586598057,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 7585
    },
    {
      "epoch": 1.774412349432815,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 7586
    },
    {
      "epoch": 1.7746462402058238,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6602,
      "step": 7587
    },
    {
      "epoch": 1.774880130978833,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 7588
    },
    {
      "epoch": 1.775114021751842,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 7589
    },
    {
      "epoch": 1.7753479125248508,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7515,
      "step": 7590
    },
    {
      "epoch": 1.7755818032978599,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 7591
    },
    {
      "epoch": 1.775815694070869,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 7592
    },
    {
      "epoch": 1.7760495848438778,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2292,
      "step": 7593
    },
    {
      "epoch": 1.776283475616887,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 7594
    },
    {
      "epoch": 1.7765173663898959,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 7595
    },
    {
      "epoch": 1.776751257162905,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 7596
    },
    {
      "epoch": 1.776985147935914,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 7597
    },
    {
      "epoch": 1.7772190387089228,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 7598
    },
    {
      "epoch": 1.777452929481932,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 7599
    },
    {
      "epoch": 1.777686820254941,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 7600
    },
    {
      "epoch": 1.777686820254941,
      "eval_runtime": 4.6218,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 7600
    },
    {
      "epoch": 1.7779207110279498,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 7601
    },
    {
      "epoch": 1.778154601800959,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 7602
    },
    {
      "epoch": 1.778388492573968,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 7603
    },
    {
      "epoch": 1.778622383346977,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8644,
      "step": 7604
    },
    {
      "epoch": 1.778856274119986,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 7605
    },
    {
      "epoch": 1.7790901648929949,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 7606
    },
    {
      "epoch": 1.779324055666004,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 7607
    },
    {
      "epoch": 1.779557946439013,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 7608
    },
    {
      "epoch": 1.7797918372120218,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.713,
      "step": 7609
    },
    {
      "epoch": 1.7800257279850311,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 7610
    },
    {
      "epoch": 1.78025961875804,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 7611
    },
    {
      "epoch": 1.780493509531049,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 7612
    },
    {
      "epoch": 1.780727400304058,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 7613
    },
    {
      "epoch": 1.780961291077067,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 7614
    },
    {
      "epoch": 1.781195181850076,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 7615
    },
    {
      "epoch": 1.781429072623085,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 7616
    },
    {
      "epoch": 1.7816629633960939,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7245,
      "step": 7617
    },
    {
      "epoch": 1.7818968541691032,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 7618
    },
    {
      "epoch": 1.782130744942112,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 7619
    },
    {
      "epoch": 1.782364635715121,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6884,
      "step": 7620
    },
    {
      "epoch": 1.78259852648813,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 7621
    },
    {
      "epoch": 1.782832417261139,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 7622
    },
    {
      "epoch": 1.783066308034148,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 7623
    },
    {
      "epoch": 1.783300198807157,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7028,
      "step": 7624
    },
    {
      "epoch": 1.783534089580166,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 7625
    },
    {
      "epoch": 1.7837679803531752,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3541,
      "step": 7626
    },
    {
      "epoch": 1.784001871126184,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 7627
    },
    {
      "epoch": 1.784235761899193,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5948,
      "step": 7628
    },
    {
      "epoch": 1.7844696526722021,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2488,
      "step": 7629
    },
    {
      "epoch": 1.784703543445211,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 7630
    },
    {
      "epoch": 1.78493743421822,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 7631
    },
    {
      "epoch": 1.785171324991229,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 7632
    },
    {
      "epoch": 1.785405215764238,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 7633
    },
    {
      "epoch": 1.7856391065372472,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 7634
    },
    {
      "epoch": 1.785872997310256,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 7635
    },
    {
      "epoch": 1.7861068880832651,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 7636
    },
    {
      "epoch": 1.7863407788562742,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 7637
    },
    {
      "epoch": 1.786574669629283,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 7638
    },
    {
      "epoch": 1.786808560402292,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 7639
    },
    {
      "epoch": 1.7870424511753011,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 7640
    },
    {
      "epoch": 1.78727634194831,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9033,
      "step": 7641
    },
    {
      "epoch": 1.7875102327213193,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 7642
    },
    {
      "epoch": 1.787744123494328,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 7643
    },
    {
      "epoch": 1.7879780142673372,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 7644
    },
    {
      "epoch": 1.7882119050403462,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 7645
    },
    {
      "epoch": 1.788445795813355,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 7646
    },
    {
      "epoch": 1.7886796865863641,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 7647
    },
    {
      "epoch": 1.7889135773593732,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 7648
    },
    {
      "epoch": 1.789147468132382,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 7649
    },
    {
      "epoch": 1.7893813589053913,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 7650
    },
    {
      "epoch": 1.7896152496784001,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 7651
    },
    {
      "epoch": 1.7898491404514092,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 7652
    },
    {
      "epoch": 1.7900830312244183,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 7653
    },
    {
      "epoch": 1.790316921997427,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.6267,
      "step": 7654
    },
    {
      "epoch": 1.7905508127704362,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 7655
    },
    {
      "epoch": 1.7907847035434452,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 7656
    },
    {
      "epoch": 1.7910185943164543,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 7657
    },
    {
      "epoch": 1.7912524850894633,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 7658
    },
    {
      "epoch": 1.7914863758624722,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 7659
    },
    {
      "epoch": 1.7917202666354812,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 7660
    },
    {
      "epoch": 1.7919541574084903,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 7661
    },
    {
      "epoch": 1.7921880481814991,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.5714,
      "step": 7662
    },
    {
      "epoch": 1.7924219389545084,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 7663
    },
    {
      "epoch": 1.7926558297275172,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 7664
    },
    {
      "epoch": 1.7928897205005263,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3905,
      "step": 7665
    },
    {
      "epoch": 1.7931236112735354,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 7666
    },
    {
      "epoch": 1.7933575020465442,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7907,
      "step": 7667
    },
    {
      "epoch": 1.7935913928195533,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6621,
      "step": 7668
    },
    {
      "epoch": 1.7938252835925623,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 7669
    },
    {
      "epoch": 1.7940591743655712,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 7670
    },
    {
      "epoch": 1.7942930651385804,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 7671
    },
    {
      "epoch": 1.7945269559115893,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8568,
      "step": 7672
    },
    {
      "epoch": 1.7947608466845983,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 7673
    },
    {
      "epoch": 1.7949947374576074,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 7674
    },
    {
      "epoch": 1.7952286282306162,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 7675
    },
    {
      "epoch": 1.7954625190036253,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 7676
    },
    {
      "epoch": 1.7956964097766344,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 7677
    },
    {
      "epoch": 1.7959303005496432,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 7678
    },
    {
      "epoch": 1.7961641913226525,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 7679
    },
    {
      "epoch": 1.7963980820956613,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 7680
    },
    {
      "epoch": 1.7966319728686704,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 7681
    },
    {
      "epoch": 1.7968658636416794,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 7682
    },
    {
      "epoch": 1.7970997544146883,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 7683
    },
    {
      "epoch": 1.7973336451876973,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 7684
    },
    {
      "epoch": 1.7975675359607064,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 7685
    },
    {
      "epoch": 1.7978014267337152,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 7686
    },
    {
      "epoch": 1.7980353175067245,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6194,
      "step": 7687
    },
    {
      "epoch": 1.7982692082797334,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 7688
    },
    {
      "epoch": 1.7985030990527424,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 7689
    },
    {
      "epoch": 1.7987369898257515,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6518,
      "step": 7690
    },
    {
      "epoch": 1.7989708805987603,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 7691
    },
    {
      "epoch": 1.7992047713717694,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 7692
    },
    {
      "epoch": 1.7994386621447784,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 7693
    },
    {
      "epoch": 1.7996725529177873,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 7694
    },
    {
      "epoch": 1.7999064436907966,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 7695
    },
    {
      "epoch": 1.8001403344638054,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 7696
    },
    {
      "epoch": 1.8003742252368145,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 7697
    },
    {
      "epoch": 1.8006081160098235,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0348,
      "step": 7698
    },
    {
      "epoch": 1.8008420067828323,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 7699
    },
    {
      "epoch": 1.8010758975558414,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9262,
      "step": 7700
    },
    {
      "epoch": 1.8010758975558414,
      "eval_runtime": 4.5791,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 7700
    },
    {
      "epoch": 1.8013097883288505,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 7701
    },
    {
      "epoch": 1.8015436791018593,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 7702
    },
    {
      "epoch": 1.8017775698748686,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 7703
    },
    {
      "epoch": 1.8020114606478774,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 7704
    },
    {
      "epoch": 1.8022453514208865,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 7705
    },
    {
      "epoch": 1.8024792421938955,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 7706
    },
    {
      "epoch": 1.8027131329669044,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 7707
    },
    {
      "epoch": 1.8029470237399134,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7366,
      "step": 7708
    },
    {
      "epoch": 1.8031809145129225,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 7709
    },
    {
      "epoch": 1.8034148052859313,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 7710
    },
    {
      "epoch": 1.8036486960589406,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 7711
    },
    {
      "epoch": 1.8038825868319495,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 7712
    },
    {
      "epoch": 1.8041164776049585,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7586,
      "step": 7713
    },
    {
      "epoch": 1.8043503683779676,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 7714
    },
    {
      "epoch": 1.8045842591509764,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6674,
      "step": 7715
    },
    {
      "epoch": 1.8048181499239855,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 7716
    },
    {
      "epoch": 1.8050520406969945,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 7717
    },
    {
      "epoch": 1.8052859314700034,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 7718
    },
    {
      "epoch": 1.8055198222430127,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 7719
    },
    {
      "epoch": 1.8057537130160215,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 7720
    },
    {
      "epoch": 1.8059876037890306,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 7721
    },
    {
      "epoch": 1.8062214945620396,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7101,
      "step": 7722
    },
    {
      "epoch": 1.8064553853350485,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3158,
      "step": 7723
    },
    {
      "epoch": 1.8066892761080575,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 7724
    },
    {
      "epoch": 1.8069231668810666,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 7725
    },
    {
      "epoch": 1.8071570576540754,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 7726
    },
    {
      "epoch": 1.8073909484270847,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 7727
    },
    {
      "epoch": 1.8076248392000935,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 7728
    },
    {
      "epoch": 1.8078587299731026,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.527,
      "step": 7729
    },
    {
      "epoch": 1.8080926207461117,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 7730
    },
    {
      "epoch": 1.8083265115191205,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 7731
    },
    {
      "epoch": 1.8085604022921296,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 7732
    },
    {
      "epoch": 1.8087942930651386,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6381,
      "step": 7733
    },
    {
      "epoch": 1.8090281838381475,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 7734
    },
    {
      "epoch": 1.8092620746111567,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2499,
      "step": 7735
    },
    {
      "epoch": 1.8094959653841656,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 7736
    },
    {
      "epoch": 1.8097298561571746,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 7737
    },
    {
      "epoch": 1.8099637469301837,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 7738
    },
    {
      "epoch": 1.8101976377031925,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7803,
      "step": 7739
    },
    {
      "epoch": 1.8104315284762016,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 7740
    },
    {
      "epoch": 1.8106654192492107,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 7741
    },
    {
      "epoch": 1.8108993100222195,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0932,
      "step": 7742
    },
    {
      "epoch": 1.8111332007952288,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 7743
    },
    {
      "epoch": 1.8113670915682376,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6008,
      "step": 7744
    },
    {
      "epoch": 1.8116009823412467,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 7745
    },
    {
      "epoch": 1.8118348731142557,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 7746
    },
    {
      "epoch": 1.8120687638872646,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 7747
    },
    {
      "epoch": 1.8123026546602736,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 7748
    },
    {
      "epoch": 1.8125365454332827,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 7749
    },
    {
      "epoch": 1.8127704362062915,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 7750
    },
    {
      "epoch": 1.8130043269793008,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2222,
      "step": 7751
    },
    {
      "epoch": 1.8132382177523096,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 7752
    },
    {
      "epoch": 1.8134721085253187,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5135,
      "step": 7753
    },
    {
      "epoch": 1.8137059992983278,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6145,
      "step": 7754
    },
    {
      "epoch": 1.8139398900713366,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 7755
    },
    {
      "epoch": 1.8141737808443457,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 7756
    },
    {
      "epoch": 1.8144076716173547,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 7757
    },
    {
      "epoch": 1.8146415623903636,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 7758
    },
    {
      "epoch": 1.8148754531633728,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 7759
    },
    {
      "epoch": 1.8151093439363817,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 7760
    },
    {
      "epoch": 1.8153432347093907,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0836,
      "step": 7761
    },
    {
      "epoch": 1.8155771254823998,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 7762
    },
    {
      "epoch": 1.8158110162554086,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 7763
    },
    {
      "epoch": 1.8160449070284177,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2691,
      "step": 7764
    },
    {
      "epoch": 1.8162787978014268,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 7765
    },
    {
      "epoch": 1.8165126885744356,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6669,
      "step": 7766
    },
    {
      "epoch": 1.8167465793474449,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 7767
    },
    {
      "epoch": 1.8169804701204537,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 7768
    },
    {
      "epoch": 1.8172143608934628,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 7769
    },
    {
      "epoch": 1.8174482516664718,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6868,
      "step": 7770
    },
    {
      "epoch": 1.8176821424394807,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6568,
      "step": 7771
    },
    {
      "epoch": 1.8179160332124897,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 7772
    },
    {
      "epoch": 1.8181499239854988,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 7773
    },
    {
      "epoch": 1.8183838147585076,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 7774
    },
    {
      "epoch": 1.818617705531517,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 7775
    },
    {
      "epoch": 1.8188515963045258,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 7776
    },
    {
      "epoch": 1.8190854870775348,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 7777
    },
    {
      "epoch": 1.8193193778505439,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 7778
    },
    {
      "epoch": 1.8195532686235527,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 7779
    },
    {
      "epoch": 1.8197871593965618,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 7780
    },
    {
      "epoch": 1.8200210501695708,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 7781
    },
    {
      "epoch": 1.8202549409425797,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 7782
    },
    {
      "epoch": 1.820488831715589,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 7783
    },
    {
      "epoch": 1.8207227224885978,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 7784
    },
    {
      "epoch": 1.8209566132616068,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8411,
      "step": 7785
    },
    {
      "epoch": 1.821190504034616,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 7786
    },
    {
      "epoch": 1.8214243948076247,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 7787
    },
    {
      "epoch": 1.8216582855806338,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 7788
    },
    {
      "epoch": 1.8218921763536429,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 7789
    },
    {
      "epoch": 1.8221260671266517,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 7790
    },
    {
      "epoch": 1.822359957899661,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6728,
      "step": 7791
    },
    {
      "epoch": 1.8225938486726698,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 7792
    },
    {
      "epoch": 1.8228277394456789,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 7793
    },
    {
      "epoch": 1.823061630218688,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 7794
    },
    {
      "epoch": 1.8232955209916968,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8322,
      "step": 7795
    },
    {
      "epoch": 1.8235294117647058,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7499,
      "step": 7796
    },
    {
      "epoch": 1.823763302537715,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6119,
      "step": 7797
    },
    {
      "epoch": 1.8239971933107237,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6787,
      "step": 7798
    },
    {
      "epoch": 1.824231084083733,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.779,
      "step": 7799
    },
    {
      "epoch": 1.8244649748567419,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 7800
    },
    {
      "epoch": 1.8244649748567419,
      "eval_runtime": 4.6347,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 7800
    },
    {
      "epoch": 1.824698865629751,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 7801
    },
    {
      "epoch": 1.82493275640276,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 7802
    },
    {
      "epoch": 1.8251666471757688,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 7803
    },
    {
      "epoch": 1.8254005379487779,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 7804
    },
    {
      "epoch": 1.825634428721787,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 7805
    },
    {
      "epoch": 1.8258683194947958,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 7806
    },
    {
      "epoch": 1.826102210267805,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 7807
    },
    {
      "epoch": 1.826336101040814,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1518,
      "step": 7808
    },
    {
      "epoch": 1.826569991813823,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 7809
    },
    {
      "epoch": 1.826803882586832,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 7810
    },
    {
      "epoch": 1.8270377733598409,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 7811
    },
    {
      "epoch": 1.82727166413285,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 7812
    },
    {
      "epoch": 1.827505554905859,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 7813
    },
    {
      "epoch": 1.8277394456788678,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 7814
    },
    {
      "epoch": 1.827973336451877,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 7815
    },
    {
      "epoch": 1.828207227224886,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 7816
    },
    {
      "epoch": 1.828441117997895,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 7817
    },
    {
      "epoch": 1.828675008770904,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 7818
    },
    {
      "epoch": 1.828908899543913,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 7819
    },
    {
      "epoch": 1.829142790316922,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 7820
    },
    {
      "epoch": 1.829376681089931,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 7821
    },
    {
      "epoch": 1.8296105718629398,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 7822
    },
    {
      "epoch": 1.8298444626359491,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 7823
    },
    {
      "epoch": 1.830078353408958,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 7824
    },
    {
      "epoch": 1.830312244181967,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 7825
    },
    {
      "epoch": 1.830546134954976,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 7826
    },
    {
      "epoch": 1.830780025727985,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 7827
    },
    {
      "epoch": 1.831013916500994,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 7828
    },
    {
      "epoch": 1.831247807274003,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 7829
    },
    {
      "epoch": 1.8314816980470119,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 7830
    },
    {
      "epoch": 1.8317155888200212,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 7831
    },
    {
      "epoch": 1.83194947959303,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 7832
    },
    {
      "epoch": 1.832183370366039,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 7833
    },
    {
      "epoch": 1.8324172611390481,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 7834
    },
    {
      "epoch": 1.832651151912057,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 7835
    },
    {
      "epoch": 1.832885042685066,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 7836
    },
    {
      "epoch": 1.833118933458075,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 7837
    },
    {
      "epoch": 1.833352824231084,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7013,
      "step": 7838
    },
    {
      "epoch": 1.8335867150040932,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.654,
      "step": 7839
    },
    {
      "epoch": 1.833820605777102,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 7840
    },
    {
      "epoch": 1.834054496550111,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 7841
    },
    {
      "epoch": 1.8342883873231202,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 7842
    },
    {
      "epoch": 1.834522278096129,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 7843
    },
    {
      "epoch": 1.834756168869138,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 7844
    },
    {
      "epoch": 1.8349900596421471,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 7845
    },
    {
      "epoch": 1.835223950415156,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 7846
    },
    {
      "epoch": 1.8354578411881652,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 7847
    },
    {
      "epoch": 1.835691731961174,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1656,
      "step": 7848
    },
    {
      "epoch": 1.8359256227341831,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 7849
    },
    {
      "epoch": 1.8361595135071922,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6596,
      "step": 7850
    },
    {
      "epoch": 1.836393404280201,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 7851
    },
    {
      "epoch": 1.83662729505321,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1563,
      "step": 7852
    },
    {
      "epoch": 1.8368611858262192,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 7853
    },
    {
      "epoch": 1.837095076599228,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1805,
      "step": 7854
    },
    {
      "epoch": 1.8373289673722373,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 7855
    },
    {
      "epoch": 1.8375628581452461,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 7856
    },
    {
      "epoch": 1.8377967489182552,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 7857
    },
    {
      "epoch": 1.8380306396912642,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6727,
      "step": 7858
    },
    {
      "epoch": 1.838264530464273,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 7859
    },
    {
      "epoch": 1.8384984212372824,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 7860
    },
    {
      "epoch": 1.8387323120102912,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 7861
    },
    {
      "epoch": 1.8389662027833003,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 7862
    },
    {
      "epoch": 1.8392000935563093,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6838,
      "step": 7863
    },
    {
      "epoch": 1.8394339843293182,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 7864
    },
    {
      "epoch": 1.8396678751023272,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 7865
    },
    {
      "epoch": 1.8399017658753363,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 7866
    },
    {
      "epoch": 1.840135656648345,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 7867
    },
    {
      "epoch": 1.8403695474213544,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 7868
    },
    {
      "epoch": 1.8406034381943632,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 7869
    },
    {
      "epoch": 1.8408373289673723,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7554,
      "step": 7870
    },
    {
      "epoch": 1.8410712197403813,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 7871
    },
    {
      "epoch": 1.8413051105133902,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 7872
    },
    {
      "epoch": 1.8415390012863992,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 7873
    },
    {
      "epoch": 1.8417728920594083,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7467,
      "step": 7874
    },
    {
      "epoch": 1.8420067828324171,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 7875
    },
    {
      "epoch": 1.8422406736054264,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 7876
    },
    {
      "epoch": 1.8424745643784353,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 7877
    },
    {
      "epoch": 1.8427084551514443,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 7878
    },
    {
      "epoch": 1.8429423459244534,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 7879
    },
    {
      "epoch": 1.8431762366974622,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 7880
    },
    {
      "epoch": 1.8434101274704713,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 7881
    },
    {
      "epoch": 1.8436440182434803,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6152,
      "step": 7882
    },
    {
      "epoch": 1.8438779090164892,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 7883
    },
    {
      "epoch": 1.8441117997894985,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 7884
    },
    {
      "epoch": 1.8443456905625073,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 7885
    },
    {
      "epoch": 1.8445795813355164,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 7886
    },
    {
      "epoch": 1.8448134721085254,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 7887
    },
    {
      "epoch": 1.8450473628815343,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 7888
    },
    {
      "epoch": 1.8452812536545433,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 7889
    },
    {
      "epoch": 1.8455151444275524,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 7890
    },
    {
      "epoch": 1.8457490352005612,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 7891
    },
    {
      "epoch": 1.8459829259735705,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 7892
    },
    {
      "epoch": 1.8462168167465793,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 7893
    },
    {
      "epoch": 1.8464507075195884,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6223,
      "step": 7894
    },
    {
      "epoch": 1.8466845982925975,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 7895
    },
    {
      "epoch": 1.8469184890656063,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 7896
    },
    {
      "epoch": 1.8471523798386154,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 7897
    },
    {
      "epoch": 1.8473862706116244,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6412,
      "step": 7898
    },
    {
      "epoch": 1.8476201613846333,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 7899
    },
    {
      "epoch": 1.8478540521576425,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 7900
    },
    {
      "epoch": 1.8478540521576425,
      "eval_runtime": 4.6167,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 7900
    },
    {
      "epoch": 1.8480879429306514,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.4814,
      "step": 7901
    },
    {
      "epoch": 1.8483218337036604,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.4434,
      "step": 7902
    },
    {
      "epoch": 1.8485557244766695,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 7903
    },
    {
      "epoch": 1.8487896152496783,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 7904
    },
    {
      "epoch": 1.8490235060226874,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 7905
    },
    {
      "epoch": 1.8492573967956965,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 7906
    },
    {
      "epoch": 1.8494912875687053,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7223,
      "step": 7907
    },
    {
      "epoch": 1.8497251783417146,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 7908
    },
    {
      "epoch": 1.8499590691147234,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 7909
    },
    {
      "epoch": 1.8501929598877325,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 7910
    },
    {
      "epoch": 1.8504268506607415,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 7911
    },
    {
      "epoch": 1.8506607414337504,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 7912
    },
    {
      "epoch": 1.8508946322067594,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 7913
    },
    {
      "epoch": 1.8511285229797685,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 7914
    },
    {
      "epoch": 1.8513624137527773,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 7915
    },
    {
      "epoch": 1.8515963045257866,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 7916
    },
    {
      "epoch": 1.8518301952987954,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 7917
    },
    {
      "epoch": 1.8520640860718045,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 7918
    },
    {
      "epoch": 1.8522979768448136,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0266,
      "step": 7919
    },
    {
      "epoch": 1.8525318676178224,
      "grad_norm": 9.625,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 7920
    },
    {
      "epoch": 1.8527657583908315,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 7921
    },
    {
      "epoch": 1.8529996491638405,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 7922
    },
    {
      "epoch": 1.8532335399368494,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 7923
    },
    {
      "epoch": 1.8534674307098586,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 7924
    },
    {
      "epoch": 1.8537013214828675,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 7925
    },
    {
      "epoch": 1.8539352122558765,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8151,
      "step": 7926
    },
    {
      "epoch": 1.8541691030288856,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 7927
    },
    {
      "epoch": 1.8544029938018944,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1937,
      "step": 7928
    },
    {
      "epoch": 1.8546368845749035,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 7929
    },
    {
      "epoch": 1.8548707753479126,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7643,
      "step": 7930
    },
    {
      "epoch": 1.8551046661209214,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 7931
    },
    {
      "epoch": 1.8553385568939307,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 7932
    },
    {
      "epoch": 1.8555724476669395,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 7933
    },
    {
      "epoch": 1.8558063384399486,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 7934
    },
    {
      "epoch": 1.8560402292129576,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7037,
      "step": 7935
    },
    {
      "epoch": 1.8562741199859665,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 7936
    },
    {
      "epoch": 1.8565080107589755,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2022,
      "step": 7937
    },
    {
      "epoch": 1.8567419015319846,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 7938
    },
    {
      "epoch": 1.8569757923049934,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 7939
    },
    {
      "epoch": 1.8572096830780027,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 7940
    },
    {
      "epoch": 1.8574435738510116,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 7941
    },
    {
      "epoch": 1.8576774646240206,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 7942
    },
    {
      "epoch": 1.8579113553970297,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 7943
    },
    {
      "epoch": 1.8581452461700385,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 7944
    },
    {
      "epoch": 1.8583791369430476,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 7945
    },
    {
      "epoch": 1.8586130277160566,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 7946
    },
    {
      "epoch": 1.8588469184890655,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 7947
    },
    {
      "epoch": 1.8590808092620748,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 7948
    },
    {
      "epoch": 1.8593147000350836,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 7949
    },
    {
      "epoch": 1.8595485908080926,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 7950
    },
    {
      "epoch": 1.8597824815811017,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6004,
      "step": 7951
    },
    {
      "epoch": 1.8600163723541105,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 7952
    },
    {
      "epoch": 1.8602502631271196,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 7953
    },
    {
      "epoch": 1.8604841539001287,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 7954
    },
    {
      "epoch": 1.8607180446731375,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 7955
    },
    {
      "epoch": 1.8609519354461468,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 7956
    },
    {
      "epoch": 1.8611858262191556,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6203,
      "step": 7957
    },
    {
      "epoch": 1.8614197169921647,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7643,
      "step": 7958
    },
    {
      "epoch": 1.8616536077651737,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 7959
    },
    {
      "epoch": 1.8618874985381826,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 7960
    },
    {
      "epoch": 1.8621213893111916,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 7961
    },
    {
      "epoch": 1.8623552800842007,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 7962
    },
    {
      "epoch": 1.8625891708572095,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 7963
    },
    {
      "epoch": 1.8628230616302188,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 7964
    },
    {
      "epoch": 1.8630569524032277,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 7965
    },
    {
      "epoch": 1.8632908431762367,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.716,
      "step": 7966
    },
    {
      "epoch": 1.8635247339492458,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6621,
      "step": 7967
    },
    {
      "epoch": 1.8637586247222546,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 7968
    },
    {
      "epoch": 1.8639925154952637,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 7969
    },
    {
      "epoch": 1.8642264062682727,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 7970
    },
    {
      "epoch": 1.8644602970412816,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 7971
    },
    {
      "epoch": 1.8646941878142909,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 7972
    },
    {
      "epoch": 1.8649280785872997,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 7973
    },
    {
      "epoch": 1.8651619693603088,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 7974
    },
    {
      "epoch": 1.8653958601333178,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 7975
    },
    {
      "epoch": 1.8656297509063267,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 7976
    },
    {
      "epoch": 1.8658636416793357,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 7977
    },
    {
      "epoch": 1.8660975324523448,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 7978
    },
    {
      "epoch": 1.8663314232253536,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 7979
    },
    {
      "epoch": 1.866565313998363,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 7980
    },
    {
      "epoch": 1.8667992047713717,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 7981
    },
    {
      "epoch": 1.8670330955443808,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 7982
    },
    {
      "epoch": 1.8672669863173899,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 7983
    },
    {
      "epoch": 1.8675008770903987,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 7984
    },
    {
      "epoch": 1.8677347678634078,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 7985
    },
    {
      "epoch": 1.8679686586364168,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0983,
      "step": 7986
    },
    {
      "epoch": 1.8682025494094257,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 7987
    },
    {
      "epoch": 1.868436440182435,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 7988
    },
    {
      "epoch": 1.8686703309554438,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 7989
    },
    {
      "epoch": 1.8689042217284528,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 7990
    },
    {
      "epoch": 1.869138112501462,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 7991
    },
    {
      "epoch": 1.8693720032744707,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 7992
    },
    {
      "epoch": 1.8696058940474798,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6082,
      "step": 7993
    },
    {
      "epoch": 1.8698397848204888,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 7994
    },
    {
      "epoch": 1.8700736755934977,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 7995
    },
    {
      "epoch": 1.870307566366507,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 7996
    },
    {
      "epoch": 1.8705414571395158,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 7997
    },
    {
      "epoch": 1.8707753479125249,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 7998
    },
    {
      "epoch": 1.871009238685534,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7162,
      "step": 7999
    },
    {
      "epoch": 1.8712431294585428,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 8000
    },
    {
      "epoch": 1.8712431294585428,
      "eval_runtime": 4.5965,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 8000
    },
    {
      "epoch": 1.8714770202315518,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 8001
    },
    {
      "epoch": 1.8717109110045609,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 8002
    },
    {
      "epoch": 1.8719448017775697,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 8003
    },
    {
      "epoch": 1.872178692550579,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 8004
    },
    {
      "epoch": 1.8724125833235878,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 8005
    },
    {
      "epoch": 1.872646474096597,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.574,
      "step": 8006
    },
    {
      "epoch": 1.872880364869606,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 8007
    },
    {
      "epoch": 1.8731142556426148,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 8008
    },
    {
      "epoch": 1.8733481464156239,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 8009
    },
    {
      "epoch": 1.873582037188633,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8473,
      "step": 8010
    },
    {
      "epoch": 1.8738159279616418,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 8011
    },
    {
      "epoch": 1.874049818734651,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 8012
    },
    {
      "epoch": 1.8742837095076599,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 8013
    },
    {
      "epoch": 1.874517600280669,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 8014
    },
    {
      "epoch": 1.874751491053678,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 8015
    },
    {
      "epoch": 1.8749853818266868,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 8016
    },
    {
      "epoch": 1.875219272599696,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 8017
    },
    {
      "epoch": 1.875453163372705,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 8018
    },
    {
      "epoch": 1.8756870541457138,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 8019
    },
    {
      "epoch": 1.875920944918723,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 8020
    },
    {
      "epoch": 1.876154835691732,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 8021
    },
    {
      "epoch": 1.876388726464741,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 8022
    },
    {
      "epoch": 1.87662261723775,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 8023
    },
    {
      "epoch": 1.8768565080107589,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 8024
    },
    {
      "epoch": 1.877090398783768,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 8025
    },
    {
      "epoch": 1.877324289556777,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.078,
      "step": 8026
    },
    {
      "epoch": 1.8775581803297858,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 8027
    },
    {
      "epoch": 1.8777920711027951,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 8028
    },
    {
      "epoch": 1.878025961875804,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 8029
    },
    {
      "epoch": 1.878259852648813,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 8030
    },
    {
      "epoch": 1.878493743421822,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 8031
    },
    {
      "epoch": 1.878727634194831,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7694,
      "step": 8032
    },
    {
      "epoch": 1.87896152496784,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 8033
    },
    {
      "epoch": 1.879195415740849,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 8034
    },
    {
      "epoch": 1.8794293065138579,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 8035
    },
    {
      "epoch": 1.8796631972868671,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 8036
    },
    {
      "epoch": 1.879897088059876,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2683,
      "step": 8037
    },
    {
      "epoch": 1.880130978832885,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 8038
    },
    {
      "epoch": 1.880364869605894,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.5986,
      "step": 8039
    },
    {
      "epoch": 1.880598760378903,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5916,
      "step": 8040
    },
    {
      "epoch": 1.880832651151912,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6777,
      "step": 8041
    },
    {
      "epoch": 1.881066541924921,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.772,
      "step": 8042
    },
    {
      "epoch": 1.88130043269793,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 8043
    },
    {
      "epoch": 1.8815343234709392,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7981,
      "step": 8044
    },
    {
      "epoch": 1.881768214243948,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 8045
    },
    {
      "epoch": 1.882002105016957,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3429,
      "step": 8046
    },
    {
      "epoch": 1.8822359957899661,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7202,
      "step": 8047
    },
    {
      "epoch": 1.882469886562975,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 8048
    },
    {
      "epoch": 1.882703777335984,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 8049
    },
    {
      "epoch": 1.882937668108993,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6438,
      "step": 8050
    },
    {
      "epoch": 1.883171558882002,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 8051
    },
    {
      "epoch": 1.8834054496550112,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8531,
      "step": 8052
    },
    {
      "epoch": 1.88363934042802,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 8053
    },
    {
      "epoch": 1.8838732312010291,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 8054
    },
    {
      "epoch": 1.8841071219740382,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6976,
      "step": 8055
    },
    {
      "epoch": 1.884341012747047,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 8056
    },
    {
      "epoch": 1.8845749035200563,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 8057
    },
    {
      "epoch": 1.8848087942930651,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 8058
    },
    {
      "epoch": 1.8850426850660742,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6133,
      "step": 8059
    },
    {
      "epoch": 1.8852765758390833,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 8060
    },
    {
      "epoch": 1.885510466612092,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 8061
    },
    {
      "epoch": 1.8857443573851012,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 8062
    },
    {
      "epoch": 1.8859782481581102,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 8063
    },
    {
      "epoch": 1.886212138931119,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 8064
    },
    {
      "epoch": 1.8864460297041283,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 8065
    },
    {
      "epoch": 1.8866799204771372,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 8066
    },
    {
      "epoch": 1.8869138112501462,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.436,
      "step": 8067
    },
    {
      "epoch": 1.8871477020231553,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 8068
    },
    {
      "epoch": 1.8873815927961641,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 8069
    },
    {
      "epoch": 1.8876154835691732,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 8070
    },
    {
      "epoch": 1.8878493743421823,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 8071
    },
    {
      "epoch": 1.888083265115191,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6374,
      "step": 8072
    },
    {
      "epoch": 1.8883171558882004,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 8073
    },
    {
      "epoch": 1.8885510466612092,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.2134,
      "step": 8074
    },
    {
      "epoch": 1.8887849374342183,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 8075
    },
    {
      "epoch": 1.8890188282072273,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 8076
    },
    {
      "epoch": 1.8892527189802362,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 8077
    },
    {
      "epoch": 1.8894866097532452,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 8078
    },
    {
      "epoch": 1.8897205005262543,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2671,
      "step": 8079
    },
    {
      "epoch": 1.8899543912992631,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 8080
    },
    {
      "epoch": 1.8901882820722724,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 8081
    },
    {
      "epoch": 1.8904221728452812,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 8082
    },
    {
      "epoch": 1.8906560636182903,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 8083
    },
    {
      "epoch": 1.8908899543912994,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 8084
    },
    {
      "epoch": 1.8911238451643082,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 8085
    },
    {
      "epoch": 1.8913577359373173,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8846,
      "step": 8086
    },
    {
      "epoch": 1.8915916267103263,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7708,
      "step": 8087
    },
    {
      "epoch": 1.8918255174833352,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7054,
      "step": 8088
    },
    {
      "epoch": 1.8920594082563444,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 8089
    },
    {
      "epoch": 1.8922932990293533,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 8090
    },
    {
      "epoch": 1.8925271898023623,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8043,
      "step": 8091
    },
    {
      "epoch": 1.8927610805753714,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 8092
    },
    {
      "epoch": 1.8929949713483802,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 8093
    },
    {
      "epoch": 1.8932288621213893,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 8094
    },
    {
      "epoch": 1.8934627528943984,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 8095
    },
    {
      "epoch": 1.8936966436674072,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6639,
      "step": 8096
    },
    {
      "epoch": 1.8939305344404165,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 8097
    },
    {
      "epoch": 1.8941644252134253,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 8098
    },
    {
      "epoch": 1.8943983159864344,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 8099
    },
    {
      "epoch": 1.8946322067594434,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7101,
      "step": 8100
    },
    {
      "epoch": 1.8946322067594434,
      "eval_runtime": 4.5935,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 8100
    },
    {
      "epoch": 1.8948660975324523,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 8101
    },
    {
      "epoch": 1.8950999883054613,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 8102
    },
    {
      "epoch": 1.8953338790784704,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 8103
    },
    {
      "epoch": 1.8955677698514792,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 8104
    },
    {
      "epoch": 1.8958016606244885,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 8105
    },
    {
      "epoch": 1.8960355513974974,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 8106
    },
    {
      "epoch": 1.8962694421705064,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 8107
    },
    {
      "epoch": 1.8965033329435155,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 8108
    },
    {
      "epoch": 1.8967372237165243,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 8109
    },
    {
      "epoch": 1.8969711144895334,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 8110
    },
    {
      "epoch": 1.8972050052625424,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 8111
    },
    {
      "epoch": 1.8974388960355513,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 8112
    },
    {
      "epoch": 1.8976727868085606,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 8113
    },
    {
      "epoch": 1.8979066775815694,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 8114
    },
    {
      "epoch": 1.8981405683545785,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8508,
      "step": 8115
    },
    {
      "epoch": 1.8983744591275875,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 8116
    },
    {
      "epoch": 1.8986083499005963,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7827,
      "step": 8117
    },
    {
      "epoch": 1.8988422406736054,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 8118
    },
    {
      "epoch": 1.8990761314466145,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7419,
      "step": 8119
    },
    {
      "epoch": 1.8993100222196233,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 8120
    },
    {
      "epoch": 1.8995439129926326,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 8121
    },
    {
      "epoch": 1.8997778037656414,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7854,
      "step": 8122
    },
    {
      "epoch": 1.9000116945386505,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8089,
      "step": 8123
    },
    {
      "epoch": 1.9002455853116595,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 8124
    },
    {
      "epoch": 1.9004794760846684,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3513,
      "step": 8125
    },
    {
      "epoch": 1.9007133668576774,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 8126
    },
    {
      "epoch": 1.9009472576306865,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 8127
    },
    {
      "epoch": 1.9011811484036953,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 8128
    },
    {
      "epoch": 1.9014150391767046,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 8129
    },
    {
      "epoch": 1.9016489299497135,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 8130
    },
    {
      "epoch": 1.9018828207227225,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 8131
    },
    {
      "epoch": 1.9021167114957316,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 8132
    },
    {
      "epoch": 1.9023506022687404,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 8133
    },
    {
      "epoch": 1.9025844930417495,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 8134
    },
    {
      "epoch": 1.9028183838147585,
      "grad_norm": 7.75,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 8135
    },
    {
      "epoch": 1.9030522745877674,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 8136
    },
    {
      "epoch": 1.9032861653607767,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 8137
    },
    {
      "epoch": 1.9035200561337855,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 8138
    },
    {
      "epoch": 1.9037539469067946,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 8139
    },
    {
      "epoch": 1.9039878376798036,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 8140
    },
    {
      "epoch": 1.9042217284528125,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 8141
    },
    {
      "epoch": 1.9044556192258215,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 8142
    },
    {
      "epoch": 1.9046895099988306,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 8143
    },
    {
      "epoch": 1.9049234007718394,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 8144
    },
    {
      "epoch": 1.9051572915448487,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 8145
    },
    {
      "epoch": 1.9053911823178575,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 8146
    },
    {
      "epoch": 1.9056250730908666,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 8147
    },
    {
      "epoch": 1.9058589638638757,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 8148
    },
    {
      "epoch": 1.9060928546368845,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 8149
    },
    {
      "epoch": 1.9063267454098936,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3555,
      "step": 8150
    },
    {
      "epoch": 1.9065606361829026,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 8151
    },
    {
      "epoch": 1.9067945269559115,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 8152
    },
    {
      "epoch": 1.9070284177289207,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 8153
    },
    {
      "epoch": 1.9072623085019296,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6948,
      "step": 8154
    },
    {
      "epoch": 1.9074961992749386,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8451,
      "step": 8155
    },
    {
      "epoch": 1.9077300900479477,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 8156
    },
    {
      "epoch": 1.9079639808209565,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2203,
      "step": 8157
    },
    {
      "epoch": 1.9081978715939656,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 8158
    },
    {
      "epoch": 1.9084317623669746,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 8159
    },
    {
      "epoch": 1.9086656531399835,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 8160
    },
    {
      "epoch": 1.9088995439129928,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6071,
      "step": 8161
    },
    {
      "epoch": 1.9091334346860016,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 8162
    },
    {
      "epoch": 1.9093673254590107,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 8163
    },
    {
      "epoch": 1.9096012162320197,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7607,
      "step": 8164
    },
    {
      "epoch": 1.9098351070050286,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 8165
    },
    {
      "epoch": 1.9100689977780376,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 8166
    },
    {
      "epoch": 1.9103028885510467,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 8167
    },
    {
      "epoch": 1.9105367793240555,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1966,
      "step": 8168
    },
    {
      "epoch": 1.9107706700970648,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 8169
    },
    {
      "epoch": 1.9110045608700736,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2861,
      "step": 8170
    },
    {
      "epoch": 1.9112384516430827,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6918,
      "step": 8171
    },
    {
      "epoch": 1.9114723424160918,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5924,
      "step": 8172
    },
    {
      "epoch": 1.9117062331891006,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 8173
    },
    {
      "epoch": 1.9119401239621097,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6117,
      "step": 8174
    },
    {
      "epoch": 1.9121740147351187,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7045,
      "step": 8175
    },
    {
      "epoch": 1.9124079055081276,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 8176
    },
    {
      "epoch": 1.9126417962811368,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 8177
    },
    {
      "epoch": 1.9128756870541457,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 8178
    },
    {
      "epoch": 1.9131095778271547,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 8179
    },
    {
      "epoch": 1.9133434686001638,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 8180
    },
    {
      "epoch": 1.9135773593731726,
      "grad_norm": 8.1875,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 8181
    },
    {
      "epoch": 1.9138112501461817,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 8182
    },
    {
      "epoch": 1.9140451409191908,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 8183
    },
    {
      "epoch": 1.9142790316921996,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 8184
    },
    {
      "epoch": 1.9145129224652089,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 8185
    },
    {
      "epoch": 1.9147468132382177,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 8186
    },
    {
      "epoch": 1.9149807040112268,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 8187
    },
    {
      "epoch": 1.9152145947842358,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 8188
    },
    {
      "epoch": 1.9154484855572447,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 8189
    },
    {
      "epoch": 1.9156823763302537,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.73,
      "step": 8190
    },
    {
      "epoch": 1.9159162671032628,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 8191
    },
    {
      "epoch": 1.9161501578762716,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 8192
    },
    {
      "epoch": 1.916384048649281,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 8193
    },
    {
      "epoch": 1.9166179394222898,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 8194
    },
    {
      "epoch": 1.9168518301952988,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 8195
    },
    {
      "epoch": 1.9170857209683079,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 8196
    },
    {
      "epoch": 1.9173196117413167,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 8197
    },
    {
      "epoch": 1.9175535025143258,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 8198
    },
    {
      "epoch": 1.9177873932873348,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7558,
      "step": 8199
    },
    {
      "epoch": 1.9180212840603437,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5746,
      "step": 8200
    },
    {
      "epoch": 1.9180212840603437,
      "eval_runtime": 4.595,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 8200
    },
    {
      "epoch": 1.918255174833353,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7792,
      "step": 8201
    },
    {
      "epoch": 1.9184890656063618,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 8202
    },
    {
      "epoch": 1.9187229563793708,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 8203
    },
    {
      "epoch": 1.91895684715238,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1819,
      "step": 8204
    },
    {
      "epoch": 1.9191907379253887,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 8205
    },
    {
      "epoch": 1.9194246286983978,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 8206
    },
    {
      "epoch": 1.9196585194714069,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 8207
    },
    {
      "epoch": 1.9198924102444157,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 8208
    },
    {
      "epoch": 1.920126301017425,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.658,
      "step": 8209
    },
    {
      "epoch": 1.9203601917904338,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 8210
    },
    {
      "epoch": 1.9205940825634429,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 8211
    },
    {
      "epoch": 1.920827973336452,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 8212
    },
    {
      "epoch": 1.9210618641094608,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 8213
    },
    {
      "epoch": 1.9212957548824698,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 8214
    },
    {
      "epoch": 1.921529645655479,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3534,
      "step": 8215
    },
    {
      "epoch": 1.9217635364284877,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 8216
    },
    {
      "epoch": 1.921997427201497,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 8217
    },
    {
      "epoch": 1.9222313179745059,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6794,
      "step": 8218
    },
    {
      "epoch": 1.922465208747515,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2737,
      "step": 8219
    },
    {
      "epoch": 1.922699099520524,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 8220
    },
    {
      "epoch": 1.9229329902935328,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7643,
      "step": 8221
    },
    {
      "epoch": 1.9231668810665419,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 8222
    },
    {
      "epoch": 1.923400771839551,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 8223
    },
    {
      "epoch": 1.9236346626125598,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 8224
    },
    {
      "epoch": 1.923868553385569,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 8225
    },
    {
      "epoch": 1.924102444158578,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 8226
    },
    {
      "epoch": 1.924336334931587,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.571,
      "step": 8227
    },
    {
      "epoch": 1.924570225704596,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 8228
    },
    {
      "epoch": 1.9248041164776049,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 8229
    },
    {
      "epoch": 1.925038007250614,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8684,
      "step": 8230
    },
    {
      "epoch": 1.925271898023623,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 8231
    },
    {
      "epoch": 1.9255057887966318,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 8232
    },
    {
      "epoch": 1.925739679569641,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 8233
    },
    {
      "epoch": 1.92597357034265,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 8234
    },
    {
      "epoch": 1.926207461115659,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1604,
      "step": 8235
    },
    {
      "epoch": 1.926441351888668,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 8236
    },
    {
      "epoch": 1.926675242661677,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 8237
    },
    {
      "epoch": 1.926909133434686,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 8238
    },
    {
      "epoch": 1.927143024207695,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 8239
    },
    {
      "epoch": 1.9273769149807038,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 8240
    },
    {
      "epoch": 1.9276108057537131,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 8241
    },
    {
      "epoch": 1.927844696526722,
      "grad_norm": 8.0,
      "learning_rate": 3e-05,
      "loss": 2.4603,
      "step": 8242
    },
    {
      "epoch": 1.928078587299731,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 8243
    },
    {
      "epoch": 1.92831247807274,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 8244
    },
    {
      "epoch": 1.928546368845749,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8524,
      "step": 8245
    },
    {
      "epoch": 1.928780259618758,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 8246
    },
    {
      "epoch": 1.929014150391767,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 8247
    },
    {
      "epoch": 1.9292480411647759,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 8248
    },
    {
      "epoch": 1.9294819319377852,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 8249
    },
    {
      "epoch": 1.929715822710794,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6116,
      "step": 8250
    },
    {
      "epoch": 1.929949713483803,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 8251
    },
    {
      "epoch": 1.9301836042568121,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 8252
    },
    {
      "epoch": 1.930417495029821,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 8253
    },
    {
      "epoch": 1.9306513858028302,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7329,
      "step": 8254
    },
    {
      "epoch": 1.930885276575839,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9642,
      "step": 8255
    },
    {
      "epoch": 1.9311191673488481,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 8256
    },
    {
      "epoch": 1.9313530581218572,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.4821,
      "step": 8257
    },
    {
      "epoch": 1.931586948894866,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9538,
      "step": 8258
    },
    {
      "epoch": 1.931820839667875,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 8259
    },
    {
      "epoch": 1.9320547304408842,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7777,
      "step": 8260
    },
    {
      "epoch": 1.932288621213893,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 8261
    },
    {
      "epoch": 1.9325225119869023,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2759,
      "step": 8262
    },
    {
      "epoch": 1.9327564027599111,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 8263
    },
    {
      "epoch": 1.9329902935329202,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 8264
    },
    {
      "epoch": 1.9332241843059292,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 8265
    },
    {
      "epoch": 1.933458075078938,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 8266
    },
    {
      "epoch": 1.9336919658519471,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 8267
    },
    {
      "epoch": 1.9339258566249562,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 8268
    },
    {
      "epoch": 1.934159747397965,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 8269
    },
    {
      "epoch": 1.9343936381709743,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 8270
    },
    {
      "epoch": 1.9346275289439832,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 8271
    },
    {
      "epoch": 1.9348614197169922,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 8272
    },
    {
      "epoch": 1.9350953104900013,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 8273
    },
    {
      "epoch": 1.9353292012630101,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2394,
      "step": 8274
    },
    {
      "epoch": 1.9355630920360192,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 8275
    },
    {
      "epoch": 1.9357969828090282,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 8276
    },
    {
      "epoch": 1.936030873582037,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 8277
    },
    {
      "epoch": 1.9362647643550464,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7701,
      "step": 8278
    },
    {
      "epoch": 1.9364986551280552,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6278,
      "step": 8279
    },
    {
      "epoch": 1.9367325459010643,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 8280
    },
    {
      "epoch": 1.9369664366740733,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 8281
    },
    {
      "epoch": 1.9372003274470821,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 8282
    },
    {
      "epoch": 1.9374342182200912,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 8283
    },
    {
      "epoch": 1.9376681089931003,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 8284
    },
    {
      "epoch": 1.937901999766109,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 8285
    },
    {
      "epoch": 1.9381358905391184,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 8286
    },
    {
      "epoch": 1.9383697813121272,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 8287
    },
    {
      "epoch": 1.9386036720851363,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8058,
      "step": 8288
    },
    {
      "epoch": 1.9388375628581453,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 8289
    },
    {
      "epoch": 1.9390714536311542,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 8290
    },
    {
      "epoch": 1.9393053444041632,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 8291
    },
    {
      "epoch": 1.9395392351771723,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 8292
    },
    {
      "epoch": 1.9397731259501811,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.612,
      "step": 8293
    },
    {
      "epoch": 1.9400070167231904,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 8294
    },
    {
      "epoch": 1.9402409074961993,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 8295
    },
    {
      "epoch": 1.9404747982692083,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4007,
      "step": 8296
    },
    {
      "epoch": 1.9407086890422174,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 8297
    },
    {
      "epoch": 1.9409425798152262,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 8298
    },
    {
      "epoch": 1.9411764705882353,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 8299
    },
    {
      "epoch": 1.9414103613612443,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 8300
    },
    {
      "epoch": 1.9414103613612443,
      "eval_runtime": 4.5904,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 8300
    },
    {
      "epoch": 1.9416442521342532,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 8301
    },
    {
      "epoch": 1.9418781429072625,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 8302
    },
    {
      "epoch": 1.9421120336802713,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 8303
    },
    {
      "epoch": 1.9423459244532804,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 8304
    },
    {
      "epoch": 1.9425798152262894,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 8305
    },
    {
      "epoch": 1.9428137059992983,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 8306
    },
    {
      "epoch": 1.9430475967723073,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 8307
    },
    {
      "epoch": 1.9432814875453164,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 8308
    },
    {
      "epoch": 1.9435153783183252,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7547,
      "step": 8309
    },
    {
      "epoch": 1.9437492690913345,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 8310
    },
    {
      "epoch": 1.9439831598643433,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6824,
      "step": 8311
    },
    {
      "epoch": 1.9442170506373524,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 8312
    },
    {
      "epoch": 1.9444509414103615,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 8313
    },
    {
      "epoch": 1.9446848321833703,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 8314
    },
    {
      "epoch": 1.9449187229563794,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 8315
    },
    {
      "epoch": 1.9451526137293884,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5727,
      "step": 8316
    },
    {
      "epoch": 1.9453865045023973,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 8317
    },
    {
      "epoch": 1.9456203952754065,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 8318
    },
    {
      "epoch": 1.9458542860484154,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 8319
    },
    {
      "epoch": 1.9460881768214244,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 8320
    },
    {
      "epoch": 1.9463220675944335,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6233,
      "step": 8321
    },
    {
      "epoch": 1.9465559583674423,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 8322
    },
    {
      "epoch": 1.9467898491404514,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 8323
    },
    {
      "epoch": 1.9470237399134604,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 8324
    },
    {
      "epoch": 1.9472576306864693,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 8325
    },
    {
      "epoch": 1.9474915214594786,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 8326
    },
    {
      "epoch": 1.9477254122324874,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8531,
      "step": 8327
    },
    {
      "epoch": 1.9479593030054965,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 8328
    },
    {
      "epoch": 1.9481931937785055,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 8329
    },
    {
      "epoch": 1.9484270845515144,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2568,
      "step": 8330
    },
    {
      "epoch": 1.9486609753245234,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 8331
    },
    {
      "epoch": 1.9488948660975325,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 8332
    },
    {
      "epoch": 1.9491287568705413,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 8333
    },
    {
      "epoch": 1.9493626476435506,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3835,
      "step": 8334
    },
    {
      "epoch": 1.9495965384165594,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 8335
    },
    {
      "epoch": 1.9498304291895685,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 8336
    },
    {
      "epoch": 1.9500643199625776,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3701,
      "step": 8337
    },
    {
      "epoch": 1.9502982107355864,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 8338
    },
    {
      "epoch": 1.9505321015085955,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 8339
    },
    {
      "epoch": 1.9507659922816045,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 8340
    },
    {
      "epoch": 1.9509998830546134,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 8341
    },
    {
      "epoch": 1.9512337738276226,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7117,
      "step": 8342
    },
    {
      "epoch": 1.9514676646006315,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 8343
    },
    {
      "epoch": 1.9517015553736405,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 8344
    },
    {
      "epoch": 1.9519354461466496,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 8345
    },
    {
      "epoch": 1.9521693369196584,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 8346
    },
    {
      "epoch": 1.9524032276926675,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.5806,
      "step": 8347
    },
    {
      "epoch": 1.9526371184656766,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 8348
    },
    {
      "epoch": 1.9528710092386854,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 8349
    },
    {
      "epoch": 1.9531049000116947,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 8350
    },
    {
      "epoch": 1.9533387907847035,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 8351
    },
    {
      "epoch": 1.9535726815577126,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 8352
    },
    {
      "epoch": 1.9538065723307216,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7011,
      "step": 8353
    },
    {
      "epoch": 1.9540404631037305,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 8354
    },
    {
      "epoch": 1.9542743538767395,
      "grad_norm": 7.78125,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 8355
    },
    {
      "epoch": 1.9545082446497486,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2979,
      "step": 8356
    },
    {
      "epoch": 1.9547421354227574,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8028,
      "step": 8357
    },
    {
      "epoch": 1.9549760261957667,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 8358
    },
    {
      "epoch": 1.9552099169687756,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 8359
    },
    {
      "epoch": 1.9554438077417846,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 8360
    },
    {
      "epoch": 1.9556776985147937,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 8361
    },
    {
      "epoch": 1.9559115892878025,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 8362
    },
    {
      "epoch": 1.9561454800608116,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 8363
    },
    {
      "epoch": 1.9563793708338206,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 8364
    },
    {
      "epoch": 1.9566132616068295,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 8365
    },
    {
      "epoch": 1.9568471523798388,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7482,
      "step": 8366
    },
    {
      "epoch": 1.9570810431528476,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 8367
    },
    {
      "epoch": 1.9573149339258566,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 8368
    },
    {
      "epoch": 1.9575488246988657,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6956,
      "step": 8369
    },
    {
      "epoch": 1.9577827154718745,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 8370
    },
    {
      "epoch": 1.9580166062448836,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 8371
    },
    {
      "epoch": 1.9582504970178927,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 8372
    },
    {
      "epoch": 1.9584843877909015,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 8373
    },
    {
      "epoch": 1.9587182785639108,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 8374
    },
    {
      "epoch": 1.9589521693369196,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 8375
    },
    {
      "epoch": 1.9591860601099287,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 8376
    },
    {
      "epoch": 1.9594199508829377,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 8377
    },
    {
      "epoch": 1.9596538416559466,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 8378
    },
    {
      "epoch": 1.9598877324289556,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7238,
      "step": 8379
    },
    {
      "epoch": 1.9601216232019647,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 8380
    },
    {
      "epoch": 1.9603555139749735,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 8381
    },
    {
      "epoch": 1.9605894047479828,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 8382
    },
    {
      "epoch": 1.9608232955209917,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 8383
    },
    {
      "epoch": 1.9610571862940007,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 8384
    },
    {
      "epoch": 1.9612910770670098,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 8385
    },
    {
      "epoch": 1.9615249678400186,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 8386
    },
    {
      "epoch": 1.9617588586130277,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 8387
    },
    {
      "epoch": 1.9619927493860367,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.457,
      "step": 8388
    },
    {
      "epoch": 1.9622266401590456,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 8389
    },
    {
      "epoch": 1.9624605309320549,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 8390
    },
    {
      "epoch": 1.9626944217050637,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 8391
    },
    {
      "epoch": 1.9629283124780728,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 8392
    },
    {
      "epoch": 1.9631622032510818,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7453,
      "step": 8393
    },
    {
      "epoch": 1.9633960940240907,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 8394
    },
    {
      "epoch": 1.9636299847970997,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 8395
    },
    {
      "epoch": 1.9638638755701088,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 8396
    },
    {
      "epoch": 1.9640977663431176,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 8397
    },
    {
      "epoch": 1.964331657116127,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7999,
      "step": 8398
    },
    {
      "epoch": 1.9645655478891357,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 8399
    },
    {
      "epoch": 1.9647994386621448,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 8400
    },
    {
      "epoch": 1.9647994386621448,
      "eval_runtime": 4.6374,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 8400
    },
    {
      "epoch": 1.9650333294351539,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 8401
    },
    {
      "epoch": 1.9652672202081627,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 8402
    },
    {
      "epoch": 1.9655011109811718,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7005,
      "step": 8403
    },
    {
      "epoch": 1.9657350017541808,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 8404
    },
    {
      "epoch": 1.9659688925271896,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 8405
    },
    {
      "epoch": 1.966202783300199,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 8406
    },
    {
      "epoch": 1.9664366740732078,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 8407
    },
    {
      "epoch": 1.9666705648462168,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 8408
    },
    {
      "epoch": 1.9669044556192259,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 8409
    },
    {
      "epoch": 1.9671383463922347,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 8410
    },
    {
      "epoch": 1.9673722371652438,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8073,
      "step": 8411
    },
    {
      "epoch": 1.9676061279382528,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 8412
    },
    {
      "epoch": 1.9678400187112617,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 8413
    },
    {
      "epoch": 1.968073909484271,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 8414
    },
    {
      "epoch": 1.9683078002572798,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 8415
    },
    {
      "epoch": 1.9685416910302889,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7507,
      "step": 8416
    },
    {
      "epoch": 1.968775581803298,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8577,
      "step": 8417
    },
    {
      "epoch": 1.9690094725763068,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 8418
    },
    {
      "epoch": 1.9692433633493158,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 8419
    },
    {
      "epoch": 1.9694772541223249,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 8420
    },
    {
      "epoch": 1.9697111448953337,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 8421
    },
    {
      "epoch": 1.969945035668343,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 8422
    },
    {
      "epoch": 1.9701789264413518,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 8423
    },
    {
      "epoch": 1.970412817214361,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 8424
    },
    {
      "epoch": 1.97064670798737,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9638,
      "step": 8425
    },
    {
      "epoch": 1.9708805987603788,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 8426
    },
    {
      "epoch": 1.9711144895333879,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 8427
    },
    {
      "epoch": 1.971348380306397,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 8428
    },
    {
      "epoch": 1.9715822710794058,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 8429
    },
    {
      "epoch": 1.971816161852415,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 8430
    },
    {
      "epoch": 1.9720500526254239,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3366,
      "step": 8431
    },
    {
      "epoch": 1.972283943398433,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 8432
    },
    {
      "epoch": 1.972517834171442,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 8433
    },
    {
      "epoch": 1.9727517249444508,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 8434
    },
    {
      "epoch": 1.97298561571746,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 8435
    },
    {
      "epoch": 1.973219506490469,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 8436
    },
    {
      "epoch": 1.9734533972634778,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 8437
    },
    {
      "epoch": 1.973687288036487,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2269,
      "step": 8438
    },
    {
      "epoch": 1.973921178809496,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6526,
      "step": 8439
    },
    {
      "epoch": 1.974155069582505,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 8440
    },
    {
      "epoch": 1.974388960355514,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 8441
    },
    {
      "epoch": 1.9746228511285229,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 8442
    },
    {
      "epoch": 1.974856741901532,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 8443
    },
    {
      "epoch": 1.975090632674541,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8385,
      "step": 8444
    },
    {
      "epoch": 1.9753245234475498,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 8445
    },
    {
      "epoch": 1.9755584142205591,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2666,
      "step": 8446
    },
    {
      "epoch": 1.975792304993568,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 8447
    },
    {
      "epoch": 1.976026195766577,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 8448
    },
    {
      "epoch": 1.976260086539586,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 8449
    },
    {
      "epoch": 1.976493977312595,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 8450
    },
    {
      "epoch": 1.976727868085604,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 8451
    },
    {
      "epoch": 1.976961758858613,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 8452
    },
    {
      "epoch": 1.977195649631622,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 8453
    },
    {
      "epoch": 1.9774295404046311,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 8454
    },
    {
      "epoch": 1.97766343117764,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 8455
    },
    {
      "epoch": 1.977897321950649,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 8456
    },
    {
      "epoch": 1.978131212723658,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5108,
      "step": 8457
    },
    {
      "epoch": 1.978365103496667,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 8458
    },
    {
      "epoch": 1.9785989942696762,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2545,
      "step": 8459
    },
    {
      "epoch": 1.978832885042685,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 8460
    },
    {
      "epoch": 1.9790667758156941,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2676,
      "step": 8461
    },
    {
      "epoch": 1.9793006665887032,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9946,
      "step": 8462
    },
    {
      "epoch": 1.979534557361712,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1004,
      "step": 8463
    },
    {
      "epoch": 1.979768448134721,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 8464
    },
    {
      "epoch": 1.9800023389077301,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 8465
    },
    {
      "epoch": 1.980236229680739,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 8466
    },
    {
      "epoch": 1.9804701204537483,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 8467
    },
    {
      "epoch": 1.980704011226757,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 8468
    },
    {
      "epoch": 1.9809379019997662,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 8469
    },
    {
      "epoch": 1.9811717927727752,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 8470
    },
    {
      "epoch": 1.981405683545784,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 8471
    },
    {
      "epoch": 1.9816395743187931,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 8472
    },
    {
      "epoch": 1.9818734650918022,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6245,
      "step": 8473
    },
    {
      "epoch": 1.982107355864811,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 8474
    },
    {
      "epoch": 1.9823412466378203,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 8475
    },
    {
      "epoch": 1.9825751374108291,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3992,
      "step": 8476
    },
    {
      "epoch": 1.9828090281838382,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 8477
    },
    {
      "epoch": 1.9830429189568473,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 8478
    },
    {
      "epoch": 1.983276809729856,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 8479
    },
    {
      "epoch": 1.9835107005028652,
      "grad_norm": 11.6875,
      "learning_rate": 3e-05,
      "loss": 2.2893,
      "step": 8480
    },
    {
      "epoch": 1.9837445912758742,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 8481
    },
    {
      "epoch": 1.983978482048883,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 8482
    },
    {
      "epoch": 1.9842123728218923,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 8483
    },
    {
      "epoch": 1.9844462635949012,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 8484
    },
    {
      "epoch": 1.9846801543679102,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 8485
    },
    {
      "epoch": 1.9849140451409193,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 8486
    },
    {
      "epoch": 1.9851479359139281,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 8487
    },
    {
      "epoch": 1.9853818266869372,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 8488
    },
    {
      "epoch": 1.9856157174599463,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 8489
    },
    {
      "epoch": 1.985849608232955,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.636,
      "step": 8490
    },
    {
      "epoch": 1.9860834990059644,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 8491
    },
    {
      "epoch": 1.9863173897789732,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.573,
      "step": 8492
    },
    {
      "epoch": 1.9865512805519823,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 8493
    },
    {
      "epoch": 1.9867851713249913,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 8494
    },
    {
      "epoch": 1.9870190620980002,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 8495
    },
    {
      "epoch": 1.9872529528710092,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 8496
    },
    {
      "epoch": 1.9874868436440183,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 8497
    },
    {
      "epoch": 1.9877207344170271,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5904,
      "step": 8498
    },
    {
      "epoch": 1.9879546251900364,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 8499
    },
    {
      "epoch": 1.9881885159630452,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 8500
    },
    {
      "epoch": 1.9881885159630452,
      "eval_runtime": 4.6114,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 8500
    },
    {
      "epoch": 1.9884224067360543,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.581,
      "step": 8501
    },
    {
      "epoch": 1.9886562975090634,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 8502
    },
    {
      "epoch": 1.9888901882820722,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 8503
    },
    {
      "epoch": 1.9891240790550813,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 8504
    },
    {
      "epoch": 1.9893579698280903,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 8505
    },
    {
      "epoch": 1.9895918606010992,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 8506
    },
    {
      "epoch": 1.9898257513741084,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 8507
    },
    {
      "epoch": 1.9900596421471173,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 8508
    },
    {
      "epoch": 1.9902935329201263,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 8509
    },
    {
      "epoch": 1.9905274236931354,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.739,
      "step": 8510
    },
    {
      "epoch": 1.9907613144661442,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6164,
      "step": 8511
    },
    {
      "epoch": 1.9909952052391533,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 8512
    },
    {
      "epoch": 1.9912290960121624,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 8513
    },
    {
      "epoch": 1.9914629867851712,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 8514
    },
    {
      "epoch": 1.9916968775581805,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 8515
    },
    {
      "epoch": 1.9919307683311893,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 8516
    },
    {
      "epoch": 1.9921646591041984,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 8517
    },
    {
      "epoch": 1.9923985498772074,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7996,
      "step": 8518
    },
    {
      "epoch": 1.9926324406502163,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 8519
    },
    {
      "epoch": 1.9928663314232253,
      "grad_norm": 10.25,
      "learning_rate": 3e-05,
      "loss": 2.3224,
      "step": 8520
    },
    {
      "epoch": 1.9931002221962344,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 8521
    },
    {
      "epoch": 1.9933341129692432,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 8522
    },
    {
      "epoch": 1.9935680037422525,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 8523
    },
    {
      "epoch": 1.9938018945152614,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 8524
    },
    {
      "epoch": 1.9940357852882704,
      "grad_norm": 7.6875,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 8525
    },
    {
      "epoch": 1.9942696760612795,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7468,
      "step": 8526
    },
    {
      "epoch": 1.9945035668342883,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 8527
    },
    {
      "epoch": 1.9947374576072974,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 8528
    },
    {
      "epoch": 1.9949713483803064,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8122,
      "step": 8529
    },
    {
      "epoch": 1.9952052391533153,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 8530
    },
    {
      "epoch": 1.9954391299263246,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 8531
    },
    {
      "epoch": 1.9956730206993334,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 8532
    },
    {
      "epoch": 1.9959069114723424,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 8533
    },
    {
      "epoch": 1.9961408022453515,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 8534
    },
    {
      "epoch": 1.9963746930183603,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 8535
    },
    {
      "epoch": 1.9966085837913694,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6458,
      "step": 8536
    },
    {
      "epoch": 1.9968424745643785,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 8537
    },
    {
      "epoch": 1.9970763653373873,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 8538
    },
    {
      "epoch": 1.9973102561103966,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 8539
    },
    {
      "epoch": 1.9975441468834054,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 8540
    },
    {
      "epoch": 1.9977780376564145,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 8541
    },
    {
      "epoch": 1.9980119284294235,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 8542
    },
    {
      "epoch": 1.9982458192024324,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 8543
    },
    {
      "epoch": 1.9984797099754414,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 8544
    },
    {
      "epoch": 1.9987136007484505,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 8545
    },
    {
      "epoch": 1.9989474915214593,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 8546
    },
    {
      "epoch": 1.9991813822944686,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 8547
    },
    {
      "epoch": 1.9994152730674775,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7343,
      "step": 8548
    },
    {
      "epoch": 1.9996491638404865,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 8549
    },
    {
      "epoch": 1.9998830546134956,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 8550
    },
    {
      "epoch": 2.000233890773009,
      "grad_norm": 10.1875,
      "learning_rate": 3e-05,
      "loss": 3.9459,
      "step": 8551
    },
    {
      "epoch": 2.000467781546018,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 8552
    },
    {
      "epoch": 2.000701672319027,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 8553
    },
    {
      "epoch": 2.0009355630920362,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 8554
    },
    {
      "epoch": 2.001169453865045,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 8555
    },
    {
      "epoch": 2.001403344638054,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 8556
    },
    {
      "epoch": 2.001637235411063,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 8557
    },
    {
      "epoch": 2.001871126184072,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 8558
    },
    {
      "epoch": 2.002105016957081,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 8559
    },
    {
      "epoch": 2.00233890773009,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 8560
    },
    {
      "epoch": 2.002572798503099,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 8561
    },
    {
      "epoch": 2.0028066892761083,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 8562
    },
    {
      "epoch": 2.003040580049117,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 8563
    },
    {
      "epoch": 2.003274470822126,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7988,
      "step": 8564
    },
    {
      "epoch": 2.0035083615951352,
      "grad_norm": 8.3125,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 8565
    },
    {
      "epoch": 2.003742252368144,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 8566
    },
    {
      "epoch": 2.003976143141153,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7345,
      "step": 8567
    },
    {
      "epoch": 2.004210033914162,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 8568
    },
    {
      "epoch": 2.004443924687171,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 8569
    },
    {
      "epoch": 2.0046778154601803,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8038,
      "step": 8570
    },
    {
      "epoch": 2.004911706233189,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7758,
      "step": 8571
    },
    {
      "epoch": 2.005145597006198,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 8572
    },
    {
      "epoch": 2.0053794877792073,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 8573
    },
    {
      "epoch": 2.005613378552216,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 8574
    },
    {
      "epoch": 2.005847269325225,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7113,
      "step": 8575
    },
    {
      "epoch": 2.0060811600982342,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6696,
      "step": 8576
    },
    {
      "epoch": 2.006315050871243,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 8577
    },
    {
      "epoch": 2.0065489416442523,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 8578
    },
    {
      "epoch": 2.006782832417261,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6943,
      "step": 8579
    },
    {
      "epoch": 2.00701672319027,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 8580
    },
    {
      "epoch": 2.0072506139632793,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6873,
      "step": 8581
    },
    {
      "epoch": 2.007484504736288,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 8582
    },
    {
      "epoch": 2.007718395509297,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 8583
    },
    {
      "epoch": 2.0079522862823063,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 8584
    },
    {
      "epoch": 2.008186177055315,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 8585
    },
    {
      "epoch": 2.0084200678283244,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 8586
    },
    {
      "epoch": 2.0086539586013332,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.769,
      "step": 8587
    },
    {
      "epoch": 2.008887849374342,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 8588
    },
    {
      "epoch": 2.0091217401473513,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9209,
      "step": 8589
    },
    {
      "epoch": 2.00935563092036,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 8590
    },
    {
      "epoch": 2.009589521693369,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5087,
      "step": 8591
    },
    {
      "epoch": 2.0098234124663783,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 8592
    },
    {
      "epoch": 2.010057303239387,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 8593
    },
    {
      "epoch": 2.0102911940123964,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 8594
    },
    {
      "epoch": 2.0105250847854053,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 8595
    },
    {
      "epoch": 2.010758975558414,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 8596
    },
    {
      "epoch": 2.0109928663314234,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8476,
      "step": 8597
    },
    {
      "epoch": 2.011226757104432,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.5932,
      "step": 8598
    },
    {
      "epoch": 2.011460647877441,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7146,
      "step": 8599
    },
    {
      "epoch": 2.0116945386504503,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 8600
    },
    {
      "epoch": 2.0116945386504503,
      "eval_runtime": 4.6021,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 8600
    },
    {
      "epoch": 2.011928429423459,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 8601
    },
    {
      "epoch": 2.0121623201964685,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 8602
    },
    {
      "epoch": 2.0123962109694773,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.8183,
      "step": 8603
    },
    {
      "epoch": 2.012630101742486,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0639,
      "step": 8604
    },
    {
      "epoch": 2.0128639925154954,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 8605
    },
    {
      "epoch": 2.0130978832885043,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 8606
    },
    {
      "epoch": 2.013331774061513,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 8607
    },
    {
      "epoch": 2.0135656648345224,
      "grad_norm": 10.0,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 8608
    },
    {
      "epoch": 2.013799555607531,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 8609
    },
    {
      "epoch": 2.0140334463805405,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 8610
    },
    {
      "epoch": 2.0142673371535493,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6145,
      "step": 8611
    },
    {
      "epoch": 2.014501227926558,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 8612
    },
    {
      "epoch": 2.0147351186995675,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 8613
    },
    {
      "epoch": 2.0149690094725763,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 8614
    },
    {
      "epoch": 2.015202900245585,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 8615
    },
    {
      "epoch": 2.0154367910185944,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 8616
    },
    {
      "epoch": 2.0156706817916032,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 8617
    },
    {
      "epoch": 2.0159045725646125,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 8618
    },
    {
      "epoch": 2.0161384633376214,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 8619
    },
    {
      "epoch": 2.01637235411063,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 8620
    },
    {
      "epoch": 2.0166062448836395,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 8621
    },
    {
      "epoch": 2.0168401356566483,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 8622
    },
    {
      "epoch": 2.017074026429657,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 8623
    },
    {
      "epoch": 2.0173079172026664,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 8624
    },
    {
      "epoch": 2.0175418079756753,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 8625
    },
    {
      "epoch": 2.0177756987486846,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8203,
      "step": 8626
    },
    {
      "epoch": 2.0180095895216934,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 8627
    },
    {
      "epoch": 2.0182434802947022,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6394,
      "step": 8628
    },
    {
      "epoch": 2.0184773710677115,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3764,
      "step": 8629
    },
    {
      "epoch": 2.0187112618407204,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 8630
    },
    {
      "epoch": 2.018945152613729,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8846,
      "step": 8631
    },
    {
      "epoch": 2.0191790433867385,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 8632
    },
    {
      "epoch": 2.0194129341597473,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 8633
    },
    {
      "epoch": 2.0196468249327566,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 8634
    },
    {
      "epoch": 2.0198807157057654,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 8635
    },
    {
      "epoch": 2.0201146064787743,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 8636
    },
    {
      "epoch": 2.0203484972517836,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8751,
      "step": 8637
    },
    {
      "epoch": 2.0205823880247924,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1163,
      "step": 8638
    },
    {
      "epoch": 2.0208162787978012,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 8639
    },
    {
      "epoch": 2.0210501695708105,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 8640
    },
    {
      "epoch": 2.0212840603438194,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6893,
      "step": 8641
    },
    {
      "epoch": 2.0215179511168286,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 8642
    },
    {
      "epoch": 2.0217518418898375,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7398,
      "step": 8643
    },
    {
      "epoch": 2.0219857326628463,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6215,
      "step": 8644
    },
    {
      "epoch": 2.0222196234358556,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 8645
    },
    {
      "epoch": 2.0224535142088644,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9787,
      "step": 8646
    },
    {
      "epoch": 2.0226874049818733,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 8647
    },
    {
      "epoch": 2.0229212957548826,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 8648
    },
    {
      "epoch": 2.0231551865278914,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8776,
      "step": 8649
    },
    {
      "epoch": 2.0233890773009007,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6374,
      "step": 8650
    },
    {
      "epoch": 2.0236229680739095,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7284,
      "step": 8651
    },
    {
      "epoch": 2.0238568588469183,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 8652
    },
    {
      "epoch": 2.0240907496199276,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7741,
      "step": 8653
    },
    {
      "epoch": 2.0243246403929365,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 8654
    },
    {
      "epoch": 2.0245585311659453,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 8655
    },
    {
      "epoch": 2.0247924219389546,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 8656
    },
    {
      "epoch": 2.0250263127119634,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 8657
    },
    {
      "epoch": 2.0252602034849727,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 8658
    },
    {
      "epoch": 2.0254940942579815,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 8659
    },
    {
      "epoch": 2.0257279850309904,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 8660
    },
    {
      "epoch": 2.0259618758039997,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.506,
      "step": 8661
    },
    {
      "epoch": 2.0261957665770085,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 8662
    },
    {
      "epoch": 2.0264296573500173,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 8663
    },
    {
      "epoch": 2.0266635481230266,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 8664
    },
    {
      "epoch": 2.0268974388960355,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 8665
    },
    {
      "epoch": 2.0271313296690447,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 8666
    },
    {
      "epoch": 2.0273652204420536,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 8667
    },
    {
      "epoch": 2.0275991112150624,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 8668
    },
    {
      "epoch": 2.0278330019880717,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 8669
    },
    {
      "epoch": 2.0280668927610805,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 8670
    },
    {
      "epoch": 2.0283007835340894,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 8671
    },
    {
      "epoch": 2.0285346743070987,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 8672
    },
    {
      "epoch": 2.0287685650801075,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 8673
    },
    {
      "epoch": 2.029002455853117,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 8674
    },
    {
      "epoch": 2.0292363466261256,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6749,
      "step": 8675
    },
    {
      "epoch": 2.0294702373991345,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 8676
    },
    {
      "epoch": 2.0297041281721437,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 8677
    },
    {
      "epoch": 2.0299380189451526,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 8678
    },
    {
      "epoch": 2.0301719097181614,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 8679
    },
    {
      "epoch": 2.0304058004911707,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 8680
    },
    {
      "epoch": 2.0306396912641795,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 8681
    },
    {
      "epoch": 2.030873582037189,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.652,
      "step": 8682
    },
    {
      "epoch": 2.0311074728101977,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 8683
    },
    {
      "epoch": 2.0313413635832065,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 8684
    },
    {
      "epoch": 2.0315752543562158,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 8685
    },
    {
      "epoch": 2.0318091451292246,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 8686
    },
    {
      "epoch": 2.0320430359022335,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 8687
    },
    {
      "epoch": 2.0322769266752427,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 8688
    },
    {
      "epoch": 2.0325108174482516,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 8689
    },
    {
      "epoch": 2.032744708221261,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 8690
    },
    {
      "epoch": 2.0329785989942697,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3267,
      "step": 8691
    },
    {
      "epoch": 2.0332124897672785,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 8692
    },
    {
      "epoch": 2.033446380540288,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 8693
    },
    {
      "epoch": 2.0336802713132967,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 8694
    },
    {
      "epoch": 2.0339141620863055,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7088,
      "step": 8695
    },
    {
      "epoch": 2.0341480528593148,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 8696
    },
    {
      "epoch": 2.0343819436323236,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7908,
      "step": 8697
    },
    {
      "epoch": 2.034615834405333,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 8698
    },
    {
      "epoch": 2.0348497251783417,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 8699
    },
    {
      "epoch": 2.0350836159513506,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 8700
    },
    {
      "epoch": 2.0350836159513506,
      "eval_runtime": 4.6487,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 8700
    },
    {
      "epoch": 2.03531750672436,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 8701
    },
    {
      "epoch": 2.0355513974973687,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 8702
    },
    {
      "epoch": 2.0357852882703775,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9556,
      "step": 8703
    },
    {
      "epoch": 2.036019179043387,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 8704
    },
    {
      "epoch": 2.0362530698163956,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 8705
    },
    {
      "epoch": 2.036486960589405,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 8706
    },
    {
      "epoch": 2.0367208513624138,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 8707
    },
    {
      "epoch": 2.0369547421354226,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 8708
    },
    {
      "epoch": 2.037188632908432,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 8709
    },
    {
      "epoch": 2.0374225236814407,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 8710
    },
    {
      "epoch": 2.0376564144544496,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 8711
    },
    {
      "epoch": 2.037890305227459,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 8712
    },
    {
      "epoch": 2.0381241960004677,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 8713
    },
    {
      "epoch": 2.038358086773477,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 8714
    },
    {
      "epoch": 2.038591977546486,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5938,
      "step": 8715
    },
    {
      "epoch": 2.0388258683194946,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 8716
    },
    {
      "epoch": 2.039059759092504,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8501,
      "step": 8717
    },
    {
      "epoch": 2.0392936498655128,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2316,
      "step": 8718
    },
    {
      "epoch": 2.0395275406385216,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 8719
    },
    {
      "epoch": 2.039761431411531,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 8720
    },
    {
      "epoch": 2.0399953221845397,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6647,
      "step": 8721
    },
    {
      "epoch": 2.040229212957549,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.707,
      "step": 8722
    },
    {
      "epoch": 2.040463103730558,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 8723
    },
    {
      "epoch": 2.0406969945035667,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 8724
    },
    {
      "epoch": 2.040930885276576,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2542,
      "step": 8725
    },
    {
      "epoch": 2.041164776049585,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 8726
    },
    {
      "epoch": 2.0413986668225936,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 8727
    },
    {
      "epoch": 2.041632557595603,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 8728
    },
    {
      "epoch": 2.0418664483686118,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.77,
      "step": 8729
    },
    {
      "epoch": 2.042100339141621,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 8730
    },
    {
      "epoch": 2.04233422991463,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 8731
    },
    {
      "epoch": 2.0425681206876387,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 8732
    },
    {
      "epoch": 2.042802011460648,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 8733
    },
    {
      "epoch": 2.043035902233657,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 8734
    },
    {
      "epoch": 2.0432697930066657,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 8735
    },
    {
      "epoch": 2.043503683779675,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 8736
    },
    {
      "epoch": 2.043737574552684,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6136,
      "step": 8737
    },
    {
      "epoch": 2.043971465325693,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 8738
    },
    {
      "epoch": 2.044205356098702,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 8739
    },
    {
      "epoch": 2.0444392468717107,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 8740
    },
    {
      "epoch": 2.04467313764472,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 8741
    },
    {
      "epoch": 2.044907028417729,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 8742
    },
    {
      "epoch": 2.0451409191907377,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 8743
    },
    {
      "epoch": 2.045374809963747,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7954,
      "step": 8744
    },
    {
      "epoch": 2.045608700736756,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 8745
    },
    {
      "epoch": 2.045842591509765,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 8746
    },
    {
      "epoch": 2.046076482282774,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 8747
    },
    {
      "epoch": 2.046310373055783,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 8748
    },
    {
      "epoch": 2.046544263828792,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 8749
    },
    {
      "epoch": 2.046778154601801,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8738,
      "step": 8750
    },
    {
      "epoch": 2.04701204537481,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5886,
      "step": 8751
    },
    {
      "epoch": 2.047245936147819,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8329,
      "step": 8752
    },
    {
      "epoch": 2.047479826920828,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 8753
    },
    {
      "epoch": 2.047713717693837,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 8754
    },
    {
      "epoch": 2.047947608466846,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 8755
    },
    {
      "epoch": 2.048181499239855,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6532,
      "step": 8756
    },
    {
      "epoch": 2.048415390012864,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 8757
    },
    {
      "epoch": 2.048649280785873,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 8758
    },
    {
      "epoch": 2.0488831715588818,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 8759
    },
    {
      "epoch": 2.049117062331891,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 8760
    },
    {
      "epoch": 2.0493509531049,
      "grad_norm": 7.8125,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 8761
    },
    {
      "epoch": 2.049584843877909,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 8762
    },
    {
      "epoch": 2.049818734650918,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 8763
    },
    {
      "epoch": 2.050052625423927,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 8764
    },
    {
      "epoch": 2.050286516196936,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 8765
    },
    {
      "epoch": 2.050520406969945,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5783,
      "step": 8766
    },
    {
      "epoch": 2.0507542977429543,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 8767
    },
    {
      "epoch": 2.050988188515963,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 8768
    },
    {
      "epoch": 2.051222079288972,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 8769
    },
    {
      "epoch": 2.051455970061981,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0112,
      "step": 8770
    },
    {
      "epoch": 2.05168986083499,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6904,
      "step": 8771
    },
    {
      "epoch": 2.051923751607999,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 8772
    },
    {
      "epoch": 2.052157642381008,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6294,
      "step": 8773
    },
    {
      "epoch": 2.052391533154017,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 8774
    },
    {
      "epoch": 2.0526254239270263,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 8775
    },
    {
      "epoch": 2.052859314700035,
      "grad_norm": 8.0,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 8776
    },
    {
      "epoch": 2.053093205473044,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 8777
    },
    {
      "epoch": 2.0533270962460533,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 8778
    },
    {
      "epoch": 2.053560987019062,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 8779
    },
    {
      "epoch": 2.053794877792071,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 8780
    },
    {
      "epoch": 2.05402876856508,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 8781
    },
    {
      "epoch": 2.054262659338089,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 8782
    },
    {
      "epoch": 2.0544965501110983,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 8783
    },
    {
      "epoch": 2.054730440884107,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 8784
    },
    {
      "epoch": 2.054964331657116,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 8785
    },
    {
      "epoch": 2.0551982224301253,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 8786
    },
    {
      "epoch": 2.055432113203134,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 8787
    },
    {
      "epoch": 2.055666003976143,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 8788
    },
    {
      "epoch": 2.0558998947491522,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6287,
      "step": 8789
    },
    {
      "epoch": 2.056133785522161,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 8790
    },
    {
      "epoch": 2.0563676762951704,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 8791
    },
    {
      "epoch": 2.056601567068179,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 8792
    },
    {
      "epoch": 2.056835457841188,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 8793
    },
    {
      "epoch": 2.0570693486141973,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 8794
    },
    {
      "epoch": 2.057303239387206,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7573,
      "step": 8795
    },
    {
      "epoch": 2.057537130160215,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 8796
    },
    {
      "epoch": 2.0577710209332243,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6832,
      "step": 8797
    },
    {
      "epoch": 2.058004911706233,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 8798
    },
    {
      "epoch": 2.0582388024792424,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2988,
      "step": 8799
    },
    {
      "epoch": 2.0584726932522512,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 8800
    },
    {
      "epoch": 2.0584726932522512,
      "eval_runtime": 4.6144,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 8800
    },
    {
      "epoch": 2.05870658402526,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 8801
    },
    {
      "epoch": 2.0589404747982694,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5776,
      "step": 8802
    },
    {
      "epoch": 2.059174365571278,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 8803
    },
    {
      "epoch": 2.059408256344287,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 8804
    },
    {
      "epoch": 2.0596421471172963,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 8805
    },
    {
      "epoch": 2.059876037890305,
      "grad_norm": 8.375,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 8806
    },
    {
      "epoch": 2.0601099286633144,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 8807
    },
    {
      "epoch": 2.0603438194363233,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 8808
    },
    {
      "epoch": 2.060577710209332,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 8809
    },
    {
      "epoch": 2.0608116009823414,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 8810
    },
    {
      "epoch": 2.0610454917553502,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 8811
    },
    {
      "epoch": 2.061279382528359,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 8812
    },
    {
      "epoch": 2.0615132733013684,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 8813
    },
    {
      "epoch": 2.061747164074377,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 8814
    },
    {
      "epoch": 2.0619810548473865,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 8815
    },
    {
      "epoch": 2.0622149456203953,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 8816
    },
    {
      "epoch": 2.062448836393404,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 8817
    },
    {
      "epoch": 2.0626827271664134,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 8818
    },
    {
      "epoch": 2.0629166179394223,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7982,
      "step": 8819
    },
    {
      "epoch": 2.063150508712431,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 8820
    },
    {
      "epoch": 2.0633843994854404,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 8821
    },
    {
      "epoch": 2.0636182902584492,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 8822
    },
    {
      "epoch": 2.0638521810314585,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 8823
    },
    {
      "epoch": 2.0640860718044673,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 8824
    },
    {
      "epoch": 2.064319962577476,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 8825
    },
    {
      "epoch": 2.0645538533504855,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4797,
      "step": 8826
    },
    {
      "epoch": 2.0647877441234943,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5759,
      "step": 8827
    },
    {
      "epoch": 2.065021634896503,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 8828
    },
    {
      "epoch": 2.0652555256695124,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 8829
    },
    {
      "epoch": 2.0654894164425213,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 8830
    },
    {
      "epoch": 2.0657233072155305,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.688,
      "step": 8831
    },
    {
      "epoch": 2.0659571979885394,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 8832
    },
    {
      "epoch": 2.0661910887615482,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 8833
    },
    {
      "epoch": 2.0664249795345575,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 8834
    },
    {
      "epoch": 2.0666588703075663,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 8835
    },
    {
      "epoch": 2.066892761080575,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.4786,
      "step": 8836
    },
    {
      "epoch": 2.0671266518535845,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 8837
    },
    {
      "epoch": 2.0673605426265933,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7187,
      "step": 8838
    },
    {
      "epoch": 2.0675944333996026,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0364,
      "step": 8839
    },
    {
      "epoch": 2.0678283241726114,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 8840
    },
    {
      "epoch": 2.0680622149456203,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6819,
      "step": 8841
    },
    {
      "epoch": 2.0682961057186295,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 8842
    },
    {
      "epoch": 2.0685299964916384,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 8843
    },
    {
      "epoch": 2.068763887264647,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 8844
    },
    {
      "epoch": 2.0689977780376565,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7349,
      "step": 8845
    },
    {
      "epoch": 2.0692316688106653,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 8846
    },
    {
      "epoch": 2.0694655595836746,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5614,
      "step": 8847
    },
    {
      "epoch": 2.0696994503566835,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 8848
    },
    {
      "epoch": 2.0699333411296923,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 8849
    },
    {
      "epoch": 2.0701672319027016,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 8850
    },
    {
      "epoch": 2.0704011226757104,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 8851
    },
    {
      "epoch": 2.0706350134487193,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 8852
    },
    {
      "epoch": 2.0708689042217285,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 8853
    },
    {
      "epoch": 2.0711027949947374,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 8854
    },
    {
      "epoch": 2.0713366857677467,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 8855
    },
    {
      "epoch": 2.0715705765407555,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 8856
    },
    {
      "epoch": 2.0718044673137643,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 8857
    },
    {
      "epoch": 2.0720383580867736,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.5932,
      "step": 8858
    },
    {
      "epoch": 2.0722722488597825,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 8859
    },
    {
      "epoch": 2.0725061396327913,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 8860
    },
    {
      "epoch": 2.0727400304058006,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 8861
    },
    {
      "epoch": 2.0729739211788094,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5072,
      "step": 8862
    },
    {
      "epoch": 2.0732078119518187,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 8863
    },
    {
      "epoch": 2.0734417027248275,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 8864
    },
    {
      "epoch": 2.0736755934978364,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.678,
      "step": 8865
    },
    {
      "epoch": 2.0739094842708456,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6595,
      "step": 8866
    },
    {
      "epoch": 2.0741433750438545,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 8867
    },
    {
      "epoch": 2.0743772658168633,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6061,
      "step": 8868
    },
    {
      "epoch": 2.0746111565898726,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 8869
    },
    {
      "epoch": 2.0748450473628814,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 8870
    },
    {
      "epoch": 2.0750789381358907,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 8871
    },
    {
      "epoch": 2.0753128289088996,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 8872
    },
    {
      "epoch": 2.0755467196819084,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 8873
    },
    {
      "epoch": 2.0757806104549177,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 8874
    },
    {
      "epoch": 2.0760145012279265,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 8875
    },
    {
      "epoch": 2.0762483920009354,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 8876
    },
    {
      "epoch": 2.0764822827739446,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 8877
    },
    {
      "epoch": 2.0767161735469535,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 8878
    },
    {
      "epoch": 2.0769500643199628,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 8879
    },
    {
      "epoch": 2.0771839550929716,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 8880
    },
    {
      "epoch": 2.0774178458659804,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 8881
    },
    {
      "epoch": 2.0776517366389897,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 8882
    },
    {
      "epoch": 2.0778856274119986,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 8883
    },
    {
      "epoch": 2.0781195181850074,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.759,
      "step": 8884
    },
    {
      "epoch": 2.0783534089580167,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5757,
      "step": 8885
    },
    {
      "epoch": 2.0785872997310255,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 8886
    },
    {
      "epoch": 2.078821190504035,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 8887
    },
    {
      "epoch": 2.0790550812770436,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 8888
    },
    {
      "epoch": 2.0792889720500525,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5576,
      "step": 8889
    },
    {
      "epoch": 2.0795228628230618,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 8890
    },
    {
      "epoch": 2.0797567535960706,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 8891
    },
    {
      "epoch": 2.0799906443690794,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 8892
    },
    {
      "epoch": 2.0802245351420887,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 8893
    },
    {
      "epoch": 2.0804584259150976,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 8894
    },
    {
      "epoch": 2.080692316688107,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 8895
    },
    {
      "epoch": 2.0809262074611157,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 8896
    },
    {
      "epoch": 2.0811600982341245,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 8897
    },
    {
      "epoch": 2.081393989007134,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 8898
    },
    {
      "epoch": 2.0816278797801426,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 8899
    },
    {
      "epoch": 2.0818617705531515,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 8900
    },
    {
      "epoch": 2.0818617705531515,
      "eval_runtime": 4.6678,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 8900
    },
    {
      "epoch": 2.0820956613261608,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 8901
    },
    {
      "epoch": 2.0823295520991696,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 8902
    },
    {
      "epoch": 2.082563442872179,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 8903
    },
    {
      "epoch": 2.0827973336451877,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5394,
      "step": 8904
    },
    {
      "epoch": 2.0830312244181965,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 8905
    },
    {
      "epoch": 2.083265115191206,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.727,
      "step": 8906
    },
    {
      "epoch": 2.0834990059642147,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.748,
      "step": 8907
    },
    {
      "epoch": 2.0837328967372235,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 8908
    },
    {
      "epoch": 2.083966787510233,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 8909
    },
    {
      "epoch": 2.0842006782832416,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 8910
    },
    {
      "epoch": 2.084434569056251,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 8911
    },
    {
      "epoch": 2.0846684598292597,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 8912
    },
    {
      "epoch": 2.0849023506022686,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 8913
    },
    {
      "epoch": 2.085136241375278,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 8914
    },
    {
      "epoch": 2.0853701321482867,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 8915
    },
    {
      "epoch": 2.0856040229212955,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 8916
    },
    {
      "epoch": 2.085837913694305,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 8917
    },
    {
      "epoch": 2.0860718044673137,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 8918
    },
    {
      "epoch": 2.086305695240323,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 8919
    },
    {
      "epoch": 2.086539586013332,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 8920
    },
    {
      "epoch": 2.0867734767863406,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 8921
    },
    {
      "epoch": 2.08700736755935,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6273,
      "step": 8922
    },
    {
      "epoch": 2.0872412583323587,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 8923
    },
    {
      "epoch": 2.0874751491053676,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 8924
    },
    {
      "epoch": 2.087709039878377,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 8925
    },
    {
      "epoch": 2.0879429306513857,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7437,
      "step": 8926
    },
    {
      "epoch": 2.088176821424395,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4215,
      "step": 8927
    },
    {
      "epoch": 2.088410712197404,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 8928
    },
    {
      "epoch": 2.0886446029704127,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5623,
      "step": 8929
    },
    {
      "epoch": 2.088878493743422,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 8930
    },
    {
      "epoch": 2.0891123845164308,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 8931
    },
    {
      "epoch": 2.08934627528944,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 8932
    },
    {
      "epoch": 2.089580166062449,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7728,
      "step": 8933
    },
    {
      "epoch": 2.0898140568354577,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.5534,
      "step": 8934
    },
    {
      "epoch": 2.090047947608467,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 8935
    },
    {
      "epoch": 2.090281838381476,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 8936
    },
    {
      "epoch": 2.0905157291544847,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.838,
      "step": 8937
    },
    {
      "epoch": 2.090749619927494,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 8938
    },
    {
      "epoch": 2.090983510700503,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9096,
      "step": 8939
    },
    {
      "epoch": 2.0912174014735117,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 8940
    },
    {
      "epoch": 2.091451292246521,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 8941
    },
    {
      "epoch": 2.0916851830195298,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 8942
    },
    {
      "epoch": 2.091919073792539,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 8943
    },
    {
      "epoch": 2.092152964565548,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2607,
      "step": 8944
    },
    {
      "epoch": 2.0923868553385567,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 8945
    },
    {
      "epoch": 2.092620746111566,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 8946
    },
    {
      "epoch": 2.092854636884575,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 8947
    },
    {
      "epoch": 2.093088527657584,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 8948
    },
    {
      "epoch": 2.093322418430593,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 8949
    },
    {
      "epoch": 2.093556309203602,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6347,
      "step": 8950
    },
    {
      "epoch": 2.093790199976611,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 8951
    },
    {
      "epoch": 2.09402409074962,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 8952
    },
    {
      "epoch": 2.0942579815226288,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.844,
      "step": 8953
    },
    {
      "epoch": 2.094491872295638,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 8954
    },
    {
      "epoch": 2.094725763068647,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 8955
    },
    {
      "epoch": 2.0949596538416557,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 8956
    },
    {
      "epoch": 2.095193544614665,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4313,
      "step": 8957
    },
    {
      "epoch": 2.095427435387674,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 8958
    },
    {
      "epoch": 2.095661326160683,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 8959
    },
    {
      "epoch": 2.095895216933692,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5027,
      "step": 8960
    },
    {
      "epoch": 2.096129107706701,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 8961
    },
    {
      "epoch": 2.09636299847971,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 8962
    },
    {
      "epoch": 2.096596889252719,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 8963
    },
    {
      "epoch": 2.096830780025728,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 8964
    },
    {
      "epoch": 2.097064670798737,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7257,
      "step": 8965
    },
    {
      "epoch": 2.097298561571746,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 8966
    },
    {
      "epoch": 2.097532452344755,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7771,
      "step": 8967
    },
    {
      "epoch": 2.097766343117764,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6621,
      "step": 8968
    },
    {
      "epoch": 2.098000233890773,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 8969
    },
    {
      "epoch": 2.098234124663782,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 8970
    },
    {
      "epoch": 2.098468015436791,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 8971
    },
    {
      "epoch": 2.0987019062098,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 8972
    },
    {
      "epoch": 2.098935796982809,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 8973
    },
    {
      "epoch": 2.099169687755818,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 8974
    },
    {
      "epoch": 2.099403578528827,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 8975
    },
    {
      "epoch": 2.099637469301836,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 8976
    },
    {
      "epoch": 2.099871360074845,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 8977
    },
    {
      "epoch": 2.100105250847854,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 8978
    },
    {
      "epoch": 2.100339141620863,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 8979
    },
    {
      "epoch": 2.1005730323938723,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 8980
    },
    {
      "epoch": 2.100806923166881,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.116,
      "step": 8981
    },
    {
      "epoch": 2.10104081393989,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 8982
    },
    {
      "epoch": 2.1012747047128992,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2405,
      "step": 8983
    },
    {
      "epoch": 2.101508595485908,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 8984
    },
    {
      "epoch": 2.101742486258917,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 8985
    },
    {
      "epoch": 2.101976377031926,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 8986
    },
    {
      "epoch": 2.102210267804935,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 8987
    },
    {
      "epoch": 2.1024441585779443,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 8988
    },
    {
      "epoch": 2.102678049350953,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 8989
    },
    {
      "epoch": 2.102911940123962,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 8990
    },
    {
      "epoch": 2.1031458308969713,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 8991
    },
    {
      "epoch": 2.10337972166998,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 8992
    },
    {
      "epoch": 2.103613612442989,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.686,
      "step": 8993
    },
    {
      "epoch": 2.1038475032159982,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 8994
    },
    {
      "epoch": 2.104081393989007,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.347,
      "step": 8995
    },
    {
      "epoch": 2.1043152847620163,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 8996
    },
    {
      "epoch": 2.104549175535025,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 8997
    },
    {
      "epoch": 2.104783066308034,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8011,
      "step": 8998
    },
    {
      "epoch": 2.1050169570810433,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 8999
    },
    {
      "epoch": 2.105250847854052,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7617,
      "step": 9000
    },
    {
      "epoch": 2.105250847854052,
      "eval_runtime": 4.6032,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 9000
    },
    {
      "epoch": 2.105484738627061,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 9001
    },
    {
      "epoch": 2.1057186294000703,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 9002
    },
    {
      "epoch": 2.105952520173079,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 9003
    },
    {
      "epoch": 2.1061864109460884,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 9004
    },
    {
      "epoch": 2.106420301719097,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1556,
      "step": 9005
    },
    {
      "epoch": 2.106654192492106,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 9006
    },
    {
      "epoch": 2.1068880832651153,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 9007
    },
    {
      "epoch": 2.107121974038124,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 9008
    },
    {
      "epoch": 2.107355864811133,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 9009
    },
    {
      "epoch": 2.1075897555841423,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 9010
    },
    {
      "epoch": 2.107823646357151,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 9011
    },
    {
      "epoch": 2.1080575371301604,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8174,
      "step": 9012
    },
    {
      "epoch": 2.1082914279031693,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 9013
    },
    {
      "epoch": 2.108525318676178,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 9014
    },
    {
      "epoch": 2.1087592094491874,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 9015
    },
    {
      "epoch": 2.108993100222196,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 9016
    },
    {
      "epoch": 2.109226990995205,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 9017
    },
    {
      "epoch": 2.1094608817682143,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 9018
    },
    {
      "epoch": 2.109694772541223,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6293,
      "step": 9019
    },
    {
      "epoch": 2.1099286633142325,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 9020
    },
    {
      "epoch": 2.1101625540872413,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6823,
      "step": 9021
    },
    {
      "epoch": 2.11039644486025,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 9022
    },
    {
      "epoch": 2.1106303356332594,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 9023
    },
    {
      "epoch": 2.1108642264062683,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 9024
    },
    {
      "epoch": 2.111098117179277,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 9025
    },
    {
      "epoch": 2.1113320079522864,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 9026
    },
    {
      "epoch": 2.111565898725295,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5085,
      "step": 9027
    },
    {
      "epoch": 2.1117997894983045,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7099,
      "step": 9028
    },
    {
      "epoch": 2.1120336802713133,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 9029
    },
    {
      "epoch": 2.112267571044322,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 9030
    },
    {
      "epoch": 2.1125014618173314,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 9031
    },
    {
      "epoch": 2.1127353525903403,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6111,
      "step": 9032
    },
    {
      "epoch": 2.112969243363349,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 9033
    },
    {
      "epoch": 2.1132031341363584,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 9034
    },
    {
      "epoch": 2.1134370249093672,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5233,
      "step": 9035
    },
    {
      "epoch": 2.1136709156823765,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 9036
    },
    {
      "epoch": 2.1139048064553854,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 9037
    },
    {
      "epoch": 2.114138697228394,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 9038
    },
    {
      "epoch": 2.1143725880014035,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 9039
    },
    {
      "epoch": 2.1146064787744123,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.535,
      "step": 9040
    },
    {
      "epoch": 2.114840369547421,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 9041
    },
    {
      "epoch": 2.1150742603204304,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5981,
      "step": 9042
    },
    {
      "epoch": 2.1153081510934393,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 9043
    },
    {
      "epoch": 2.1155420418664486,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 9044
    },
    {
      "epoch": 2.1157759326394574,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2745,
      "step": 9045
    },
    {
      "epoch": 2.1160098234124662,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 9046
    },
    {
      "epoch": 2.1162437141854755,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 9047
    },
    {
      "epoch": 2.1164776049584844,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 9048
    },
    {
      "epoch": 2.116711495731493,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8252,
      "step": 9049
    },
    {
      "epoch": 2.1169453865045025,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 9050
    },
    {
      "epoch": 2.1171792772775113,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 9051
    },
    {
      "epoch": 2.1174131680505206,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 9052
    },
    {
      "epoch": 2.1176470588235294,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 9053
    },
    {
      "epoch": 2.1178809495965383,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 9054
    },
    {
      "epoch": 2.1181148403695476,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 9055
    },
    {
      "epoch": 2.1183487311425564,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 9056
    },
    {
      "epoch": 2.1185826219155652,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 9057
    },
    {
      "epoch": 2.1188165126885745,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6196,
      "step": 9058
    },
    {
      "epoch": 2.1190504034615834,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 9059
    },
    {
      "epoch": 2.1192842942345926,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 9060
    },
    {
      "epoch": 2.1195181850076015,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 9061
    },
    {
      "epoch": 2.1197520757806103,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 9062
    },
    {
      "epoch": 2.1199859665536196,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.6208,
      "step": 9063
    },
    {
      "epoch": 2.1202198573266284,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6673,
      "step": 9064
    },
    {
      "epoch": 2.1204537480996373,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6174,
      "step": 9065
    },
    {
      "epoch": 2.1206876388726466,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6486,
      "step": 9066
    },
    {
      "epoch": 2.1209215296456554,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 9067
    },
    {
      "epoch": 2.1211554204186647,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 9068
    },
    {
      "epoch": 2.1213893111916735,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9196,
      "step": 9069
    },
    {
      "epoch": 2.1216232019646823,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 9070
    },
    {
      "epoch": 2.1218570927376916,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 9071
    },
    {
      "epoch": 2.1220909835107005,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 9072
    },
    {
      "epoch": 2.1223248742837093,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.702,
      "step": 9073
    },
    {
      "epoch": 2.1225587650567186,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 9074
    },
    {
      "epoch": 2.1227926558297274,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 9075
    },
    {
      "epoch": 2.1230265466027367,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 9076
    },
    {
      "epoch": 2.1232604373757455,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7118,
      "step": 9077
    },
    {
      "epoch": 2.1234943281487544,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 9078
    },
    {
      "epoch": 2.1237282189217637,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 9079
    },
    {
      "epoch": 2.1239621096947725,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 9080
    },
    {
      "epoch": 2.1241960004677813,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 9081
    },
    {
      "epoch": 2.1244298912407906,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 9082
    },
    {
      "epoch": 2.1246637820137995,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6985,
      "step": 9083
    },
    {
      "epoch": 2.1248976727868087,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 9084
    },
    {
      "epoch": 2.1251315635598176,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5578,
      "step": 9085
    },
    {
      "epoch": 2.1253654543328264,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 9086
    },
    {
      "epoch": 2.1255993451058357,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5665,
      "step": 9087
    },
    {
      "epoch": 2.1258332358788445,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 9088
    },
    {
      "epoch": 2.1260671266518534,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 9089
    },
    {
      "epoch": 2.1263010174248627,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 9090
    },
    {
      "epoch": 2.1265349081978715,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 9091
    },
    {
      "epoch": 2.126768798970881,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 9092
    },
    {
      "epoch": 2.1270026897438896,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 9093
    },
    {
      "epoch": 2.1272365805168985,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 9094
    },
    {
      "epoch": 2.1274704712899077,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8903,
      "step": 9095
    },
    {
      "epoch": 2.1277043620629166,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 9096
    },
    {
      "epoch": 2.127938252835926,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 9097
    },
    {
      "epoch": 2.1281721436089347,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 9098
    },
    {
      "epoch": 2.1284060343819435,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 9099
    },
    {
      "epoch": 2.128639925154953,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 9100
    },
    {
      "epoch": 2.128639925154953,
      "eval_runtime": 4.6608,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 9100
    },
    {
      "epoch": 2.1288738159279617,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 9101
    },
    {
      "epoch": 2.1291077067009705,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 9102
    },
    {
      "epoch": 2.1293415974739798,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 9103
    },
    {
      "epoch": 2.1295754882469886,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 9104
    },
    {
      "epoch": 2.1298093790199975,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 9105
    },
    {
      "epoch": 2.1300432697930067,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 9106
    },
    {
      "epoch": 2.1302771605660156,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 9107
    },
    {
      "epoch": 2.130511051339025,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6175,
      "step": 9108
    },
    {
      "epoch": 2.1307449421120337,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 9109
    },
    {
      "epoch": 2.1309788328850425,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1974,
      "step": 9110
    },
    {
      "epoch": 2.131212723658052,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 9111
    },
    {
      "epoch": 2.1314466144310606,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 9112
    },
    {
      "epoch": 2.13168050520407,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 9113
    },
    {
      "epoch": 2.1319143959770788,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 9114
    },
    {
      "epoch": 2.1321482867500876,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 9115
    },
    {
      "epoch": 2.132382177523097,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 9116
    },
    {
      "epoch": 2.1326160682961057,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 9117
    },
    {
      "epoch": 2.1328499590691146,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 9118
    },
    {
      "epoch": 2.133083849842124,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6605,
      "step": 9119
    },
    {
      "epoch": 2.1333177406151327,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 9120
    },
    {
      "epoch": 2.1335516313881415,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 9121
    },
    {
      "epoch": 2.133785522161151,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7224,
      "step": 9122
    },
    {
      "epoch": 2.1340194129341596,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 9123
    },
    {
      "epoch": 2.134253303707169,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6388,
      "step": 9124
    },
    {
      "epoch": 2.1344871944801778,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.325,
      "step": 9125
    },
    {
      "epoch": 2.1347210852531866,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 9126
    },
    {
      "epoch": 2.134954976026196,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8462,
      "step": 9127
    },
    {
      "epoch": 2.1351888667992047,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 9128
    },
    {
      "epoch": 2.135422757572214,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 9129
    },
    {
      "epoch": 2.135656648345223,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 9130
    },
    {
      "epoch": 2.1358905391182317,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 9131
    },
    {
      "epoch": 2.136124429891241,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 9132
    },
    {
      "epoch": 2.13635832066425,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 9133
    },
    {
      "epoch": 2.1365922114372586,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 9134
    },
    {
      "epoch": 2.136826102210268,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 9135
    },
    {
      "epoch": 2.1370599929832768,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 9136
    },
    {
      "epoch": 2.1372938837562856,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 9137
    },
    {
      "epoch": 2.137527774529295,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 9138
    },
    {
      "epoch": 2.1377616653023037,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 9139
    },
    {
      "epoch": 2.137995556075313,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 9140
    },
    {
      "epoch": 2.138229446848322,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 9141
    },
    {
      "epoch": 2.1384633376213307,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 9142
    },
    {
      "epoch": 2.13869722839434,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 9143
    },
    {
      "epoch": 2.138931119167349,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 9144
    },
    {
      "epoch": 2.139165009940358,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.697,
      "step": 9145
    },
    {
      "epoch": 2.139398900713367,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6058,
      "step": 9146
    },
    {
      "epoch": 2.1396327914863758,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 9147
    },
    {
      "epoch": 2.139866682259385,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 9148
    },
    {
      "epoch": 2.140100573032394,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 9149
    },
    {
      "epoch": 2.1403344638054027,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 9150
    },
    {
      "epoch": 2.140568354578412,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 9151
    },
    {
      "epoch": 2.140802245351421,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 9152
    },
    {
      "epoch": 2.1410361361244297,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 9153
    },
    {
      "epoch": 2.141270026897439,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 9154
    },
    {
      "epoch": 2.141503917670448,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7576,
      "step": 9155
    },
    {
      "epoch": 2.141737808443457,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.697,
      "step": 9156
    },
    {
      "epoch": 2.141971699216466,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 9157
    },
    {
      "epoch": 2.1422055899894747,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 9158
    },
    {
      "epoch": 2.142439480762484,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 9159
    },
    {
      "epoch": 2.142673371535493,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 9160
    },
    {
      "epoch": 2.142907262308502,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 9161
    },
    {
      "epoch": 2.143141153081511,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 9162
    },
    {
      "epoch": 2.14337504385452,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2273,
      "step": 9163
    },
    {
      "epoch": 2.143608934627529,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 9164
    },
    {
      "epoch": 2.143842825400538,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 9165
    },
    {
      "epoch": 2.144076716173547,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 9166
    },
    {
      "epoch": 2.144310606946556,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 9167
    },
    {
      "epoch": 2.144544497719565,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.315,
      "step": 9168
    },
    {
      "epoch": 2.1447783884925737,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 9169
    },
    {
      "epoch": 2.145012279265583,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 9170
    },
    {
      "epoch": 2.145246170038592,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4145,
      "step": 9171
    },
    {
      "epoch": 2.145480060811601,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 9172
    },
    {
      "epoch": 2.14571395158461,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 9173
    },
    {
      "epoch": 2.145947842357619,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 9174
    },
    {
      "epoch": 2.146181733130628,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5905,
      "step": 9175
    },
    {
      "epoch": 2.146415623903637,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5648,
      "step": 9176
    },
    {
      "epoch": 2.146649514676646,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7185,
      "step": 9177
    },
    {
      "epoch": 2.146883405449655,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 9178
    },
    {
      "epoch": 2.147117296222664,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 9179
    },
    {
      "epoch": 2.147351186995673,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 9180
    },
    {
      "epoch": 2.147585077768682,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 9181
    },
    {
      "epoch": 2.147818968541691,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7365,
      "step": 9182
    },
    {
      "epoch": 2.1480528593147,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 9183
    },
    {
      "epoch": 2.148286750087709,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 9184
    },
    {
      "epoch": 2.148520640860718,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 9185
    },
    {
      "epoch": 2.148754531633727,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 9186
    },
    {
      "epoch": 2.148988422406736,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 9187
    },
    {
      "epoch": 2.149222313179745,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 9188
    },
    {
      "epoch": 2.149456203952754,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 9189
    },
    {
      "epoch": 2.149690094725763,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6941,
      "step": 9190
    },
    {
      "epoch": 2.149923985498772,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 9191
    },
    {
      "epoch": 2.150157876271781,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 9192
    },
    {
      "epoch": 2.1503917670447903,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8366,
      "step": 9193
    },
    {
      "epoch": 2.150625657817799,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 9194
    },
    {
      "epoch": 2.150859548590808,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6885,
      "step": 9195
    },
    {
      "epoch": 2.1510934393638173,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 9196
    },
    {
      "epoch": 2.151327330136826,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 9197
    },
    {
      "epoch": 2.151561220909835,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 9198
    },
    {
      "epoch": 2.151795111682844,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2145,
      "step": 9199
    },
    {
      "epoch": 2.152029002455853,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 9200
    },
    {
      "epoch": 2.152029002455853,
      "eval_runtime": 4.6523,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 9200
    },
    {
      "epoch": 2.1522628932288623,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.273,
      "step": 9201
    },
    {
      "epoch": 2.152496784001871,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 9202
    },
    {
      "epoch": 2.15273067477488,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3253,
      "step": 9203
    },
    {
      "epoch": 2.1529645655478893,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 9204
    },
    {
      "epoch": 2.153198456320898,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 9205
    },
    {
      "epoch": 2.153432347093907,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 9206
    },
    {
      "epoch": 2.1536662378669162,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 9207
    },
    {
      "epoch": 2.153900128639925,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 9208
    },
    {
      "epoch": 2.1541340194129344,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2263,
      "step": 9209
    },
    {
      "epoch": 2.154367910185943,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 9210
    },
    {
      "epoch": 2.154601800958952,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 9211
    },
    {
      "epoch": 2.1548356917319613,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 9212
    },
    {
      "epoch": 2.15506958250497,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 9213
    },
    {
      "epoch": 2.155303473277979,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 9214
    },
    {
      "epoch": 2.1555373640509883,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 9215
    },
    {
      "epoch": 2.155771254823997,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 9216
    },
    {
      "epoch": 2.1560051455970064,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 9217
    },
    {
      "epoch": 2.1562390363700152,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 9218
    },
    {
      "epoch": 2.156472927143024,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6776,
      "step": 9219
    },
    {
      "epoch": 2.1567068179160334,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 9220
    },
    {
      "epoch": 2.156940708689042,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.69,
      "step": 9221
    },
    {
      "epoch": 2.157174599462051,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 9222
    },
    {
      "epoch": 2.1574084902350603,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 9223
    },
    {
      "epoch": 2.157642381008069,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 9224
    },
    {
      "epoch": 2.1578762717810784,
      "grad_norm": 7.34375,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 9225
    },
    {
      "epoch": 2.1581101625540873,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 9226
    },
    {
      "epoch": 2.158344053327096,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 9227
    },
    {
      "epoch": 2.1585779441001054,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 9228
    },
    {
      "epoch": 2.1588118348731142,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 9229
    },
    {
      "epoch": 2.159045725646123,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.782,
      "step": 9230
    },
    {
      "epoch": 2.1592796164191324,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 9231
    },
    {
      "epoch": 2.159513507192141,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 9232
    },
    {
      "epoch": 2.1597473979651505,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 9233
    },
    {
      "epoch": 2.1599812887381593,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 9234
    },
    {
      "epoch": 2.160215179511168,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 9235
    },
    {
      "epoch": 2.1604490702841774,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 9236
    },
    {
      "epoch": 2.1606829610571863,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 9237
    },
    {
      "epoch": 2.160916851830195,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 9238
    },
    {
      "epoch": 2.1611507426032044,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 9239
    },
    {
      "epoch": 2.1613846333762132,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 9240
    },
    {
      "epoch": 2.1616185241492225,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 9241
    },
    {
      "epoch": 2.1618524149222313,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 9242
    },
    {
      "epoch": 2.16208630569524,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 9243
    },
    {
      "epoch": 2.1623201964682495,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 9244
    },
    {
      "epoch": 2.1625540872412583,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 9245
    },
    {
      "epoch": 2.162787978014267,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 9246
    },
    {
      "epoch": 2.1630218687872764,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 9247
    },
    {
      "epoch": 2.1632557595602853,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 9248
    },
    {
      "epoch": 2.1634896503332945,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 9249
    },
    {
      "epoch": 2.1637235411063034,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 9250
    },
    {
      "epoch": 2.163957431879312,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 9251
    },
    {
      "epoch": 2.1641913226523215,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 9252
    },
    {
      "epoch": 2.1644252134253303,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 9253
    },
    {
      "epoch": 2.164659104198339,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 9254
    },
    {
      "epoch": 2.1648929949713485,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 9255
    },
    {
      "epoch": 2.1651268857443573,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3962,
      "step": 9256
    },
    {
      "epoch": 2.1653607765173666,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 9257
    },
    {
      "epoch": 2.1655946672903754,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 9258
    },
    {
      "epoch": 2.1658285580633843,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 9259
    },
    {
      "epoch": 2.1660624488363935,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 9260
    },
    {
      "epoch": 2.1662963396094024,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 9261
    },
    {
      "epoch": 2.166530230382411,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6777,
      "step": 9262
    },
    {
      "epoch": 2.1667641211554205,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 9263
    },
    {
      "epoch": 2.1669980119284293,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 9264
    },
    {
      "epoch": 2.1672319027014386,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 9265
    },
    {
      "epoch": 2.1674657934744475,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6768,
      "step": 9266
    },
    {
      "epoch": 2.1676996842474563,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 9267
    },
    {
      "epoch": 2.1679335750204656,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 9268
    },
    {
      "epoch": 2.1681674657934744,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6464,
      "step": 9269
    },
    {
      "epoch": 2.1684013565664833,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 9270
    },
    {
      "epoch": 2.1686352473394925,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 9271
    },
    {
      "epoch": 2.1688691381125014,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 9272
    },
    {
      "epoch": 2.1691030288855107,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 9273
    },
    {
      "epoch": 2.1693369196585195,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 9274
    },
    {
      "epoch": 2.1695708104315283,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 9275
    },
    {
      "epoch": 2.1698047012045376,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.758,
      "step": 9276
    },
    {
      "epoch": 2.1700385919775464,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 9277
    },
    {
      "epoch": 2.1702724827505553,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 9278
    },
    {
      "epoch": 2.1705063735235646,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 9279
    },
    {
      "epoch": 2.1707402642965734,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6964,
      "step": 9280
    },
    {
      "epoch": 2.1709741550695827,
      "grad_norm": 9.0625,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 9281
    },
    {
      "epoch": 2.1712080458425915,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6148,
      "step": 9282
    },
    {
      "epoch": 2.1714419366156004,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 9283
    },
    {
      "epoch": 2.1716758273886096,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.886,
      "step": 9284
    },
    {
      "epoch": 2.1719097181616185,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 9285
    },
    {
      "epoch": 2.1721436089346273,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 9286
    },
    {
      "epoch": 2.1723774997076366,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 9287
    },
    {
      "epoch": 2.1726113904806454,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 9288
    },
    {
      "epoch": 2.1728452812536547,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 9289
    },
    {
      "epoch": 2.1730791720266636,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 9290
    },
    {
      "epoch": 2.1733130627996724,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7282,
      "step": 9291
    },
    {
      "epoch": 2.1735469535726817,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 9292
    },
    {
      "epoch": 2.1737808443456905,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7515,
      "step": 9293
    },
    {
      "epoch": 2.1740147351186994,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 9294
    },
    {
      "epoch": 2.1742486258917086,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 9295
    },
    {
      "epoch": 2.1744825166647175,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 9296
    },
    {
      "epoch": 2.1747164074377268,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 9297
    },
    {
      "epoch": 2.1749502982107356,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 9298
    },
    {
      "epoch": 2.1751841889837444,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 9299
    },
    {
      "epoch": 2.1754180797567537,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 9300
    },
    {
      "epoch": 2.1754180797567537,
      "eval_runtime": 4.5852,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 9300
    },
    {
      "epoch": 2.1756519705297626,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 9301
    },
    {
      "epoch": 2.1758858613027714,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 9302
    },
    {
      "epoch": 2.1761197520757807,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 9303
    },
    {
      "epoch": 2.1763536428487895,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 9304
    },
    {
      "epoch": 2.176587533621799,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 9305
    },
    {
      "epoch": 2.1768214243948076,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 9306
    },
    {
      "epoch": 2.1770553151678165,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 9307
    },
    {
      "epoch": 2.1772892059408258,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 9308
    },
    {
      "epoch": 2.1775230967138346,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 9309
    },
    {
      "epoch": 2.177756987486844,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 9310
    },
    {
      "epoch": 2.1779908782598527,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 9311
    },
    {
      "epoch": 2.1782247690328616,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 9312
    },
    {
      "epoch": 2.178458659805871,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5778,
      "step": 9313
    },
    {
      "epoch": 2.1786925505788797,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 9314
    },
    {
      "epoch": 2.1789264413518885,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.658,
      "step": 9315
    },
    {
      "epoch": 2.179160332124898,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 9316
    },
    {
      "epoch": 2.1793942228979066,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 9317
    },
    {
      "epoch": 2.1796281136709155,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 9318
    },
    {
      "epoch": 2.1798620044439248,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 9319
    },
    {
      "epoch": 2.1800958952169336,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 9320
    },
    {
      "epoch": 2.180329785989943,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.555,
      "step": 9321
    },
    {
      "epoch": 2.1805636767629517,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8607,
      "step": 9322
    },
    {
      "epoch": 2.1807975675359605,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 9323
    },
    {
      "epoch": 2.18103145830897,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 9324
    },
    {
      "epoch": 2.1812653490819787,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7439,
      "step": 9325
    },
    {
      "epoch": 2.181499239854988,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9622,
      "step": 9326
    },
    {
      "epoch": 2.181733130627997,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 9327
    },
    {
      "epoch": 2.1819670214010056,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 9328
    },
    {
      "epoch": 2.182200912174015,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 9329
    },
    {
      "epoch": 2.1824348029470237,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3189,
      "step": 9330
    },
    {
      "epoch": 2.1826686937200326,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 9331
    },
    {
      "epoch": 2.182902584493042,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7444,
      "step": 9332
    },
    {
      "epoch": 2.1831364752660507,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 9333
    },
    {
      "epoch": 2.1833703660390595,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.585,
      "step": 9334
    },
    {
      "epoch": 2.183604256812069,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 9335
    },
    {
      "epoch": 2.1838381475850777,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 9336
    },
    {
      "epoch": 2.184072038358087,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 9337
    },
    {
      "epoch": 2.184305929131096,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 9338
    },
    {
      "epoch": 2.1845398199041046,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 9339
    },
    {
      "epoch": 2.184773710677114,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 9340
    },
    {
      "epoch": 2.1850076014501227,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 9341
    },
    {
      "epoch": 2.185241492223132,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 9342
    },
    {
      "epoch": 2.185475382996141,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 9343
    },
    {
      "epoch": 2.1857092737691497,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 9344
    },
    {
      "epoch": 2.185943164542159,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6282,
      "step": 9345
    },
    {
      "epoch": 2.186177055315168,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.3535,
      "step": 9346
    },
    {
      "epoch": 2.1864109460881767,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 9347
    },
    {
      "epoch": 2.186644836861186,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7642,
      "step": 9348
    },
    {
      "epoch": 2.1868787276341948,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5628,
      "step": 9349
    },
    {
      "epoch": 2.1871126184072036,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 9350
    },
    {
      "epoch": 2.187346509180213,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 9351
    },
    {
      "epoch": 2.1875803999532217,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 9352
    },
    {
      "epoch": 2.187814290726231,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 9353
    },
    {
      "epoch": 2.18804818149924,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 9354
    },
    {
      "epoch": 2.1882820722722487,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7866,
      "step": 9355
    },
    {
      "epoch": 2.188515963045258,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 9356
    },
    {
      "epoch": 2.188749853818267,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 9357
    },
    {
      "epoch": 2.188983744591276,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 9358
    },
    {
      "epoch": 2.189217635364285,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 9359
    },
    {
      "epoch": 2.1894515261372938,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 9360
    },
    {
      "epoch": 2.189685416910303,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1115,
      "step": 9361
    },
    {
      "epoch": 2.189919307683312,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 9362
    },
    {
      "epoch": 2.1901531984563207,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1972,
      "step": 9363
    },
    {
      "epoch": 2.19038708922933,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 9364
    },
    {
      "epoch": 2.190620980002339,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 9365
    },
    {
      "epoch": 2.1908548707753477,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 9366
    },
    {
      "epoch": 2.191088761548357,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 9367
    },
    {
      "epoch": 2.191322652321366,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.4846,
      "step": 9368
    },
    {
      "epoch": 2.191556543094375,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 9369
    },
    {
      "epoch": 2.191790433867384,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 9370
    },
    {
      "epoch": 2.1920243246403928,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 9371
    },
    {
      "epoch": 2.192258215413402,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 9372
    },
    {
      "epoch": 2.192492106186411,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 9373
    },
    {
      "epoch": 2.19272599695942,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8693,
      "step": 9374
    },
    {
      "epoch": 2.192959887732429,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9786,
      "step": 9375
    },
    {
      "epoch": 2.193193778505438,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1761,
      "step": 9376
    },
    {
      "epoch": 2.193427669278447,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 9377
    },
    {
      "epoch": 2.193661560051456,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1767,
      "step": 9378
    },
    {
      "epoch": 2.193895450824465,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6854,
      "step": 9379
    },
    {
      "epoch": 2.194129341597474,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 9380
    },
    {
      "epoch": 2.194363232370483,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 9381
    },
    {
      "epoch": 2.1945971231434918,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 9382
    },
    {
      "epoch": 2.194831013916501,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 9383
    },
    {
      "epoch": 2.19506490468951,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 9384
    },
    {
      "epoch": 2.195298795462519,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 9385
    },
    {
      "epoch": 2.195532686235528,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 9386
    },
    {
      "epoch": 2.195766577008537,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 9387
    },
    {
      "epoch": 2.196000467781546,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.63,
      "step": 9388
    },
    {
      "epoch": 2.196234358554555,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 9389
    },
    {
      "epoch": 2.1964682493275642,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 9390
    },
    {
      "epoch": 2.196702140100573,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5827,
      "step": 9391
    },
    {
      "epoch": 2.196936030873582,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 9392
    },
    {
      "epoch": 2.197169921646591,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 9393
    },
    {
      "epoch": 2.1974038124196,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 9394
    },
    {
      "epoch": 2.197637703192609,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 9395
    },
    {
      "epoch": 2.197871593965618,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 9396
    },
    {
      "epoch": 2.198105484738627,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 9397
    },
    {
      "epoch": 2.1983393755116363,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 9398
    },
    {
      "epoch": 2.198573266284645,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 9399
    },
    {
      "epoch": 2.198807157057654,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 9400
    },
    {
      "epoch": 2.198807157057654,
      "eval_runtime": 4.6048,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 9400
    },
    {
      "epoch": 2.1990410478306632,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6907,
      "step": 9401
    },
    {
      "epoch": 2.199274938603672,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 9402
    },
    {
      "epoch": 2.199508829376681,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 9403
    },
    {
      "epoch": 2.19974272014969,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 9404
    },
    {
      "epoch": 2.199976610922699,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 9405
    },
    {
      "epoch": 2.2002105016957083,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1282,
      "step": 9406
    },
    {
      "epoch": 2.200444392468717,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 9407
    },
    {
      "epoch": 2.200678283241726,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 9408
    },
    {
      "epoch": 2.2009121740147353,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 9409
    },
    {
      "epoch": 2.201146064787744,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 9410
    },
    {
      "epoch": 2.201379955560753,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 9411
    },
    {
      "epoch": 2.2016138463337622,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 9412
    },
    {
      "epoch": 2.201847737106771,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 9413
    },
    {
      "epoch": 2.2020816278797803,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 9414
    },
    {
      "epoch": 2.202315518652789,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 9415
    },
    {
      "epoch": 2.202549409425798,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 9416
    },
    {
      "epoch": 2.2027833001988073,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2934,
      "step": 9417
    },
    {
      "epoch": 2.203017190971816,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7148,
      "step": 9418
    },
    {
      "epoch": 2.203251081744825,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4422,
      "step": 9419
    },
    {
      "epoch": 2.2034849725178343,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 9420
    },
    {
      "epoch": 2.203718863290843,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 9421
    },
    {
      "epoch": 2.2039527540638524,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 9422
    },
    {
      "epoch": 2.204186644836861,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 9423
    },
    {
      "epoch": 2.20442053560987,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 9424
    },
    {
      "epoch": 2.2046544263828793,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6131,
      "step": 9425
    },
    {
      "epoch": 2.204888317155888,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 9426
    },
    {
      "epoch": 2.205122207928897,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 9427
    },
    {
      "epoch": 2.2053560987019063,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 9428
    },
    {
      "epoch": 2.205589989474915,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3323,
      "step": 9429
    },
    {
      "epoch": 2.2058238802479244,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 9430
    },
    {
      "epoch": 2.2060577710209333,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.538,
      "step": 9431
    },
    {
      "epoch": 2.206291661793942,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 9432
    },
    {
      "epoch": 2.2065255525669514,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 9433
    },
    {
      "epoch": 2.20675944333996,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 9434
    },
    {
      "epoch": 2.206993334112969,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 9435
    },
    {
      "epoch": 2.2072272248859783,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 9436
    },
    {
      "epoch": 2.207461115658987,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 9437
    },
    {
      "epoch": 2.2076950064319965,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 9438
    },
    {
      "epoch": 2.2079288972050053,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 9439
    },
    {
      "epoch": 2.208162787978014,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 9440
    },
    {
      "epoch": 2.2083966787510234,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 9441
    },
    {
      "epoch": 2.2086305695240322,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 9442
    },
    {
      "epoch": 2.208864460297041,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 9443
    },
    {
      "epoch": 2.2090983510700504,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6477,
      "step": 9444
    },
    {
      "epoch": 2.209332241843059,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 9445
    },
    {
      "epoch": 2.2095661326160685,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 9446
    },
    {
      "epoch": 2.2098000233890773,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 9447
    },
    {
      "epoch": 2.210033914162086,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5483,
      "step": 9448
    },
    {
      "epoch": 2.2102678049350954,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 9449
    },
    {
      "epoch": 2.2105016957081043,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 9450
    },
    {
      "epoch": 2.210735586481113,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 9451
    },
    {
      "epoch": 2.2109694772541224,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 9452
    },
    {
      "epoch": 2.2112033680271312,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 9453
    },
    {
      "epoch": 2.2114372588001405,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.3413,
      "step": 9454
    },
    {
      "epoch": 2.2116711495731494,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 9455
    },
    {
      "epoch": 2.211905040346158,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 9456
    },
    {
      "epoch": 2.2121389311191675,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 9457
    },
    {
      "epoch": 2.2123728218921763,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 9458
    },
    {
      "epoch": 2.212606712665185,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 9459
    },
    {
      "epoch": 2.2128406034381944,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 9460
    },
    {
      "epoch": 2.2130744942112033,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9314,
      "step": 9461
    },
    {
      "epoch": 2.2133083849842126,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 9462
    },
    {
      "epoch": 2.2135422757572214,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 9463
    },
    {
      "epoch": 2.2137761665302302,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 9464
    },
    {
      "epoch": 2.2140100573032395,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 9465
    },
    {
      "epoch": 2.2142439480762484,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.3966,
      "step": 9466
    },
    {
      "epoch": 2.214477838849257,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 9467
    },
    {
      "epoch": 2.2147117296222665,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 9468
    },
    {
      "epoch": 2.2149456203952753,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 9469
    },
    {
      "epoch": 2.2151795111682846,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7568,
      "step": 9470
    },
    {
      "epoch": 2.2154134019412934,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3072,
      "step": 9471
    },
    {
      "epoch": 2.2156472927143023,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 9472
    },
    {
      "epoch": 2.2158811834873116,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 9473
    },
    {
      "epoch": 2.2161150742603204,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.5009,
      "step": 9474
    },
    {
      "epoch": 2.2163489650333292,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6334,
      "step": 9475
    },
    {
      "epoch": 2.2165828558063385,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 9476
    },
    {
      "epoch": 2.2168167465793474,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 9477
    },
    {
      "epoch": 2.2170506373523566,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 9478
    },
    {
      "epoch": 2.2172845281253655,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8902,
      "step": 9479
    },
    {
      "epoch": 2.2175184188983743,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 9480
    },
    {
      "epoch": 2.2177523096713836,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 9481
    },
    {
      "epoch": 2.2179862004443924,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 9482
    },
    {
      "epoch": 2.2182200912174013,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 9483
    },
    {
      "epoch": 2.2184539819904106,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 9484
    },
    {
      "epoch": 2.2186878727634194,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 9485
    },
    {
      "epoch": 2.2189217635364287,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 9486
    },
    {
      "epoch": 2.2191556543094375,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 9487
    },
    {
      "epoch": 2.2193895450824463,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 9488
    },
    {
      "epoch": 2.2196234358554556,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 9489
    },
    {
      "epoch": 2.2198573266284645,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 9490
    },
    {
      "epoch": 2.2200912174014733,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 9491
    },
    {
      "epoch": 2.2203251081744826,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 9492
    },
    {
      "epoch": 2.2205589989474914,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 9493
    },
    {
      "epoch": 2.2207928897205007,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 9494
    },
    {
      "epoch": 2.2210267804935095,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3513,
      "step": 9495
    },
    {
      "epoch": 2.2212606712665184,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 9496
    },
    {
      "epoch": 2.2214945620395277,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 9497
    },
    {
      "epoch": 2.2217284528125365,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 9498
    },
    {
      "epoch": 2.2219623435855453,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 9499
    },
    {
      "epoch": 2.2221962343585546,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 9500
    },
    {
      "epoch": 2.2221962343585546,
      "eval_runtime": 4.635,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 9500
    },
    {
      "epoch": 2.2224301251315635,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 9501
    },
    {
      "epoch": 2.2226640159045727,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 9502
    },
    {
      "epoch": 2.2228979066775816,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 9503
    },
    {
      "epoch": 2.2231317974505904,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 9504
    },
    {
      "epoch": 2.2233656882235997,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 9505
    },
    {
      "epoch": 2.2235995789966085,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 9506
    },
    {
      "epoch": 2.223833469769618,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8902,
      "step": 9507
    },
    {
      "epoch": 2.2240673605426267,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 9508
    },
    {
      "epoch": 2.2243012513156355,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.4984,
      "step": 9509
    },
    {
      "epoch": 2.224535142088645,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 9510
    },
    {
      "epoch": 2.2247690328616536,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 9511
    },
    {
      "epoch": 2.2250029236346625,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 9512
    },
    {
      "epoch": 2.2252368144076717,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 9513
    },
    {
      "epoch": 2.2254707051806806,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 9514
    },
    {
      "epoch": 2.2257045959536894,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9381,
      "step": 9515
    },
    {
      "epoch": 2.2259384867266987,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 9516
    },
    {
      "epoch": 2.2261723774997075,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 9517
    },
    {
      "epoch": 2.226406268272717,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7128,
      "step": 9518
    },
    {
      "epoch": 2.2266401590457257,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6907,
      "step": 9519
    },
    {
      "epoch": 2.2268740498187345,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 9520
    },
    {
      "epoch": 2.2271079405917438,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 9521
    },
    {
      "epoch": 2.2273418313647526,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 9522
    },
    {
      "epoch": 2.227575722137762,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 9523
    },
    {
      "epoch": 2.2278096129107707,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 9524
    },
    {
      "epoch": 2.2280435036837796,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 9525
    },
    {
      "epoch": 2.228277394456789,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 9526
    },
    {
      "epoch": 2.2285112852297977,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 9527
    },
    {
      "epoch": 2.2287451760028065,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 9528
    },
    {
      "epoch": 2.228979066775816,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 9529
    },
    {
      "epoch": 2.2292129575488246,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 9530
    },
    {
      "epoch": 2.2294468483218335,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 9531
    },
    {
      "epoch": 2.2296807390948428,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 9532
    },
    {
      "epoch": 2.2299146298678516,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 9533
    },
    {
      "epoch": 2.230148520640861,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 9534
    },
    {
      "epoch": 2.2303824114138697,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8945,
      "step": 9535
    },
    {
      "epoch": 2.2306163021868786,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 9536
    },
    {
      "epoch": 2.230850192959888,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 9537
    },
    {
      "epoch": 2.2310840837328967,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6203,
      "step": 9538
    },
    {
      "epoch": 2.231317974505906,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 9539
    },
    {
      "epoch": 2.231551865278915,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2688,
      "step": 9540
    },
    {
      "epoch": 2.2317857560519236,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 9541
    },
    {
      "epoch": 2.232019646824933,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 9542
    },
    {
      "epoch": 2.2322535375979418,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 9543
    },
    {
      "epoch": 2.2324874283709506,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 9544
    },
    {
      "epoch": 2.23272131914396,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 9545
    },
    {
      "epoch": 2.2329552099169687,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7416,
      "step": 9546
    },
    {
      "epoch": 2.2331891006899776,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 9547
    },
    {
      "epoch": 2.233422991462987,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.471,
      "step": 9548
    },
    {
      "epoch": 2.2336568822359957,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 9549
    },
    {
      "epoch": 2.233890773009005,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 9550
    },
    {
      "epoch": 2.234124663782014,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5585,
      "step": 9551
    },
    {
      "epoch": 2.2343585545550226,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 9552
    },
    {
      "epoch": 2.234592445328032,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 9553
    },
    {
      "epoch": 2.2348263361010408,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 9554
    },
    {
      "epoch": 2.23506022687405,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 9555
    },
    {
      "epoch": 2.235294117647059,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 9556
    },
    {
      "epoch": 2.2355280084200677,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.219,
      "step": 9557
    },
    {
      "epoch": 2.235761899193077,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8363,
      "step": 9558
    },
    {
      "epoch": 2.235995789966086,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.6774,
      "step": 9559
    },
    {
      "epoch": 2.2362296807390947,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 9560
    },
    {
      "epoch": 2.236463571512104,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 9561
    },
    {
      "epoch": 2.236697462285113,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.583,
      "step": 9562
    },
    {
      "epoch": 2.2369313530581216,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 9563
    },
    {
      "epoch": 2.237165243831131,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 9564
    },
    {
      "epoch": 2.2373991346041397,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 9565
    },
    {
      "epoch": 2.237633025377149,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 9566
    },
    {
      "epoch": 2.237866916150158,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 9567
    },
    {
      "epoch": 2.2381008069231667,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2253,
      "step": 9568
    },
    {
      "epoch": 2.238334697696176,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 9569
    },
    {
      "epoch": 2.238568588469185,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 9570
    },
    {
      "epoch": 2.238802479242194,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 9571
    },
    {
      "epoch": 2.239036370015203,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 9572
    },
    {
      "epoch": 2.239270260788212,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 9573
    },
    {
      "epoch": 2.239504151561221,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 9574
    },
    {
      "epoch": 2.23973804233423,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8234,
      "step": 9575
    },
    {
      "epoch": 2.2399719331072387,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 9576
    },
    {
      "epoch": 2.240205823880248,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.255,
      "step": 9577
    },
    {
      "epoch": 2.240439714653257,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 9578
    },
    {
      "epoch": 2.2406736054262657,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 9579
    },
    {
      "epoch": 2.240907496199275,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.78,
      "step": 9580
    },
    {
      "epoch": 2.241141386972284,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.885,
      "step": 9581
    },
    {
      "epoch": 2.241375277745293,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 9582
    },
    {
      "epoch": 2.241609168518302,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.5717,
      "step": 9583
    },
    {
      "epoch": 2.241843059291311,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 9584
    },
    {
      "epoch": 2.24207695006432,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 9585
    },
    {
      "epoch": 2.242310840837329,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4193,
      "step": 9586
    },
    {
      "epoch": 2.242544731610338,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 9587
    },
    {
      "epoch": 2.242778622383347,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 9588
    },
    {
      "epoch": 2.243012513156356,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 9589
    },
    {
      "epoch": 2.243246403929365,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 9590
    },
    {
      "epoch": 2.243480294702374,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 9591
    },
    {
      "epoch": 2.243714185475383,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8677,
      "step": 9592
    },
    {
      "epoch": 2.243948076248392,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 9593
    },
    {
      "epoch": 2.244181967021401,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 9594
    },
    {
      "epoch": 2.24441585779441,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 9595
    },
    {
      "epoch": 2.244649748567419,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6411,
      "step": 9596
    },
    {
      "epoch": 2.244883639340428,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6608,
      "step": 9597
    },
    {
      "epoch": 2.245117530113437,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 9598
    },
    {
      "epoch": 2.245351420886446,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6226,
      "step": 9599
    },
    {
      "epoch": 2.245585311659455,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6422,
      "step": 9600
    },
    {
      "epoch": 2.245585311659455,
      "eval_runtime": 4.652,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 9600
    },
    {
      "epoch": 2.245819202432464,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 9601
    },
    {
      "epoch": 2.246053093205473,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 9602
    },
    {
      "epoch": 2.2462869839784823,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 9603
    },
    {
      "epoch": 2.246520874751491,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 9604
    },
    {
      "epoch": 2.2467547655245,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 9605
    },
    {
      "epoch": 2.246988656297509,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 9606
    },
    {
      "epoch": 2.247222547070518,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 9607
    },
    {
      "epoch": 2.247456437843527,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3239,
      "step": 9608
    },
    {
      "epoch": 2.247690328616536,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 9609
    },
    {
      "epoch": 2.247924219389545,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 9610
    },
    {
      "epoch": 2.2481581101625543,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 9611
    },
    {
      "epoch": 2.248392000935563,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8165,
      "step": 9612
    },
    {
      "epoch": 2.248625891708572,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 9613
    },
    {
      "epoch": 2.2488597824815812,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 9614
    },
    {
      "epoch": 2.24909367325459,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 9615
    },
    {
      "epoch": 2.249327564027599,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 9616
    },
    {
      "epoch": 2.249561454800608,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 9617
    },
    {
      "epoch": 2.249795345573617,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 9618
    },
    {
      "epoch": 2.2500292363466263,
      "grad_norm": 7.625,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 9619
    },
    {
      "epoch": 2.250263127119635,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 9620
    },
    {
      "epoch": 2.250497017892644,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 9621
    },
    {
      "epoch": 2.2507309086656533,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 9622
    },
    {
      "epoch": 2.250964799438662,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 9623
    },
    {
      "epoch": 2.251198690211671,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 9624
    },
    {
      "epoch": 2.2514325809846802,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 9625
    },
    {
      "epoch": 2.251666471757689,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 9626
    },
    {
      "epoch": 2.251900362530698,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 9627
    },
    {
      "epoch": 2.252134253303707,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 9628
    },
    {
      "epoch": 2.252368144076716,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 9629
    },
    {
      "epoch": 2.2526020348497253,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 9630
    },
    {
      "epoch": 2.252835925622734,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6204,
      "step": 9631
    },
    {
      "epoch": 2.253069816395743,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9301,
      "step": 9632
    },
    {
      "epoch": 2.2533037071687523,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 9633
    },
    {
      "epoch": 2.253537597941761,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 9634
    },
    {
      "epoch": 2.2537714887147704,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 9635
    },
    {
      "epoch": 2.2540053794877792,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 9636
    },
    {
      "epoch": 2.254239270260788,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 9637
    },
    {
      "epoch": 2.2544731610337974,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 9638
    },
    {
      "epoch": 2.254707051806806,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 9639
    },
    {
      "epoch": 2.254940942579815,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.922,
      "step": 9640
    },
    {
      "epoch": 2.2551748333528243,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2198,
      "step": 9641
    },
    {
      "epoch": 2.255408724125833,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 9642
    },
    {
      "epoch": 2.255642614898842,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6464,
      "step": 9643
    },
    {
      "epoch": 2.2558765056718513,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 9644
    },
    {
      "epoch": 2.25611039644486,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 9645
    },
    {
      "epoch": 2.2563442872178694,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.6587,
      "step": 9646
    },
    {
      "epoch": 2.2565781779908782,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 9647
    },
    {
      "epoch": 2.256812068763887,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5393,
      "step": 9648
    },
    {
      "epoch": 2.2570459595368964,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 9649
    },
    {
      "epoch": 2.257279850309905,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 9650
    },
    {
      "epoch": 2.2575137410829145,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 9651
    },
    {
      "epoch": 2.2577476318559233,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6719,
      "step": 9652
    },
    {
      "epoch": 2.257981522628932,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 9653
    },
    {
      "epoch": 2.2582154134019414,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6648,
      "step": 9654
    },
    {
      "epoch": 2.2584493041749503,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 9655
    },
    {
      "epoch": 2.2586831949479595,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 9656
    },
    {
      "epoch": 2.2589170857209684,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 9657
    },
    {
      "epoch": 2.2591509764939772,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 9658
    },
    {
      "epoch": 2.2593848672669865,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 9659
    },
    {
      "epoch": 2.2596187580399953,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 9660
    },
    {
      "epoch": 2.259852648813004,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 9661
    },
    {
      "epoch": 2.2600865395860135,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 9662
    },
    {
      "epoch": 2.2603204303590223,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 9663
    },
    {
      "epoch": 2.260554321132031,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 9664
    },
    {
      "epoch": 2.2607882119050404,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 9665
    },
    {
      "epoch": 2.2610221026780493,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 9666
    },
    {
      "epoch": 2.2612559934510585,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 9667
    },
    {
      "epoch": 2.2614898842240674,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7148,
      "step": 9668
    },
    {
      "epoch": 2.261723774997076,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 9669
    },
    {
      "epoch": 2.2619576657700855,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7101,
      "step": 9670
    },
    {
      "epoch": 2.2621915565430943,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 9671
    },
    {
      "epoch": 2.2624254473161036,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4708,
      "step": 9672
    },
    {
      "epoch": 2.2626593380891125,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 9673
    },
    {
      "epoch": 2.2628932288621213,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 9674
    },
    {
      "epoch": 2.2631271196351306,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 9675
    },
    {
      "epoch": 2.2633610104081394,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6385,
      "step": 9676
    },
    {
      "epoch": 2.2635949011811483,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 9677
    },
    {
      "epoch": 2.2638287919541575,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9642,
      "step": 9678
    },
    {
      "epoch": 2.2640626827271664,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 9679
    },
    {
      "epoch": 2.264296573500175,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 9680
    },
    {
      "epoch": 2.2645304642731845,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 9681
    },
    {
      "epoch": 2.2647643550461933,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6526,
      "step": 9682
    },
    {
      "epoch": 2.2649982458192026,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2597,
      "step": 9683
    },
    {
      "epoch": 2.2652321365922115,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.772,
      "step": 9684
    },
    {
      "epoch": 2.2654660273652203,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 9685
    },
    {
      "epoch": 2.2656999181382296,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 9686
    },
    {
      "epoch": 2.2659338089112384,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8905,
      "step": 9687
    },
    {
      "epoch": 2.2661676996842477,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 9688
    },
    {
      "epoch": 2.2664015904572565,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 9689
    },
    {
      "epoch": 2.2666354812302654,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7266,
      "step": 9690
    },
    {
      "epoch": 2.2668693720032747,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 9691
    },
    {
      "epoch": 2.2671032627762835,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 9692
    },
    {
      "epoch": 2.2673371535492923,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 9693
    },
    {
      "epoch": 2.2675710443223016,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 9694
    },
    {
      "epoch": 2.2678049350953104,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 9695
    },
    {
      "epoch": 2.2680388258683193,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 9696
    },
    {
      "epoch": 2.2682727166413286,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 9697
    },
    {
      "epoch": 2.2685066074143374,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 9698
    },
    {
      "epoch": 2.2687404981873467,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 9699
    },
    {
      "epoch": 2.2689743889603555,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7349,
      "step": 9700
    },
    {
      "epoch": 2.2689743889603555,
      "eval_runtime": 4.6149,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 9700
    },
    {
      "epoch": 2.2692082797333644,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 9701
    },
    {
      "epoch": 2.2694421705063736,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 9702
    },
    {
      "epoch": 2.2696760612793825,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 9703
    },
    {
      "epoch": 2.2699099520523918,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 9704
    },
    {
      "epoch": 2.2701438428254006,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 9705
    },
    {
      "epoch": 2.2703777335984094,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9184,
      "step": 9706
    },
    {
      "epoch": 2.2706116243714187,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 9707
    },
    {
      "epoch": 2.2708455151444276,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 9708
    },
    {
      "epoch": 2.2710794059174364,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 9709
    },
    {
      "epoch": 2.2713132966904457,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 9710
    },
    {
      "epoch": 2.2715471874634545,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6698,
      "step": 9711
    },
    {
      "epoch": 2.2717810782364634,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 9712
    },
    {
      "epoch": 2.2720149690094726,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 9713
    },
    {
      "epoch": 2.2722488597824815,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 1.9001,
      "step": 9714
    },
    {
      "epoch": 2.2724827505554908,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 9715
    },
    {
      "epoch": 2.2727166413284996,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 9716
    },
    {
      "epoch": 2.2729505321015084,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7359,
      "step": 9717
    },
    {
      "epoch": 2.2731844228745177,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 9718
    },
    {
      "epoch": 2.2734183136475266,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 9719
    },
    {
      "epoch": 2.273652204420536,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 9720
    },
    {
      "epoch": 2.2738860951935447,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 9721
    },
    {
      "epoch": 2.2741199859665535,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 9722
    },
    {
      "epoch": 2.274353876739563,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 9723
    },
    {
      "epoch": 2.2745877675125716,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 9724
    },
    {
      "epoch": 2.2748216582855805,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 9725
    },
    {
      "epoch": 2.2750555490585898,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 9726
    },
    {
      "epoch": 2.2752894398315986,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 9727
    },
    {
      "epoch": 2.2755233306046074,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 9728
    },
    {
      "epoch": 2.2757572213776167,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 9729
    },
    {
      "epoch": 2.2759911121506256,
      "grad_norm": 7.71875,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 9730
    },
    {
      "epoch": 2.276225002923635,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 9731
    },
    {
      "epoch": 2.2764588936966437,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 9732
    },
    {
      "epoch": 2.2766927844696525,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 9733
    },
    {
      "epoch": 2.276926675242662,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 9734
    },
    {
      "epoch": 2.2771605660156706,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 9735
    },
    {
      "epoch": 2.27739445678868,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 9736
    },
    {
      "epoch": 2.2776283475616887,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 9737
    },
    {
      "epoch": 2.2778622383346976,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 9738
    },
    {
      "epoch": 2.278096129107707,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 9739
    },
    {
      "epoch": 2.2783300198807157,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 9740
    },
    {
      "epoch": 2.2785639106537245,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 9741
    },
    {
      "epoch": 2.278797801426734,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7657,
      "step": 9742
    },
    {
      "epoch": 2.2790316921997427,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7527,
      "step": 9743
    },
    {
      "epoch": 2.2792655829727515,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 9744
    },
    {
      "epoch": 2.279499473745761,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6582,
      "step": 9745
    },
    {
      "epoch": 2.2797333645187696,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 9746
    },
    {
      "epoch": 2.279967255291779,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 9747
    },
    {
      "epoch": 2.2802011460647877,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 9748
    },
    {
      "epoch": 2.2804350368377966,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 9749
    },
    {
      "epoch": 2.280668927610806,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 9750
    },
    {
      "epoch": 2.2809028183838147,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 9751
    },
    {
      "epoch": 2.281136709156824,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 9752
    },
    {
      "epoch": 2.281370599929833,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 9753
    },
    {
      "epoch": 2.2816044907028417,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 9754
    },
    {
      "epoch": 2.281838381475851,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 9755
    },
    {
      "epoch": 2.28207227224886,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 9756
    },
    {
      "epoch": 2.2823061630218686,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 9757
    },
    {
      "epoch": 2.282540053794878,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 9758
    },
    {
      "epoch": 2.2827739445678867,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 9759
    },
    {
      "epoch": 2.2830078353408956,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 9760
    },
    {
      "epoch": 2.283241726113905,
      "grad_norm": 7.6875,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 9761
    },
    {
      "epoch": 2.2834756168869137,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 9762
    },
    {
      "epoch": 2.283709507659923,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7306,
      "step": 9763
    },
    {
      "epoch": 2.283943398432932,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 9764
    },
    {
      "epoch": 2.2841772892059407,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 9765
    },
    {
      "epoch": 2.28441117997895,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 9766
    },
    {
      "epoch": 2.2846450707519588,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 9767
    },
    {
      "epoch": 2.284878961524968,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 9768
    },
    {
      "epoch": 2.285112852297977,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 9769
    },
    {
      "epoch": 2.2853467430709857,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 9770
    },
    {
      "epoch": 2.285580633843995,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.4239,
      "step": 9771
    },
    {
      "epoch": 2.285814524617004,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 9772
    },
    {
      "epoch": 2.2860484153900127,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 9773
    },
    {
      "epoch": 2.286282306163022,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 9774
    },
    {
      "epoch": 2.286516196936031,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 9775
    },
    {
      "epoch": 2.2867500877090396,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 9776
    },
    {
      "epoch": 2.286983978482049,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6559,
      "step": 9777
    },
    {
      "epoch": 2.2872178692550578,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7619,
      "step": 9778
    },
    {
      "epoch": 2.287451760028067,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 9779
    },
    {
      "epoch": 2.287685650801076,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 9780
    },
    {
      "epoch": 2.2879195415740847,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 9781
    },
    {
      "epoch": 2.288153432347094,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 9782
    },
    {
      "epoch": 2.288387323120103,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 9783
    },
    {
      "epoch": 2.288621213893112,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3219,
      "step": 9784
    },
    {
      "epoch": 2.288855104666121,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 9785
    },
    {
      "epoch": 2.28908899543913,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 9786
    },
    {
      "epoch": 2.289322886212139,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 9787
    },
    {
      "epoch": 2.289556776985148,
      "grad_norm": 12.6875,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 9788
    },
    {
      "epoch": 2.2897906677581568,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6291,
      "step": 9789
    },
    {
      "epoch": 2.290024558531166,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 9790
    },
    {
      "epoch": 2.290258449304175,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8734,
      "step": 9791
    },
    {
      "epoch": 2.2904923400771837,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.474,
      "step": 9792
    },
    {
      "epoch": 2.290726230850193,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 9793
    },
    {
      "epoch": 2.290960121623202,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8988,
      "step": 9794
    },
    {
      "epoch": 2.291194012396211,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 9795
    },
    {
      "epoch": 2.29142790316922,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7277,
      "step": 9796
    },
    {
      "epoch": 2.291661793942229,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 9797
    },
    {
      "epoch": 2.291895684715238,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 9798
    },
    {
      "epoch": 2.292129575488247,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 9799
    },
    {
      "epoch": 2.292363466261256,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 9800
    },
    {
      "epoch": 2.292363466261256,
      "eval_runtime": 4.6031,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 9800
    },
    {
      "epoch": 2.292597357034265,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 9801
    },
    {
      "epoch": 2.292831247807274,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 9802
    },
    {
      "epoch": 2.293065138580283,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 9803
    },
    {
      "epoch": 2.293299029353292,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7066,
      "step": 9804
    },
    {
      "epoch": 2.293532920126301,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 9805
    },
    {
      "epoch": 2.29376681089931,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 9806
    },
    {
      "epoch": 2.294000701672319,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 9807
    },
    {
      "epoch": 2.294234592445328,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 9808
    },
    {
      "epoch": 2.294468483218337,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 9809
    },
    {
      "epoch": 2.294702373991346,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 9810
    },
    {
      "epoch": 2.294936264764355,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 9811
    },
    {
      "epoch": 2.295170155537364,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 9812
    },
    {
      "epoch": 2.295404046310373,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 9813
    },
    {
      "epoch": 2.295637937083382,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 9814
    },
    {
      "epoch": 2.295871827856391,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 9815
    },
    {
      "epoch": 2.2961057186294003,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 9816
    },
    {
      "epoch": 2.296339609402409,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 9817
    },
    {
      "epoch": 2.296573500175418,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 9818
    },
    {
      "epoch": 2.2968073909484272,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 9819
    },
    {
      "epoch": 2.297041281721436,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.673,
      "step": 9820
    },
    {
      "epoch": 2.297275172494445,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 9821
    },
    {
      "epoch": 2.297509063267454,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 9822
    },
    {
      "epoch": 2.297742954040463,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 9823
    },
    {
      "epoch": 2.297976844813472,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6031,
      "step": 9824
    },
    {
      "epoch": 2.298210735586481,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 9825
    },
    {
      "epoch": 2.29844462635949,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8653,
      "step": 9826
    },
    {
      "epoch": 2.2986785171324993,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 9827
    },
    {
      "epoch": 2.298912407905508,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 9828
    },
    {
      "epoch": 2.299146298678517,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 9829
    },
    {
      "epoch": 2.2993801894515262,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 9830
    },
    {
      "epoch": 2.299614080224535,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 9831
    },
    {
      "epoch": 2.2998479709975443,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 9832
    },
    {
      "epoch": 2.300081861770553,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 9833
    },
    {
      "epoch": 2.300315752543562,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 9834
    },
    {
      "epoch": 2.3005496433165713,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 9835
    },
    {
      "epoch": 2.30078353408958,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 9836
    },
    {
      "epoch": 2.301017424862589,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 9837
    },
    {
      "epoch": 2.3012513156355983,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 9838
    },
    {
      "epoch": 2.301485206408607,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 9839
    },
    {
      "epoch": 2.301719097181616,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 9840
    },
    {
      "epoch": 2.301952987954625,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.5591,
      "step": 9841
    },
    {
      "epoch": 2.302186878727634,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 9842
    },
    {
      "epoch": 2.3024207695006433,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 9843
    },
    {
      "epoch": 2.302654660273652,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 9844
    },
    {
      "epoch": 2.302888551046661,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.6562,
      "step": 9845
    },
    {
      "epoch": 2.3031224418196703,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7221,
      "step": 9846
    },
    {
      "epoch": 2.303356332592679,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7444,
      "step": 9847
    },
    {
      "epoch": 2.3035902233656884,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 9848
    },
    {
      "epoch": 2.3038241141386973,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 9849
    },
    {
      "epoch": 2.304058004911706,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 9850
    },
    {
      "epoch": 2.3042918956847154,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 9851
    },
    {
      "epoch": 2.304525786457724,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 9852
    },
    {
      "epoch": 2.3047596772307335,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 9853
    },
    {
      "epoch": 2.3049935680037423,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5234,
      "step": 9854
    },
    {
      "epoch": 2.305227458776751,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 9855
    },
    {
      "epoch": 2.3054613495497605,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 9856
    },
    {
      "epoch": 2.3056952403227693,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 9857
    },
    {
      "epoch": 2.305929131095778,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 9858
    },
    {
      "epoch": 2.3061630218687874,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 9859
    },
    {
      "epoch": 2.3063969126417962,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 9860
    },
    {
      "epoch": 2.306630803414805,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3211,
      "step": 9861
    },
    {
      "epoch": 2.3068646941878144,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 9862
    },
    {
      "epoch": 2.307098584960823,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.594,
      "step": 9863
    },
    {
      "epoch": 2.3073324757338325,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 9864
    },
    {
      "epoch": 2.3075663665068413,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 9865
    },
    {
      "epoch": 2.30780025727985,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 9866
    },
    {
      "epoch": 2.3080341480528594,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 9867
    },
    {
      "epoch": 2.3082680388258683,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 9868
    },
    {
      "epoch": 2.3085019295988776,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 9869
    },
    {
      "epoch": 2.3087358203718864,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8306,
      "step": 9870
    },
    {
      "epoch": 2.3089697111448952,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.4779,
      "step": 9871
    },
    {
      "epoch": 2.3092036019179045,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 9872
    },
    {
      "epoch": 2.3094374926909134,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 9873
    },
    {
      "epoch": 2.309671383463922,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7851,
      "step": 9874
    },
    {
      "epoch": 2.3099052742369315,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 9875
    },
    {
      "epoch": 2.3101391650099403,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6865,
      "step": 9876
    },
    {
      "epoch": 2.310373055782949,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 9877
    },
    {
      "epoch": 2.3106069465559584,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7093,
      "step": 9878
    },
    {
      "epoch": 2.3108408373289673,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 9879
    },
    {
      "epoch": 2.3110747281019766,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2611,
      "step": 9880
    },
    {
      "epoch": 2.3113086188749854,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 9881
    },
    {
      "epoch": 2.3115425096479942,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5802,
      "step": 9882
    },
    {
      "epoch": 2.3117764004210035,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 9883
    },
    {
      "epoch": 2.3120102911940124,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 9884
    },
    {
      "epoch": 2.3122441819670216,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 9885
    },
    {
      "epoch": 2.3124780727400305,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 9886
    },
    {
      "epoch": 2.3127119635130393,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2489,
      "step": 9887
    },
    {
      "epoch": 2.3129458542860486,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.15,
      "step": 9888
    },
    {
      "epoch": 2.3131797450590574,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7345,
      "step": 9889
    },
    {
      "epoch": 2.3134136358320663,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 9890
    },
    {
      "epoch": 2.3136475266050756,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 9891
    },
    {
      "epoch": 2.3138814173780844,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 9892
    },
    {
      "epoch": 2.3141153081510932,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 9893
    },
    {
      "epoch": 2.3143491989241025,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 9894
    },
    {
      "epoch": 2.3145830896971114,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 9895
    },
    {
      "epoch": 2.3148169804701206,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 9896
    },
    {
      "epoch": 2.3150508712431295,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 9897
    },
    {
      "epoch": 2.3152847620161383,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 9898
    },
    {
      "epoch": 2.3155186527891476,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7261,
      "step": 9899
    },
    {
      "epoch": 2.3157525435621564,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 9900
    },
    {
      "epoch": 2.3157525435621564,
      "eval_runtime": 4.611,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 9900
    },
    {
      "epoch": 2.3159864343351657,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4926,
      "step": 9901
    },
    {
      "epoch": 2.3162203251081745,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5193,
      "step": 9902
    },
    {
      "epoch": 2.3164542158811834,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 9903
    },
    {
      "epoch": 2.3166881066541927,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 9904
    },
    {
      "epoch": 2.3169219974272015,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.363,
      "step": 9905
    },
    {
      "epoch": 2.3171558882002103,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 9906
    },
    {
      "epoch": 2.3173897789732196,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 9907
    },
    {
      "epoch": 2.3176236697462285,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 9908
    },
    {
      "epoch": 2.3178575605192373,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 9909
    },
    {
      "epoch": 2.3180914512922466,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 9910
    },
    {
      "epoch": 2.3183253420652554,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 9911
    },
    {
      "epoch": 2.3185592328382647,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 9912
    },
    {
      "epoch": 2.3187931236112735,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 9913
    },
    {
      "epoch": 2.3190270143842824,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.866,
      "step": 9914
    },
    {
      "epoch": 2.3192609051572917,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6459,
      "step": 9915
    },
    {
      "epoch": 2.3194947959303005,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4374,
      "step": 9916
    },
    {
      "epoch": 2.31972868670331,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 9917
    },
    {
      "epoch": 2.3199625774763186,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 9918
    },
    {
      "epoch": 2.3201964682493275,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 9919
    },
    {
      "epoch": 2.3204303590223367,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 9920
    },
    {
      "epoch": 2.3206642497953456,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 9921
    },
    {
      "epoch": 2.3208981405683544,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 9922
    },
    {
      "epoch": 2.3211320313413637,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6794,
      "step": 9923
    },
    {
      "epoch": 2.3213659221143725,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 9924
    },
    {
      "epoch": 2.3215998128873814,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.885,
      "step": 9925
    },
    {
      "epoch": 2.3218337036603907,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 9926
    },
    {
      "epoch": 2.3220675944333995,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6774,
      "step": 9927
    },
    {
      "epoch": 2.3223014852064088,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 9928
    },
    {
      "epoch": 2.3225353759794176,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 9929
    },
    {
      "epoch": 2.3227692667524265,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0112,
      "step": 9930
    },
    {
      "epoch": 2.3230031575254357,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 9931
    },
    {
      "epoch": 2.3232370482984446,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 9932
    },
    {
      "epoch": 2.323470939071454,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 9933
    },
    {
      "epoch": 2.3237048298444627,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 9934
    },
    {
      "epoch": 2.3239387206174715,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 9935
    },
    {
      "epoch": 2.324172611390481,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 9936
    },
    {
      "epoch": 2.3244065021634897,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 9937
    },
    {
      "epoch": 2.3246403929364985,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 9938
    },
    {
      "epoch": 2.3248742837095078,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 9939
    },
    {
      "epoch": 2.3251081744825166,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.558,
      "step": 9940
    },
    {
      "epoch": 2.3253420652555254,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 9941
    },
    {
      "epoch": 2.3255759560285347,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 1.6346,
      "step": 9942
    },
    {
      "epoch": 2.3258098468015436,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8411,
      "step": 9943
    },
    {
      "epoch": 2.326043737574553,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 9944
    },
    {
      "epoch": 2.3262776283475617,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 9945
    },
    {
      "epoch": 2.3265115191205705,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 9946
    },
    {
      "epoch": 2.32674540989358,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 9947
    },
    {
      "epoch": 2.3269793006665886,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 9948
    },
    {
      "epoch": 2.327213191439598,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 9949
    },
    {
      "epoch": 2.3274470822126068,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8053,
      "step": 9950
    },
    {
      "epoch": 2.3276809729856156,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 9951
    },
    {
      "epoch": 2.327914863758625,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 9952
    },
    {
      "epoch": 2.3281487545316337,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 9953
    },
    {
      "epoch": 2.3283826453046426,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 9954
    },
    {
      "epoch": 2.328616536077652,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 9955
    },
    {
      "epoch": 2.3288504268506607,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 9956
    },
    {
      "epoch": 2.3290843176236695,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 9957
    },
    {
      "epoch": 2.329318208396679,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 9958
    },
    {
      "epoch": 2.3295520991696876,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 9959
    },
    {
      "epoch": 2.329785989942697,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 9960
    },
    {
      "epoch": 2.3300198807157058,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 9961
    },
    {
      "epoch": 2.3302537714887146,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 9962
    },
    {
      "epoch": 2.330487662261724,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 9963
    },
    {
      "epoch": 2.3307215530347327,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 9964
    },
    {
      "epoch": 2.330955443807742,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5897,
      "step": 9965
    },
    {
      "epoch": 2.331189334580751,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 9966
    },
    {
      "epoch": 2.3314232253537597,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 9967
    },
    {
      "epoch": 2.331657116126769,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8205,
      "step": 9968
    },
    {
      "epoch": 2.331891006899778,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.814,
      "step": 9969
    },
    {
      "epoch": 2.3321248976727866,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 9970
    },
    {
      "epoch": 2.332358788445796,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 9971
    },
    {
      "epoch": 2.3325926792188048,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 9972
    },
    {
      "epoch": 2.3328265699918136,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 9973
    },
    {
      "epoch": 2.333060460764823,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4383,
      "step": 9974
    },
    {
      "epoch": 2.3332943515378317,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.4915,
      "step": 9975
    },
    {
      "epoch": 2.333528242310841,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 9976
    },
    {
      "epoch": 2.33376213308385,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 9977
    },
    {
      "epoch": 2.3339960238568587,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 9978
    },
    {
      "epoch": 2.334229914629868,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 9979
    },
    {
      "epoch": 2.334463805402877,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 9980
    },
    {
      "epoch": 2.334697696175886,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 9981
    },
    {
      "epoch": 2.334931586948895,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 9982
    },
    {
      "epoch": 2.3351654777219037,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 9983
    },
    {
      "epoch": 2.335399368494913,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 9984
    },
    {
      "epoch": 2.335633259267922,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 9985
    },
    {
      "epoch": 2.3358671500409307,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 9986
    },
    {
      "epoch": 2.33610104081394,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7397,
      "step": 9987
    },
    {
      "epoch": 2.336334931586949,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 9988
    },
    {
      "epoch": 2.3365688223599577,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7449,
      "step": 9989
    },
    {
      "epoch": 2.336802713132967,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 9990
    },
    {
      "epoch": 2.337036603905976,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 9991
    },
    {
      "epoch": 2.337270494678985,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6256,
      "step": 9992
    },
    {
      "epoch": 2.337504385451994,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 9993
    },
    {
      "epoch": 2.3377382762250027,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6493,
      "step": 9994
    },
    {
      "epoch": 2.337972166998012,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 9995
    },
    {
      "epoch": 2.338206057771021,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 9996
    },
    {
      "epoch": 2.33843994854403,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 9997
    },
    {
      "epoch": 2.338673839317039,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 9998
    },
    {
      "epoch": 2.338907730090048,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8385,
      "step": 9999
    },
    {
      "epoch": 2.339141620863057,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 10000
    },
    {
      "epoch": 2.339141620863057,
      "eval_runtime": 4.64,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 10000
    },
    {
      "epoch": 2.339375511636066,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 10001
    },
    {
      "epoch": 2.339609402409075,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7723,
      "step": 10002
    },
    {
      "epoch": 2.339843293182084,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 10003
    },
    {
      "epoch": 2.340077183955093,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 10004
    },
    {
      "epoch": 2.3403110747281017,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 10005
    },
    {
      "epoch": 2.340544965501111,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 10006
    },
    {
      "epoch": 2.34077885627412,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 10007
    },
    {
      "epoch": 2.341012747047129,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6639,
      "step": 10008
    },
    {
      "epoch": 2.341246637820138,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 10009
    },
    {
      "epoch": 2.341480528593147,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 10010
    },
    {
      "epoch": 2.341714419366156,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 10011
    },
    {
      "epoch": 2.341948310139165,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 10012
    },
    {
      "epoch": 2.342182200912174,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 10013
    },
    {
      "epoch": 2.342416091685183,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 10014
    },
    {
      "epoch": 2.342649982458192,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 10015
    },
    {
      "epoch": 2.342883873231201,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 10016
    },
    {
      "epoch": 2.34311776400421,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4659,
      "step": 10017
    },
    {
      "epoch": 2.343351654777219,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 10018
    },
    {
      "epoch": 2.343585545550228,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 10019
    },
    {
      "epoch": 2.343819436323237,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7128,
      "step": 10020
    },
    {
      "epoch": 2.344053327096246,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 10021
    },
    {
      "epoch": 2.344287217869255,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 10022
    },
    {
      "epoch": 2.344521108642264,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5083,
      "step": 10023
    },
    {
      "epoch": 2.344754999415273,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 10024
    },
    {
      "epoch": 2.344988890188282,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 10025
    },
    {
      "epoch": 2.345222780961291,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 10026
    },
    {
      "epoch": 2.3454566717343,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 10027
    },
    {
      "epoch": 2.345690562507309,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 10028
    },
    {
      "epoch": 2.3459244532803183,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 10029
    },
    {
      "epoch": 2.346158344053327,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 10030
    },
    {
      "epoch": 2.346392234826336,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 10031
    },
    {
      "epoch": 2.3466261255993452,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6474,
      "step": 10032
    },
    {
      "epoch": 2.346860016372354,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 10033
    },
    {
      "epoch": 2.347093907145363,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 10034
    },
    {
      "epoch": 2.347327797918372,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 10035
    },
    {
      "epoch": 2.347561688691381,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 10036
    },
    {
      "epoch": 2.34779557946439,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 10037
    },
    {
      "epoch": 2.348029470237399,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 10038
    },
    {
      "epoch": 2.348263361010408,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5471,
      "step": 10039
    },
    {
      "epoch": 2.3484972517834173,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 10040
    },
    {
      "epoch": 2.348731142556426,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 10041
    },
    {
      "epoch": 2.348965033329435,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 10042
    },
    {
      "epoch": 2.3491989241024442,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 10043
    },
    {
      "epoch": 2.349432814875453,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 10044
    },
    {
      "epoch": 2.3496667056484624,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 10045
    },
    {
      "epoch": 2.349900596421471,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.3779,
      "step": 10046
    },
    {
      "epoch": 2.35013448719448,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 10047
    },
    {
      "epoch": 2.3503683779674893,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.607,
      "step": 10048
    },
    {
      "epoch": 2.350602268740498,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 10049
    },
    {
      "epoch": 2.3508361595135074,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 10050
    },
    {
      "epoch": 2.3510700502865163,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 10051
    },
    {
      "epoch": 2.351303941059525,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6951,
      "step": 10052
    },
    {
      "epoch": 2.3515378318325344,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 10053
    },
    {
      "epoch": 2.3517717226055432,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 10054
    },
    {
      "epoch": 2.352005613378552,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 10055
    },
    {
      "epoch": 2.3522395041515614,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 10056
    },
    {
      "epoch": 2.35247339492457,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 10057
    },
    {
      "epoch": 2.352707285697579,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 10058
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 10059
    },
    {
      "epoch": 2.353175067243597,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 10060
    },
    {
      "epoch": 2.3534089580166064,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 10061
    },
    {
      "epoch": 2.3536428487896153,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 10062
    },
    {
      "epoch": 2.353876739562624,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 10063
    },
    {
      "epoch": 2.3541106303356334,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 10064
    },
    {
      "epoch": 2.3543445211086422,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 10065
    },
    {
      "epoch": 2.3545784118816515,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 10066
    },
    {
      "epoch": 2.3548123026546603,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 10067
    },
    {
      "epoch": 2.355046193427669,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 10068
    },
    {
      "epoch": 2.3552800842006785,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 10069
    },
    {
      "epoch": 2.3555139749736873,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 10070
    },
    {
      "epoch": 2.355747865746696,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 10071
    },
    {
      "epoch": 2.3559817565197054,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 10072
    },
    {
      "epoch": 2.3562156472927143,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 10073
    },
    {
      "epoch": 2.356449538065723,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 10074
    },
    {
      "epoch": 2.3566834288387324,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 10075
    },
    {
      "epoch": 2.3569173196117412,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5574,
      "step": 10076
    },
    {
      "epoch": 2.3571512103847505,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 10077
    },
    {
      "epoch": 2.3573851011577593,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 10078
    },
    {
      "epoch": 2.357618991930768,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 10079
    },
    {
      "epoch": 2.3578528827037775,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 10080
    },
    {
      "epoch": 2.3580867734767863,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 10081
    },
    {
      "epoch": 2.3583206642497956,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 10082
    },
    {
      "epoch": 2.3585545550228044,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 10083
    },
    {
      "epoch": 2.3587884457958133,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 10084
    },
    {
      "epoch": 2.3590223365688225,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5276,
      "step": 10085
    },
    {
      "epoch": 2.3592562273418314,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 10086
    },
    {
      "epoch": 2.35949011811484,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7936,
      "step": 10087
    },
    {
      "epoch": 2.3597240088878495,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 10088
    },
    {
      "epoch": 2.3599578996608583,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 10089
    },
    {
      "epoch": 2.360191790433867,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 10090
    },
    {
      "epoch": 2.3604256812068765,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 10091
    },
    {
      "epoch": 2.3606595719798853,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 10092
    },
    {
      "epoch": 2.3608934627528946,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 10093
    },
    {
      "epoch": 2.3611273535259034,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 10094
    },
    {
      "epoch": 2.3613612442989123,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 10095
    },
    {
      "epoch": 2.3615951350719215,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 10096
    },
    {
      "epoch": 2.3618290258449304,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 10097
    },
    {
      "epoch": 2.3620629166179397,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.6109,
      "step": 10098
    },
    {
      "epoch": 2.3622968073909485,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 10099
    },
    {
      "epoch": 2.3625306981639573,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 10100
    },
    {
      "epoch": 2.3625306981639573,
      "eval_runtime": 4.6297,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 10100
    },
    {
      "epoch": 2.3627645889369666,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 10101
    },
    {
      "epoch": 2.3629984797099755,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.0491,
      "step": 10102
    },
    {
      "epoch": 2.3632323704829843,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 10103
    },
    {
      "epoch": 2.3634662612559936,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 10104
    },
    {
      "epoch": 2.3637001520290024,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 10105
    },
    {
      "epoch": 2.3639340428020112,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3365,
      "step": 10106
    },
    {
      "epoch": 2.3641679335750205,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 10107
    },
    {
      "epoch": 2.3644018243480294,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 10108
    },
    {
      "epoch": 2.3646357151210387,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 10109
    },
    {
      "epoch": 2.3648696058940475,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 10110
    },
    {
      "epoch": 2.3651034966670563,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 10111
    },
    {
      "epoch": 2.3653373874400656,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6981,
      "step": 10112
    },
    {
      "epoch": 2.3655712782130744,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 10113
    },
    {
      "epoch": 2.3658051689860837,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7515,
      "step": 10114
    },
    {
      "epoch": 2.3660390597590926,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 10115
    },
    {
      "epoch": 2.3662729505321014,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 10116
    },
    {
      "epoch": 2.3665068413051107,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 10117
    },
    {
      "epoch": 2.3667407320781195,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 10118
    },
    {
      "epoch": 2.3669746228511284,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 10119
    },
    {
      "epoch": 2.3672085136241376,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8178,
      "step": 10120
    },
    {
      "epoch": 2.3674424043971465,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0601,
      "step": 10121
    },
    {
      "epoch": 2.3676762951701553,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 10122
    },
    {
      "epoch": 2.3679101859431646,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 10123
    },
    {
      "epoch": 2.3681440767161734,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 10124
    },
    {
      "epoch": 2.3683779674891827,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 10125
    },
    {
      "epoch": 2.3686118582621916,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 10126
    },
    {
      "epoch": 2.3688457490352004,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 10127
    },
    {
      "epoch": 2.3690796398082097,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 10128
    },
    {
      "epoch": 2.3693135305812185,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7061,
      "step": 10129
    },
    {
      "epoch": 2.369547421354228,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 10130
    },
    {
      "epoch": 2.3697813121272366,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.726,
      "step": 10131
    },
    {
      "epoch": 2.3700152029002455,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6095,
      "step": 10132
    },
    {
      "epoch": 2.3702490936732548,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 10133
    },
    {
      "epoch": 2.3704829844462636,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 10134
    },
    {
      "epoch": 2.3707168752192724,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 10135
    },
    {
      "epoch": 2.3709507659922817,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.819,
      "step": 10136
    },
    {
      "epoch": 2.3711846567652906,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 10137
    },
    {
      "epoch": 2.3714185475382994,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 10138
    },
    {
      "epoch": 2.3716524383113087,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6593,
      "step": 10139
    },
    {
      "epoch": 2.3718863290843175,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 10140
    },
    {
      "epoch": 2.372120219857327,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 10141
    },
    {
      "epoch": 2.3723541106303356,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 10142
    },
    {
      "epoch": 2.3725880014033445,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 10143
    },
    {
      "epoch": 2.3728218921763538,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7442,
      "step": 10144
    },
    {
      "epoch": 2.3730557829493626,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.369,
      "step": 10145
    },
    {
      "epoch": 2.373289673722372,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 10146
    },
    {
      "epoch": 2.3735235644953807,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.667,
      "step": 10147
    },
    {
      "epoch": 2.3737574552683895,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3292,
      "step": 10148
    },
    {
      "epoch": 2.373991346041399,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 10149
    },
    {
      "epoch": 2.3742252368144077,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 10150
    },
    {
      "epoch": 2.3744591275874165,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 10151
    },
    {
      "epoch": 2.374693018360426,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 10152
    },
    {
      "epoch": 2.3749269091334346,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 10153
    },
    {
      "epoch": 2.3751607999064435,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 10154
    },
    {
      "epoch": 2.3753946906794527,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 10155
    },
    {
      "epoch": 2.3756285814524616,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 10156
    },
    {
      "epoch": 2.375862472225471,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6523,
      "step": 10157
    },
    {
      "epoch": 2.3760963629984797,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6011,
      "step": 10158
    },
    {
      "epoch": 2.3763302537714885,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 10159
    },
    {
      "epoch": 2.376564144544498,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 10160
    },
    {
      "epoch": 2.3767980353175067,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 10161
    },
    {
      "epoch": 2.377031926090516,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 10162
    },
    {
      "epoch": 2.377265816863525,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 10163
    },
    {
      "epoch": 2.3774997076365336,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 10164
    },
    {
      "epoch": 2.377733598409543,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 10165
    },
    {
      "epoch": 2.3779674891825517,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 10166
    },
    {
      "epoch": 2.3782013799555606,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 10167
    },
    {
      "epoch": 2.37843527072857,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 10168
    },
    {
      "epoch": 2.3786691615015787,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 10169
    },
    {
      "epoch": 2.3789030522745875,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 10170
    },
    {
      "epoch": 2.379136943047597,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 10171
    },
    {
      "epoch": 2.3793708338206057,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 10172
    },
    {
      "epoch": 2.379604724593615,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7447,
      "step": 10173
    },
    {
      "epoch": 2.3798386153666238,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 10174
    },
    {
      "epoch": 2.3800725061396326,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 10175
    },
    {
      "epoch": 2.380306396912642,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 10176
    },
    {
      "epoch": 2.3805402876856507,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.5001,
      "step": 10177
    },
    {
      "epoch": 2.38077417845866,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 10178
    },
    {
      "epoch": 2.381008069231669,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 10179
    },
    {
      "epoch": 2.3812419600046777,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 10180
    },
    {
      "epoch": 2.381475850777687,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 10181
    },
    {
      "epoch": 2.381709741550696,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 10182
    },
    {
      "epoch": 2.3819436323237047,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 10183
    },
    {
      "epoch": 2.382177523096714,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 10184
    },
    {
      "epoch": 2.3824114138697228,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5406,
      "step": 10185
    },
    {
      "epoch": 2.3826453046427316,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 10186
    },
    {
      "epoch": 2.382879195415741,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7598,
      "step": 10187
    },
    {
      "epoch": 2.3831130861887497,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 10188
    },
    {
      "epoch": 2.383346976961759,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 10189
    },
    {
      "epoch": 2.383580867734768,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 10190
    },
    {
      "epoch": 2.3838147585077767,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 10191
    },
    {
      "epoch": 2.384048649280786,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 10192
    },
    {
      "epoch": 2.384282540053795,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 10193
    },
    {
      "epoch": 2.384516430826804,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 10194
    },
    {
      "epoch": 2.384750321599813,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 10195
    },
    {
      "epoch": 2.3849842123728218,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 10196
    },
    {
      "epoch": 2.385218103145831,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 10197
    },
    {
      "epoch": 2.38545199391884,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 10198
    },
    {
      "epoch": 2.3856858846918487,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8501,
      "step": 10199
    },
    {
      "epoch": 2.385919775464858,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3189,
      "step": 10200
    },
    {
      "epoch": 2.385919775464858,
      "eval_runtime": 4.5809,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 10200
    },
    {
      "epoch": 2.386153666237867,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 10201
    },
    {
      "epoch": 2.3863875570108757,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 10202
    },
    {
      "epoch": 2.386621447783885,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 10203
    },
    {
      "epoch": 2.386855338556894,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 10204
    },
    {
      "epoch": 2.387089229329903,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 10205
    },
    {
      "epoch": 2.387323120102912,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 10206
    },
    {
      "epoch": 2.3875570108759208,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.4953,
      "step": 10207
    },
    {
      "epoch": 2.38779090164893,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.7479,
      "step": 10208
    },
    {
      "epoch": 2.388024792421939,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 10209
    },
    {
      "epoch": 2.388258683194948,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 10210
    },
    {
      "epoch": 2.388492573967957,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 10211
    },
    {
      "epoch": 2.388726464740966,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 10212
    },
    {
      "epoch": 2.388960355513975,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 10213
    },
    {
      "epoch": 2.389194246286984,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 10214
    },
    {
      "epoch": 2.389428137059993,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 10215
    },
    {
      "epoch": 2.389662027833002,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 10216
    },
    {
      "epoch": 2.389895918606011,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 10217
    },
    {
      "epoch": 2.3901298093790198,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 10218
    },
    {
      "epoch": 2.390363700152029,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7984,
      "step": 10219
    },
    {
      "epoch": 2.390597590925038,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 10220
    },
    {
      "epoch": 2.390831481698047,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6479,
      "step": 10221
    },
    {
      "epoch": 2.391065372471056,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 10222
    },
    {
      "epoch": 2.391299263244065,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 10223
    },
    {
      "epoch": 2.391533154017074,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 10224
    },
    {
      "epoch": 2.391767044790083,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 10225
    },
    {
      "epoch": 2.3920009355630922,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 10226
    },
    {
      "epoch": 2.392234826336101,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 10227
    },
    {
      "epoch": 2.39246871710911,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 10228
    },
    {
      "epoch": 2.392702607882119,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 10229
    },
    {
      "epoch": 2.392936498655128,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 10230
    },
    {
      "epoch": 2.393170389428137,
      "grad_norm": 10.5625,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 10231
    },
    {
      "epoch": 2.393404280201146,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5256,
      "step": 10232
    },
    {
      "epoch": 2.393638170974155,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6923,
      "step": 10233
    },
    {
      "epoch": 2.393872061747164,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 10234
    },
    {
      "epoch": 2.394105952520173,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 10235
    },
    {
      "epoch": 2.394339843293182,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 10236
    },
    {
      "epoch": 2.3945737340661912,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9504,
      "step": 10237
    },
    {
      "epoch": 2.3948076248392,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.4005,
      "step": 10238
    },
    {
      "epoch": 2.395041515612209,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6096,
      "step": 10239
    },
    {
      "epoch": 2.395275406385218,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 10240
    },
    {
      "epoch": 2.395509297158227,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7482,
      "step": 10241
    },
    {
      "epoch": 2.3957431879312363,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 10242
    },
    {
      "epoch": 2.395977078704245,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 10243
    },
    {
      "epoch": 2.396210969477254,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 10244
    },
    {
      "epoch": 2.3964448602502633,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 10245
    },
    {
      "epoch": 2.396678751023272,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 10246
    },
    {
      "epoch": 2.3969126417962814,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7256,
      "step": 10247
    },
    {
      "epoch": 2.3971465325692902,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 10248
    },
    {
      "epoch": 2.397380423342299,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 10249
    },
    {
      "epoch": 2.3976143141153083,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 10250
    },
    {
      "epoch": 2.397848204888317,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 10251
    },
    {
      "epoch": 2.398082095661326,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 10252
    },
    {
      "epoch": 2.3983159864343353,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 10253
    },
    {
      "epoch": 2.398549877207344,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 10254
    },
    {
      "epoch": 2.398783767980353,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6354,
      "step": 10255
    },
    {
      "epoch": 2.3990176587533623,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 10256
    },
    {
      "epoch": 2.399251549526371,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7777,
      "step": 10257
    },
    {
      "epoch": 2.3994854402993804,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 10258
    },
    {
      "epoch": 2.399719331072389,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 10259
    },
    {
      "epoch": 2.399953221845398,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 10260
    },
    {
      "epoch": 2.4001871126184073,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 10261
    },
    {
      "epoch": 2.400421003391416,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 10262
    },
    {
      "epoch": 2.4006548941644255,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 10263
    },
    {
      "epoch": 2.4008887849374343,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 10264
    },
    {
      "epoch": 2.401122675710443,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 10265
    },
    {
      "epoch": 2.4013565664834524,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 10266
    },
    {
      "epoch": 2.4015904572564613,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7395,
      "step": 10267
    },
    {
      "epoch": 2.40182434802947,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 10268
    },
    {
      "epoch": 2.4020582388024794,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 10269
    },
    {
      "epoch": 2.402292129575488,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 10270
    },
    {
      "epoch": 2.402526020348497,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 10271
    },
    {
      "epoch": 2.4027599111215063,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 10272
    },
    {
      "epoch": 2.402993801894515,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 10273
    },
    {
      "epoch": 2.4032276926675245,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 10274
    },
    {
      "epoch": 2.4034615834405333,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 10275
    },
    {
      "epoch": 2.403695474213542,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6255,
      "step": 10276
    },
    {
      "epoch": 2.4039293649865514,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 10277
    },
    {
      "epoch": 2.4041632557595602,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 10278
    },
    {
      "epoch": 2.4043971465325695,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5018,
      "step": 10279
    },
    {
      "epoch": 2.4046310373055784,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9503,
      "step": 10280
    },
    {
      "epoch": 2.404864928078587,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7789,
      "step": 10281
    },
    {
      "epoch": 2.4050988188515965,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 10282
    },
    {
      "epoch": 2.4053327096246053,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 10283
    },
    {
      "epoch": 2.405566600397614,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 10284
    },
    {
      "epoch": 2.4058004911706234,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 10285
    },
    {
      "epoch": 2.4060343819436323,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 10286
    },
    {
      "epoch": 2.406268272716641,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 10287
    },
    {
      "epoch": 2.4065021634896504,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 10288
    },
    {
      "epoch": 2.4067360542626592,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6519,
      "step": 10289
    },
    {
      "epoch": 2.4069699450356685,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 10290
    },
    {
      "epoch": 2.4072038358086774,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 10291
    },
    {
      "epoch": 2.407437726581686,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.712,
      "step": 10292
    },
    {
      "epoch": 2.4076716173546955,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 10293
    },
    {
      "epoch": 2.4079055081277043,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5408,
      "step": 10294
    },
    {
      "epoch": 2.4081393989007136,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 10295
    },
    {
      "epoch": 2.4083732896737224,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 10296
    },
    {
      "epoch": 2.4086071804467313,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 10297
    },
    {
      "epoch": 2.4088410712197406,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6793,
      "step": 10298
    },
    {
      "epoch": 2.4090749619927494,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 10299
    },
    {
      "epoch": 2.4093088527657582,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 10300
    },
    {
      "epoch": 2.4093088527657582,
      "eval_runtime": 4.7262,
      "eval_samples_per_second": 0.212,
      "eval_steps_per_second": 0.212,
      "step": 10300
    },
    {
      "epoch": 2.4095427435387675,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 10301
    },
    {
      "epoch": 2.4097766343117764,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 10302
    },
    {
      "epoch": 2.410010525084785,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 10303
    },
    {
      "epoch": 2.4102444158577945,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 10304
    },
    {
      "epoch": 2.4104783066308033,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 10305
    },
    {
      "epoch": 2.4107121974038126,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5784,
      "step": 10306
    },
    {
      "epoch": 2.4109460881768214,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6511,
      "step": 10307
    },
    {
      "epoch": 2.4111799789498303,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 10308
    },
    {
      "epoch": 2.4114138697228396,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 10309
    },
    {
      "epoch": 2.4116477604958484,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 10310
    },
    {
      "epoch": 2.4118816512688577,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 10311
    },
    {
      "epoch": 2.4121155420418665,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 10312
    },
    {
      "epoch": 2.4123494328148753,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 10313
    },
    {
      "epoch": 2.4125833235878846,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 10314
    },
    {
      "epoch": 2.4128172143608935,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 10315
    },
    {
      "epoch": 2.4130511051339023,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 10316
    },
    {
      "epoch": 2.4132849959069116,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 10317
    },
    {
      "epoch": 2.4135188866799204,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 10318
    },
    {
      "epoch": 2.4137527774529293,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 10319
    },
    {
      "epoch": 2.4139866682259385,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 10320
    },
    {
      "epoch": 2.4142205589989474,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9221,
      "step": 10321
    },
    {
      "epoch": 2.4144544497719567,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6947,
      "step": 10322
    },
    {
      "epoch": 2.4146883405449655,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 10323
    },
    {
      "epoch": 2.4149222313179743,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 10324
    },
    {
      "epoch": 2.4151561220909836,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6462,
      "step": 10325
    },
    {
      "epoch": 2.4153900128639925,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 10326
    },
    {
      "epoch": 2.4156239036370017,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 10327
    },
    {
      "epoch": 2.4158577944100106,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 10328
    },
    {
      "epoch": 2.4160916851830194,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.635,
      "step": 10329
    },
    {
      "epoch": 2.4163255759560287,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 10330
    },
    {
      "epoch": 2.4165594667290375,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 10331
    },
    {
      "epoch": 2.4167933575020464,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8719,
      "step": 10332
    },
    {
      "epoch": 2.4170272482750557,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8073,
      "step": 10333
    },
    {
      "epoch": 2.4172611390480645,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 10334
    },
    {
      "epoch": 2.4174950298210733,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 10335
    },
    {
      "epoch": 2.4177289205940826,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5545,
      "step": 10336
    },
    {
      "epoch": 2.4179628113670915,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 10337
    },
    {
      "epoch": 2.4181967021401007,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 10338
    },
    {
      "epoch": 2.4184305929131096,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 10339
    },
    {
      "epoch": 2.4186644836861184,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 10340
    },
    {
      "epoch": 2.4188983744591277,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7906,
      "step": 10341
    },
    {
      "epoch": 2.4191322652321365,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 10342
    },
    {
      "epoch": 2.419366156005146,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 10343
    },
    {
      "epoch": 2.4196000467781547,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 10344
    },
    {
      "epoch": 2.4198339375511635,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 10345
    },
    {
      "epoch": 2.4200678283241728,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 10346
    },
    {
      "epoch": 2.4203017190971816,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 10347
    },
    {
      "epoch": 2.4205356098701905,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6606,
      "step": 10348
    },
    {
      "epoch": 2.4207695006431997,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 10349
    },
    {
      "epoch": 2.4210033914162086,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7644,
      "step": 10350
    },
    {
      "epoch": 2.4212372821892174,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 10351
    },
    {
      "epoch": 2.4214711729622267,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 10352
    },
    {
      "epoch": 2.4217050637352355,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 10353
    },
    {
      "epoch": 2.421938954508245,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 10354
    },
    {
      "epoch": 2.4221728452812537,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 10355
    },
    {
      "epoch": 2.4224067360542625,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 10356
    },
    {
      "epoch": 2.4226406268272718,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 10357
    },
    {
      "epoch": 2.4228745176002806,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 10358
    },
    {
      "epoch": 2.42310840837329,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 10359
    },
    {
      "epoch": 2.4233422991462987,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 10360
    },
    {
      "epoch": 2.4235761899193076,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 10361
    },
    {
      "epoch": 2.423810080692317,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 10362
    },
    {
      "epoch": 2.4240439714653257,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5865,
      "step": 10363
    },
    {
      "epoch": 2.4242778622383345,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 10364
    },
    {
      "epoch": 2.424511753011344,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 10365
    },
    {
      "epoch": 2.4247456437843526,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 10366
    },
    {
      "epoch": 2.4249795345573615,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1518,
      "step": 10367
    },
    {
      "epoch": 2.4252134253303708,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 10368
    },
    {
      "epoch": 2.4254473161033796,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5937,
      "step": 10369
    },
    {
      "epoch": 2.425681206876389,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 10370
    },
    {
      "epoch": 2.4259150976493977,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 10371
    },
    {
      "epoch": 2.4261489884224066,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 10372
    },
    {
      "epoch": 2.426382879195416,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.736,
      "step": 10373
    },
    {
      "epoch": 2.4266167699684247,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 10374
    },
    {
      "epoch": 2.426850660741434,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 10375
    },
    {
      "epoch": 2.427084551514443,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 10376
    },
    {
      "epoch": 2.4273184422874516,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 10377
    },
    {
      "epoch": 2.427552333060461,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 10378
    },
    {
      "epoch": 2.4277862238334698,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 10379
    },
    {
      "epoch": 2.4280201146064786,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 10380
    },
    {
      "epoch": 2.428254005379488,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 10381
    },
    {
      "epoch": 2.4284878961524967,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 10382
    },
    {
      "epoch": 2.4287217869255056,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7127,
      "step": 10383
    },
    {
      "epoch": 2.428955677698515,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 10384
    },
    {
      "epoch": 2.4291895684715237,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 10385
    },
    {
      "epoch": 2.429423459244533,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7988,
      "step": 10386
    },
    {
      "epoch": 2.429657350017542,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 10387
    },
    {
      "epoch": 2.4298912407905506,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 10388
    },
    {
      "epoch": 2.43012513156356,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 10389
    },
    {
      "epoch": 2.4303590223365688,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 10390
    },
    {
      "epoch": 2.430592913109578,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 10391
    },
    {
      "epoch": 2.430826803882587,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 10392
    },
    {
      "epoch": 2.4310606946555957,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 10393
    },
    {
      "epoch": 2.431294585428605,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7933,
      "step": 10394
    },
    {
      "epoch": 2.431528476201614,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 10395
    },
    {
      "epoch": 2.4317623669746227,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5561,
      "step": 10396
    },
    {
      "epoch": 2.431996257747632,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.078,
      "step": 10397
    },
    {
      "epoch": 2.432230148520641,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7741,
      "step": 10398
    },
    {
      "epoch": 2.4324640392936496,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 10399
    },
    {
      "epoch": 2.432697930066659,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 10400
    },
    {
      "epoch": 2.432697930066659,
      "eval_runtime": 4.6392,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 10400
    },
    {
      "epoch": 2.4329318208396677,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 10401
    },
    {
      "epoch": 2.433165711612677,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 10402
    },
    {
      "epoch": 2.433399602385686,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 10403
    },
    {
      "epoch": 2.4336334931586947,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 10404
    },
    {
      "epoch": 2.433867383931704,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 10405
    },
    {
      "epoch": 2.434101274704713,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 10406
    },
    {
      "epoch": 2.434335165477722,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 10407
    },
    {
      "epoch": 2.434569056250731,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1877,
      "step": 10408
    },
    {
      "epoch": 2.43480294702374,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 10409
    },
    {
      "epoch": 2.435036837796749,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 10410
    },
    {
      "epoch": 2.435270728569758,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 10411
    },
    {
      "epoch": 2.4355046193427667,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 10412
    },
    {
      "epoch": 2.435738510115776,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7135,
      "step": 10413
    },
    {
      "epoch": 2.435972400888785,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 10414
    },
    {
      "epoch": 2.4362062916617937,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7448,
      "step": 10415
    },
    {
      "epoch": 2.436440182434803,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 10416
    },
    {
      "epoch": 2.436674073207812,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 10417
    },
    {
      "epoch": 2.436907963980821,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2425,
      "step": 10418
    },
    {
      "epoch": 2.43714185475383,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 10419
    },
    {
      "epoch": 2.4373757455268388,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 10420
    },
    {
      "epoch": 2.437609636299848,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 10421
    },
    {
      "epoch": 2.437843527072857,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 10422
    },
    {
      "epoch": 2.438077417845866,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 10423
    },
    {
      "epoch": 2.438311308618875,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 10424
    },
    {
      "epoch": 2.438545199391884,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 10425
    },
    {
      "epoch": 2.438779090164893,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 10426
    },
    {
      "epoch": 2.439012980937902,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 10427
    },
    {
      "epoch": 2.439246871710911,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 10428
    },
    {
      "epoch": 2.43948076248392,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5242,
      "step": 10429
    },
    {
      "epoch": 2.439714653256929,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 10430
    },
    {
      "epoch": 2.4399485440299378,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 10431
    },
    {
      "epoch": 2.440182434802947,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 10432
    },
    {
      "epoch": 2.440416325575956,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 10433
    },
    {
      "epoch": 2.440650216348965,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 10434
    },
    {
      "epoch": 2.440884107121974,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 10435
    },
    {
      "epoch": 2.441117997894983,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 10436
    },
    {
      "epoch": 2.441351888667992,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.3141,
      "step": 10437
    },
    {
      "epoch": 2.441585779441001,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 10438
    },
    {
      "epoch": 2.4418196702140103,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 10439
    },
    {
      "epoch": 2.442053560987019,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 10440
    },
    {
      "epoch": 2.442287451760028,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4102,
      "step": 10441
    },
    {
      "epoch": 2.442521342533037,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6759,
      "step": 10442
    },
    {
      "epoch": 2.442755233306046,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 10443
    },
    {
      "epoch": 2.4429891240790553,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 10444
    },
    {
      "epoch": 2.443223014852064,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 10445
    },
    {
      "epoch": 2.443456905625073,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 10446
    },
    {
      "epoch": 2.4436907963980823,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.593,
      "step": 10447
    },
    {
      "epoch": 2.443924687171091,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 10448
    },
    {
      "epoch": 2.4441585779441,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 10449
    },
    {
      "epoch": 2.4443924687171092,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 10450
    },
    {
      "epoch": 2.444626359490118,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 10451
    },
    {
      "epoch": 2.444860250263127,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 10452
    },
    {
      "epoch": 2.445094141036136,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 10453
    },
    {
      "epoch": 2.445328031809145,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3118,
      "step": 10454
    },
    {
      "epoch": 2.4455619225821543,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8415,
      "step": 10455
    },
    {
      "epoch": 2.445795813355163,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 10456
    },
    {
      "epoch": 2.446029704128172,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 10457
    },
    {
      "epoch": 2.4462635949011813,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 10458
    },
    {
      "epoch": 2.44649748567419,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 10459
    },
    {
      "epoch": 2.4467313764471994,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 10460
    },
    {
      "epoch": 2.4469652672202082,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 10461
    },
    {
      "epoch": 2.447199157993217,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 10462
    },
    {
      "epoch": 2.4474330487662264,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 10463
    },
    {
      "epoch": 2.447666939539235,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.2801,
      "step": 10464
    },
    {
      "epoch": 2.447900830312244,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 10465
    },
    {
      "epoch": 2.4481347210852533,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 10466
    },
    {
      "epoch": 2.448368611858262,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 10467
    },
    {
      "epoch": 2.448602502631271,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6671,
      "step": 10468
    },
    {
      "epoch": 2.4488363934042803,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 10469
    },
    {
      "epoch": 2.449070284177289,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 10470
    },
    {
      "epoch": 2.4493041749502984,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 10471
    },
    {
      "epoch": 2.4495380657233072,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 10472
    },
    {
      "epoch": 2.449771956496316,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.988,
      "step": 10473
    },
    {
      "epoch": 2.4500058472693254,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6956,
      "step": 10474
    },
    {
      "epoch": 2.450239738042334,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 10475
    },
    {
      "epoch": 2.4504736288153435,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 10476
    },
    {
      "epoch": 2.4507075195883523,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 10477
    },
    {
      "epoch": 2.450941410361361,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 10478
    },
    {
      "epoch": 2.4511753011343704,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 10479
    },
    {
      "epoch": 2.4514091919073793,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 10480
    },
    {
      "epoch": 2.451643082680388,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 10481
    },
    {
      "epoch": 2.4518769734533974,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 10482
    },
    {
      "epoch": 2.4521108642264062,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 10483
    },
    {
      "epoch": 2.452344754999415,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 10484
    },
    {
      "epoch": 2.4525786457724243,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 10485
    },
    {
      "epoch": 2.452812536545433,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 10486
    },
    {
      "epoch": 2.4530464273184425,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6431,
      "step": 10487
    },
    {
      "epoch": 2.4532803180914513,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 10488
    },
    {
      "epoch": 2.45351420886446,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 10489
    },
    {
      "epoch": 2.4537480996374694,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 10490
    },
    {
      "epoch": 2.4539819904104783,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 10491
    },
    {
      "epoch": 2.4542158811834875,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0902,
      "step": 10492
    },
    {
      "epoch": 2.4544497719564964,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 10493
    },
    {
      "epoch": 2.4546836627295052,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 10494
    },
    {
      "epoch": 2.4549175535025145,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 10495
    },
    {
      "epoch": 2.4551514442755233,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 10496
    },
    {
      "epoch": 2.455385335048532,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 10497
    },
    {
      "epoch": 2.4556192258215415,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 10498
    },
    {
      "epoch": 2.4558531165945503,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5587,
      "step": 10499
    },
    {
      "epoch": 2.456087007367559,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 10500
    },
    {
      "epoch": 2.456087007367559,
      "eval_runtime": 4.5892,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 10500
    },
    {
      "epoch": 2.4563208981405684,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 10501
    },
    {
      "epoch": 2.4565547889135773,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 10502
    },
    {
      "epoch": 2.4567886796865865,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 10503
    },
    {
      "epoch": 2.4570225704595954,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 10504
    },
    {
      "epoch": 2.457256461232604,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8096,
      "step": 10505
    },
    {
      "epoch": 2.4574903520056135,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 10506
    },
    {
      "epoch": 2.4577242427786223,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 10507
    },
    {
      "epoch": 2.4579581335516316,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 10508
    },
    {
      "epoch": 2.4581920243246405,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5825,
      "step": 10509
    },
    {
      "epoch": 2.4584259150976493,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 10510
    },
    {
      "epoch": 2.4586598058706586,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 10511
    },
    {
      "epoch": 2.4588936966436674,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 10512
    },
    {
      "epoch": 2.4591275874166763,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6541,
      "step": 10513
    },
    {
      "epoch": 2.4593614781896855,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.5399,
      "step": 10514
    },
    {
      "epoch": 2.4595953689626944,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6745,
      "step": 10515
    },
    {
      "epoch": 2.459829259735703,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 10516
    },
    {
      "epoch": 2.4600631505087125,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8135,
      "step": 10517
    },
    {
      "epoch": 2.4602970412817213,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 10518
    },
    {
      "epoch": 2.4605309320547306,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 10519
    },
    {
      "epoch": 2.4607648228277395,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 10520
    },
    {
      "epoch": 2.4609987136007483,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 10521
    },
    {
      "epoch": 2.4612326043737576,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8366,
      "step": 10522
    },
    {
      "epoch": 2.4614664951467664,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 10523
    },
    {
      "epoch": 2.4617003859197757,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 10524
    },
    {
      "epoch": 2.4619342766927845,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7627,
      "step": 10525
    },
    {
      "epoch": 2.4621681674657934,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 10526
    },
    {
      "epoch": 2.4624020582388026,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 10527
    },
    {
      "epoch": 2.4626359490118115,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 10528
    },
    {
      "epoch": 2.4628698397848203,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 10529
    },
    {
      "epoch": 2.4631037305578296,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 10530
    },
    {
      "epoch": 2.4633376213308384,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 10531
    },
    {
      "epoch": 2.4635715121038473,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 10532
    },
    {
      "epoch": 2.4638054028768566,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7481,
      "step": 10533
    },
    {
      "epoch": 2.4640392936498654,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 10534
    },
    {
      "epoch": 2.4642731844228747,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.5087,
      "step": 10535
    },
    {
      "epoch": 2.4645070751958835,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 10536
    },
    {
      "epoch": 2.4647409659688924,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6612,
      "step": 10537
    },
    {
      "epoch": 2.4649748567419016,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 10538
    },
    {
      "epoch": 2.4652087475149105,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 10539
    },
    {
      "epoch": 2.4654426382879198,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 10540
    },
    {
      "epoch": 2.4656765290609286,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 10541
    },
    {
      "epoch": 2.4659104198339374,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 10542
    },
    {
      "epoch": 2.4661443106069467,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 10543
    },
    {
      "epoch": 2.4663782013799556,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 10544
    },
    {
      "epoch": 2.4666120921529644,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 10545
    },
    {
      "epoch": 2.4668459829259737,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5771,
      "step": 10546
    },
    {
      "epoch": 2.4670798736989825,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 10547
    },
    {
      "epoch": 2.4673137644719914,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 10548
    },
    {
      "epoch": 2.4675476552450006,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 10549
    },
    {
      "epoch": 2.4677815460180095,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 10550
    },
    {
      "epoch": 2.4680154367910188,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 10551
    },
    {
      "epoch": 2.4682493275640276,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 10552
    },
    {
      "epoch": 2.4684832183370364,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 10553
    },
    {
      "epoch": 2.4687171091100457,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 10554
    },
    {
      "epoch": 2.4689509998830546,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 10555
    },
    {
      "epoch": 2.469184890656064,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 10556
    },
    {
      "epoch": 2.4694187814290727,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 10557
    },
    {
      "epoch": 2.4696526722020815,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 10558
    },
    {
      "epoch": 2.469886562975091,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6563,
      "step": 10559
    },
    {
      "epoch": 2.4701204537480996,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 10560
    },
    {
      "epoch": 2.4703543445211085,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 10561
    },
    {
      "epoch": 2.4705882352941178,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 10562
    },
    {
      "epoch": 2.4708221260671266,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8728,
      "step": 10563
    },
    {
      "epoch": 2.4710560168401354,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 10564
    },
    {
      "epoch": 2.4712899076131447,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 10565
    },
    {
      "epoch": 2.4715237983861535,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 10566
    },
    {
      "epoch": 2.471757689159163,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8701,
      "step": 10567
    },
    {
      "epoch": 2.4719915799321717,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 10568
    },
    {
      "epoch": 2.4722254707051805,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2447,
      "step": 10569
    },
    {
      "epoch": 2.47245936147819,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 10570
    },
    {
      "epoch": 2.4726932522511986,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 10571
    },
    {
      "epoch": 2.472927143024208,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 10572
    },
    {
      "epoch": 2.4731610337972167,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9075,
      "step": 10573
    },
    {
      "epoch": 2.4733949245702256,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 10574
    },
    {
      "epoch": 2.473628815343235,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.988,
      "step": 10575
    },
    {
      "epoch": 2.4738627061162437,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 10576
    },
    {
      "epoch": 2.4740965968892525,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 10577
    },
    {
      "epoch": 2.474330487662262,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 10578
    },
    {
      "epoch": 2.4745643784352707,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 10579
    },
    {
      "epoch": 2.4747982692082795,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 10580
    },
    {
      "epoch": 2.475032159981289,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 10581
    },
    {
      "epoch": 2.4752660507542976,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 10582
    },
    {
      "epoch": 2.475499941527307,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8223,
      "step": 10583
    },
    {
      "epoch": 2.4757338323003157,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 10584
    },
    {
      "epoch": 2.4759677230733246,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8456,
      "step": 10585
    },
    {
      "epoch": 2.476201613846334,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 10586
    },
    {
      "epoch": 2.4764355046193427,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6641,
      "step": 10587
    },
    {
      "epoch": 2.476669395392352,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0391,
      "step": 10588
    },
    {
      "epoch": 2.476903286165361,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 10589
    },
    {
      "epoch": 2.4771371769383697,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6878,
      "step": 10590
    },
    {
      "epoch": 2.477371067711379,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 10591
    },
    {
      "epoch": 2.4776049584843878,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7663,
      "step": 10592
    },
    {
      "epoch": 2.4778388492573966,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.306,
      "step": 10593
    },
    {
      "epoch": 2.478072740030406,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 10594
    },
    {
      "epoch": 2.4783066308034147,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 10595
    },
    {
      "epoch": 2.4785405215764236,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 10596
    },
    {
      "epoch": 2.478774412349433,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 10597
    },
    {
      "epoch": 2.4790083031224417,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 10598
    },
    {
      "epoch": 2.479242193895451,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 10599
    },
    {
      "epoch": 2.47947608466846,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 10600
    },
    {
      "epoch": 2.47947608466846,
      "eval_runtime": 4.6151,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 10600
    },
    {
      "epoch": 2.4797099754414687,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 10601
    },
    {
      "epoch": 2.479943866214478,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 10602
    },
    {
      "epoch": 2.4801777569874868,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 10603
    },
    {
      "epoch": 2.480411647760496,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 10604
    },
    {
      "epoch": 2.480645538533505,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 10605
    },
    {
      "epoch": 2.4808794293065137,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 10606
    },
    {
      "epoch": 2.481113320079523,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 10607
    },
    {
      "epoch": 2.481347210852532,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 10608
    },
    {
      "epoch": 2.4815811016255407,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 10609
    },
    {
      "epoch": 2.48181499239855,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 10610
    },
    {
      "epoch": 2.482048883171559,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7528,
      "step": 10611
    },
    {
      "epoch": 2.4822827739445676,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 10612
    },
    {
      "epoch": 2.482516664717577,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 10613
    },
    {
      "epoch": 2.4827505554905858,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 10614
    },
    {
      "epoch": 2.482984446263595,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.3896,
      "step": 10615
    },
    {
      "epoch": 2.483218337036604,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 10616
    },
    {
      "epoch": 2.4834522278096127,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.542,
      "step": 10617
    },
    {
      "epoch": 2.483686118582622,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 10618
    },
    {
      "epoch": 2.483920009355631,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 10619
    },
    {
      "epoch": 2.48415390012864,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 10620
    },
    {
      "epoch": 2.484387790901649,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 10621
    },
    {
      "epoch": 2.484621681674658,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 10622
    },
    {
      "epoch": 2.484855572447667,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 10623
    },
    {
      "epoch": 2.485089463220676,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 10624
    },
    {
      "epoch": 2.4853233539936848,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 10625
    },
    {
      "epoch": 2.485557244766694,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 10626
    },
    {
      "epoch": 2.485791135539703,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5734,
      "step": 10627
    },
    {
      "epoch": 2.4860250263127117,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7604,
      "step": 10628
    },
    {
      "epoch": 2.486258917085721,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7188,
      "step": 10629
    },
    {
      "epoch": 2.48649280785873,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 10630
    },
    {
      "epoch": 2.486726698631739,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 10631
    },
    {
      "epoch": 2.486960589404748,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 10632
    },
    {
      "epoch": 2.487194480177757,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.078,
      "step": 10633
    },
    {
      "epoch": 2.487428370950766,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 10634
    },
    {
      "epoch": 2.487662261723775,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4091,
      "step": 10635
    },
    {
      "epoch": 2.487896152496784,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.682,
      "step": 10636
    },
    {
      "epoch": 2.488130043269793,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 10637
    },
    {
      "epoch": 2.488363934042802,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7178,
      "step": 10638
    },
    {
      "epoch": 2.488597824815811,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 10639
    },
    {
      "epoch": 2.48883171558882,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 10640
    },
    {
      "epoch": 2.4890656063618293,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7328,
      "step": 10641
    },
    {
      "epoch": 2.489299497134838,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 10642
    },
    {
      "epoch": 2.489533387907847,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6324,
      "step": 10643
    },
    {
      "epoch": 2.4897672786808562,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 10644
    },
    {
      "epoch": 2.490001169453865,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 10645
    },
    {
      "epoch": 2.490235060226874,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 10646
    },
    {
      "epoch": 2.490468950999883,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 10647
    },
    {
      "epoch": 2.490702841772892,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 10648
    },
    {
      "epoch": 2.490936732545901,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 10649
    },
    {
      "epoch": 2.49117062331891,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 10650
    },
    {
      "epoch": 2.491404514091919,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 10651
    },
    {
      "epoch": 2.4916384048649283,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.2873,
      "step": 10652
    },
    {
      "epoch": 2.491872295637937,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 10653
    },
    {
      "epoch": 2.492106186410946,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 10654
    },
    {
      "epoch": 2.4923400771839552,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 10655
    },
    {
      "epoch": 2.492573967956964,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 10656
    },
    {
      "epoch": 2.4928078587299733,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 10657
    },
    {
      "epoch": 2.493041749502982,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2576,
      "step": 10658
    },
    {
      "epoch": 2.493275640275991,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 10659
    },
    {
      "epoch": 2.4935095310490003,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 10660
    },
    {
      "epoch": 2.493743421822009,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 10661
    },
    {
      "epoch": 2.493977312595018,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 10662
    },
    {
      "epoch": 2.4942112033680273,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 10663
    },
    {
      "epoch": 2.494445094141036,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 10664
    },
    {
      "epoch": 2.494678984914045,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 10665
    },
    {
      "epoch": 2.494912875687054,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 10666
    },
    {
      "epoch": 2.495146766460063,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 10667
    },
    {
      "epoch": 2.4953806572330723,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 10668
    },
    {
      "epoch": 2.495614548006081,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 10669
    },
    {
      "epoch": 2.49584843877909,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3692,
      "step": 10670
    },
    {
      "epoch": 2.4960823295520993,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 10671
    },
    {
      "epoch": 2.496316220325108,
      "grad_norm": 9.375,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 10672
    },
    {
      "epoch": 2.4965501110981174,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 10673
    },
    {
      "epoch": 2.4967840018711263,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6866,
      "step": 10674
    },
    {
      "epoch": 2.497017892644135,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 10675
    },
    {
      "epoch": 2.4972517834171444,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5919,
      "step": 10676
    },
    {
      "epoch": 2.497485674190153,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2343,
      "step": 10677
    },
    {
      "epoch": 2.497719564963162,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 10678
    },
    {
      "epoch": 2.4979534557361713,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 10679
    },
    {
      "epoch": 2.49818734650918,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 10680
    },
    {
      "epoch": 2.498421237282189,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 10681
    },
    {
      "epoch": 2.4986551280551983,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 10682
    },
    {
      "epoch": 2.498889018828207,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 10683
    },
    {
      "epoch": 2.4991229096012164,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 10684
    },
    {
      "epoch": 2.4993568003742253,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 10685
    },
    {
      "epoch": 2.499590691147234,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 10686
    },
    {
      "epoch": 2.4998245819202434,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 10687
    },
    {
      "epoch": 2.500058472693252,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 10688
    },
    {
      "epoch": 2.5002923634662615,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 10689
    },
    {
      "epoch": 2.5005262542392703,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 10690
    },
    {
      "epoch": 2.500760145012279,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 10691
    },
    {
      "epoch": 2.500994035785288,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 10692
    },
    {
      "epoch": 2.5012279265582973,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 10693
    },
    {
      "epoch": 2.501461817331306,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8306,
      "step": 10694
    },
    {
      "epoch": 2.5016957081043154,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 10695
    },
    {
      "epoch": 2.5019295988773242,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 10696
    },
    {
      "epoch": 2.502163489650333,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 10697
    },
    {
      "epoch": 2.5023973804233424,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 10698
    },
    {
      "epoch": 2.502631271196351,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 10699
    },
    {
      "epoch": 2.5028651619693605,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9556,
      "step": 10700
    },
    {
      "epoch": 2.5028651619693605,
      "eval_runtime": 4.6239,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 10700
    },
    {
      "epoch": 2.5030990527423693,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.714,
      "step": 10701
    },
    {
      "epoch": 2.503332943515378,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 10702
    },
    {
      "epoch": 2.5035668342883874,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5886,
      "step": 10703
    },
    {
      "epoch": 2.5038007250613963,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 10704
    },
    {
      "epoch": 2.5040346158344056,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 10705
    },
    {
      "epoch": 2.5042685066074144,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 10706
    },
    {
      "epoch": 2.5045023973804232,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 10707
    },
    {
      "epoch": 2.504736288153432,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 10708
    },
    {
      "epoch": 2.5049701789264414,
      "grad_norm": 2.40625,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 10709
    },
    {
      "epoch": 2.50520406969945,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 10710
    },
    {
      "epoch": 2.5054379604724595,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 10711
    },
    {
      "epoch": 2.5056718512454683,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7285,
      "step": 10712
    },
    {
      "epoch": 2.505905742018477,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 10713
    },
    {
      "epoch": 2.5061396327914864,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 10714
    },
    {
      "epoch": 2.5063735235644953,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 10715
    },
    {
      "epoch": 2.5066074143375046,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 10716
    },
    {
      "epoch": 2.5068413051105134,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 10717
    },
    {
      "epoch": 2.5070751958835222,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 10718
    },
    {
      "epoch": 2.5073090866565315,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 10719
    },
    {
      "epoch": 2.5075429774295404,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 10720
    },
    {
      "epoch": 2.5077768682025496,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 10721
    },
    {
      "epoch": 2.5080107589755585,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 10722
    },
    {
      "epoch": 2.5082446497485673,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 10723
    },
    {
      "epoch": 2.508478540521576,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 10724
    },
    {
      "epoch": 2.5087124312945854,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6745,
      "step": 10725
    },
    {
      "epoch": 2.5089463220675943,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 10726
    },
    {
      "epoch": 2.5091802128406036,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 10727
    },
    {
      "epoch": 2.5094141036136124,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6071,
      "step": 10728
    },
    {
      "epoch": 2.5096479943866212,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8738,
      "step": 10729
    },
    {
      "epoch": 2.5098818851596305,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 10730
    },
    {
      "epoch": 2.5101157759326393,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6493,
      "step": 10731
    },
    {
      "epoch": 2.5103496667056486,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 10732
    },
    {
      "epoch": 2.5105835574786575,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 10733
    },
    {
      "epoch": 2.5108174482516663,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 10734
    },
    {
      "epoch": 2.5110513390246756,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 10735
    },
    {
      "epoch": 2.5112852297976844,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 10736
    },
    {
      "epoch": 2.5115191205706937,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6621,
      "step": 10737
    },
    {
      "epoch": 2.5117530113437025,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 10738
    },
    {
      "epoch": 2.5119869021167114,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 10739
    },
    {
      "epoch": 2.5122207928897207,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4701,
      "step": 10740
    },
    {
      "epoch": 2.5124546836627295,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6438,
      "step": 10741
    },
    {
      "epoch": 2.512688574435739,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 10742
    },
    {
      "epoch": 2.5129224652087476,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6946,
      "step": 10743
    },
    {
      "epoch": 2.5131563559817565,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 10744
    },
    {
      "epoch": 2.5133902467547653,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 10745
    },
    {
      "epoch": 2.5136241375277746,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4393,
      "step": 10746
    },
    {
      "epoch": 2.5138580283007834,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 10747
    },
    {
      "epoch": 2.5140919190737927,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 10748
    },
    {
      "epoch": 2.5143258098468015,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 10749
    },
    {
      "epoch": 2.5145597006198104,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6559,
      "step": 10750
    },
    {
      "epoch": 2.5147935913928197,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7622,
      "step": 10751
    },
    {
      "epoch": 2.5150274821658285,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 10752
    },
    {
      "epoch": 2.515261372938838,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 10753
    },
    {
      "epoch": 2.5154952637118466,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 10754
    },
    {
      "epoch": 2.5157291544848555,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 10755
    },
    {
      "epoch": 2.5159630452578647,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 10756
    },
    {
      "epoch": 2.5161969360308736,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 10757
    },
    {
      "epoch": 2.516430826803883,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 10758
    },
    {
      "epoch": 2.5166647175768917,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 10759
    },
    {
      "epoch": 2.5168986083499005,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 10760
    },
    {
      "epoch": 2.5171324991229094,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5744,
      "step": 10761
    },
    {
      "epoch": 2.5173663898959187,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7362,
      "step": 10762
    },
    {
      "epoch": 2.5176002806689275,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 10763
    },
    {
      "epoch": 2.5178341714419368,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 10764
    },
    {
      "epoch": 2.5180680622149456,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7366,
      "step": 10765
    },
    {
      "epoch": 2.5183019529879545,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 10766
    },
    {
      "epoch": 2.5185358437609637,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 10767
    },
    {
      "epoch": 2.5187697345339726,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 10768
    },
    {
      "epoch": 2.519003625306982,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 10769
    },
    {
      "epoch": 2.5192375160799907,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 10770
    },
    {
      "epoch": 2.5194714068529995,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 10771
    },
    {
      "epoch": 2.519705297626009,
      "grad_norm": 7.34375,
      "learning_rate": 3e-05,
      "loss": 2.2257,
      "step": 10772
    },
    {
      "epoch": 2.5199391883990176,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 10773
    },
    {
      "epoch": 2.520173079172027,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.772,
      "step": 10774
    },
    {
      "epoch": 2.5204069699450358,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 10775
    },
    {
      "epoch": 2.5206408607180446,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 10776
    },
    {
      "epoch": 2.5208747514910534,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7185,
      "step": 10777
    },
    {
      "epoch": 2.5211086422640627,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9301,
      "step": 10778
    },
    {
      "epoch": 2.5213425330370716,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 10779
    },
    {
      "epoch": 2.521576423810081,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 10780
    },
    {
      "epoch": 2.5218103145830897,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 10781
    },
    {
      "epoch": 2.5220442053560985,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 10782
    },
    {
      "epoch": 2.522278096129108,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.5861,
      "step": 10783
    },
    {
      "epoch": 2.5225119869021166,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 10784
    },
    {
      "epoch": 2.522745877675126,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 10785
    },
    {
      "epoch": 2.5229797684481348,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 10786
    },
    {
      "epoch": 2.5232136592211436,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 10787
    },
    {
      "epoch": 2.523447549994153,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0299,
      "step": 10788
    },
    {
      "epoch": 2.5236814407671617,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 10789
    },
    {
      "epoch": 2.523915331540171,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 10790
    },
    {
      "epoch": 2.52414922231318,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 10791
    },
    {
      "epoch": 2.5243831130861887,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8629,
      "step": 10792
    },
    {
      "epoch": 2.5246170038591975,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 10793
    },
    {
      "epoch": 2.524850894632207,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 10794
    },
    {
      "epoch": 2.5250847854052156,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8531,
      "step": 10795
    },
    {
      "epoch": 2.525318676178225,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 10796
    },
    {
      "epoch": 2.5255525669512338,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 10797
    },
    {
      "epoch": 2.5257864577242426,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7469,
      "step": 10798
    },
    {
      "epoch": 2.526020348497252,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 10799
    },
    {
      "epoch": 2.5262542392702607,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8043,
      "step": 10800
    },
    {
      "epoch": 2.5262542392702607,
      "eval_runtime": 4.6739,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 10800
    },
    {
      "epoch": 2.52648813004327,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 10801
    },
    {
      "epoch": 2.526722020816279,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 10802
    },
    {
      "epoch": 2.5269559115892877,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.769,
      "step": 10803
    },
    {
      "epoch": 2.527189802362297,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 10804
    },
    {
      "epoch": 2.527423693135306,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 10805
    },
    {
      "epoch": 2.527657583908315,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 10806
    },
    {
      "epoch": 2.527891474681324,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6554,
      "step": 10807
    },
    {
      "epoch": 2.5281253654543328,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 10808
    },
    {
      "epoch": 2.5283592562273416,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 10809
    },
    {
      "epoch": 2.528593147000351,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 10810
    },
    {
      "epoch": 2.5288270377733597,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 10811
    },
    {
      "epoch": 2.529060928546369,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 10812
    },
    {
      "epoch": 2.529294819319378,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 10813
    },
    {
      "epoch": 2.5295287100923867,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 10814
    },
    {
      "epoch": 2.529762600865396,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 10815
    },
    {
      "epoch": 2.529996491638405,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 10816
    },
    {
      "epoch": 2.530230382411414,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 10817
    },
    {
      "epoch": 2.530464273184423,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 10818
    },
    {
      "epoch": 2.5306981639574317,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7597,
      "step": 10819
    },
    {
      "epoch": 2.530932054730441,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.5828,
      "step": 10820
    },
    {
      "epoch": 2.53116594550345,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7669,
      "step": 10821
    },
    {
      "epoch": 2.531399836276459,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 10822
    },
    {
      "epoch": 2.531633727049468,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 10823
    },
    {
      "epoch": 2.531867617822477,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 10824
    },
    {
      "epoch": 2.5321015085954857,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 10825
    },
    {
      "epoch": 2.532335399368495,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 10826
    },
    {
      "epoch": 2.532569290141504,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6869,
      "step": 10827
    },
    {
      "epoch": 2.532803180914513,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 10828
    },
    {
      "epoch": 2.533037071687522,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.611,
      "step": 10829
    },
    {
      "epoch": 2.5332709624605307,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 10830
    },
    {
      "epoch": 2.53350485323354,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 10831
    },
    {
      "epoch": 2.533738744006549,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6301,
      "step": 10832
    },
    {
      "epoch": 2.533972634779558,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 10833
    },
    {
      "epoch": 2.534206525552567,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 10834
    },
    {
      "epoch": 2.534440416325576,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 10835
    },
    {
      "epoch": 2.534674307098585,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 10836
    },
    {
      "epoch": 2.534908197871594,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6264,
      "step": 10837
    },
    {
      "epoch": 2.535142088644603,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 10838
    },
    {
      "epoch": 2.535375979417612,
      "grad_norm": 17.375,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 10839
    },
    {
      "epoch": 2.535609870190621,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 10840
    },
    {
      "epoch": 2.5358437609636297,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 10841
    },
    {
      "epoch": 2.536077651736639,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 10842
    },
    {
      "epoch": 2.536311542509648,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 10843
    },
    {
      "epoch": 2.536545433282657,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 10844
    },
    {
      "epoch": 2.536779324055666,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 10845
    },
    {
      "epoch": 2.537013214828675,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 10846
    },
    {
      "epoch": 2.537247105601684,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 10847
    },
    {
      "epoch": 2.537480996374693,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 10848
    },
    {
      "epoch": 2.537714887147702,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 10849
    },
    {
      "epoch": 2.537948777920711,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 10850
    },
    {
      "epoch": 2.53818266869372,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 10851
    },
    {
      "epoch": 2.538416559466729,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 10852
    },
    {
      "epoch": 2.538650450239738,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 10853
    },
    {
      "epoch": 2.5388843410127473,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 10854
    },
    {
      "epoch": 2.539118231785756,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6245,
      "step": 10855
    },
    {
      "epoch": 2.539352122558765,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 10856
    },
    {
      "epoch": 2.539586013331774,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 10857
    },
    {
      "epoch": 2.539819904104783,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 10858
    },
    {
      "epoch": 2.540053794877792,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 10859
    },
    {
      "epoch": 2.540287685650801,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 10860
    },
    {
      "epoch": 2.54052157642381,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 10861
    },
    {
      "epoch": 2.540755467196819,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 10862
    },
    {
      "epoch": 2.540989357969828,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3334,
      "step": 10863
    },
    {
      "epoch": 2.541223248742837,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 10864
    },
    {
      "epoch": 2.5414571395158463,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8025,
      "step": 10865
    },
    {
      "epoch": 2.541691030288855,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 10866
    },
    {
      "epoch": 2.541924921061864,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 10867
    },
    {
      "epoch": 2.5421588118348732,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 10868
    },
    {
      "epoch": 2.542392702607882,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 10869
    },
    {
      "epoch": 2.5426265933808914,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 10870
    },
    {
      "epoch": 2.5428604841539,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 10871
    },
    {
      "epoch": 2.543094374926909,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 10872
    },
    {
      "epoch": 2.543328265699918,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 10873
    },
    {
      "epoch": 2.543562156472927,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 10874
    },
    {
      "epoch": 2.543796047245936,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 10875
    },
    {
      "epoch": 2.5440299380189453,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 10876
    },
    {
      "epoch": 2.544263828791954,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 10877
    },
    {
      "epoch": 2.544497719564963,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 10878
    },
    {
      "epoch": 2.5447316103379722,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 10879
    },
    {
      "epoch": 2.544965501110981,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 10880
    },
    {
      "epoch": 2.5451993918839904,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 10881
    },
    {
      "epoch": 2.545433282656999,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 10882
    },
    {
      "epoch": 2.545667173430008,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8549,
      "step": 10883
    },
    {
      "epoch": 2.5459010642030173,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 10884
    },
    {
      "epoch": 2.546134954976026,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 10885
    },
    {
      "epoch": 2.5463688457490354,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.3182,
      "step": 10886
    },
    {
      "epoch": 2.5466027365220443,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8644,
      "step": 10887
    },
    {
      "epoch": 2.546836627295053,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 10888
    },
    {
      "epoch": 2.547070518068062,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1821,
      "step": 10889
    },
    {
      "epoch": 2.5473044088410712,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6949,
      "step": 10890
    },
    {
      "epoch": 2.54753829961408,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 10891
    },
    {
      "epoch": 2.5477721903870894,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8361,
      "step": 10892
    },
    {
      "epoch": 2.548006081160098,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 10893
    },
    {
      "epoch": 2.548239971933107,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 10894
    },
    {
      "epoch": 2.5484738627061163,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 10895
    },
    {
      "epoch": 2.548707753479125,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 10896
    },
    {
      "epoch": 2.5489416442521344,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6788,
      "step": 10897
    },
    {
      "epoch": 2.5491755350251433,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8738,
      "step": 10898
    },
    {
      "epoch": 2.549409425798152,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 10899
    },
    {
      "epoch": 2.5496433165711614,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 10900
    },
    {
      "epoch": 2.5496433165711614,
      "eval_runtime": 4.6512,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 10900
    },
    {
      "epoch": 2.5498772073441702,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 10901
    },
    {
      "epoch": 2.5501110981171795,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 10902
    },
    {
      "epoch": 2.5503449888901883,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 10903
    },
    {
      "epoch": 2.550578879663197,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 10904
    },
    {
      "epoch": 2.550812770436206,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 10905
    },
    {
      "epoch": 2.5510466612092153,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 10906
    },
    {
      "epoch": 2.551280551982224,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 10907
    },
    {
      "epoch": 2.5515144427552334,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6609,
      "step": 10908
    },
    {
      "epoch": 2.5517483335282423,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 10909
    },
    {
      "epoch": 2.551982224301251,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 10910
    },
    {
      "epoch": 2.5522161150742604,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6602,
      "step": 10911
    },
    {
      "epoch": 2.552450005847269,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 10912
    },
    {
      "epoch": 2.5526838966202785,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7375,
      "step": 10913
    },
    {
      "epoch": 2.5529177873932873,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 10914
    },
    {
      "epoch": 2.553151678166296,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 10915
    },
    {
      "epoch": 2.5533855689393055,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 10916
    },
    {
      "epoch": 2.5536194597123143,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 10917
    },
    {
      "epoch": 2.5538533504853236,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 10918
    },
    {
      "epoch": 2.5540872412583324,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 10919
    },
    {
      "epoch": 2.5543211320313413,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 10920
    },
    {
      "epoch": 2.55455502280435,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6258,
      "step": 10921
    },
    {
      "epoch": 2.5547889135773594,
      "grad_norm": 9.25,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 10922
    },
    {
      "epoch": 2.555022804350368,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7745,
      "step": 10923
    },
    {
      "epoch": 2.5552566951233775,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 10924
    },
    {
      "epoch": 2.5554905858963863,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 10925
    },
    {
      "epoch": 2.555724476669395,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 10926
    },
    {
      "epoch": 2.5559583674424045,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 10927
    },
    {
      "epoch": 2.5561922582154133,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 10928
    },
    {
      "epoch": 2.5564261489884226,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 10929
    },
    {
      "epoch": 2.5566600397614314,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 10930
    },
    {
      "epoch": 2.5568939305344403,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 10931
    },
    {
      "epoch": 2.5571278213074495,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 10932
    },
    {
      "epoch": 2.5573617120804584,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6447,
      "step": 10933
    },
    {
      "epoch": 2.5575956028534677,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 10934
    },
    {
      "epoch": 2.5578294936264765,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5304,
      "step": 10935
    },
    {
      "epoch": 2.5580633843994853,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6413,
      "step": 10936
    },
    {
      "epoch": 2.5582972751724946,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 10937
    },
    {
      "epoch": 2.5585311659455034,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 10938
    },
    {
      "epoch": 2.5587650567185127,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 10939
    },
    {
      "epoch": 2.5589989474915216,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7306,
      "step": 10940
    },
    {
      "epoch": 2.5592328382645304,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 10941
    },
    {
      "epoch": 2.5594667290375392,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 10942
    },
    {
      "epoch": 2.5597006198105485,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 10943
    },
    {
      "epoch": 2.5599345105835574,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6188,
      "step": 10944
    },
    {
      "epoch": 2.5601684013565666,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 10945
    },
    {
      "epoch": 2.5604022921295755,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 10946
    },
    {
      "epoch": 2.5606361829025843,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 10947
    },
    {
      "epoch": 2.5608700736755936,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 10948
    },
    {
      "epoch": 2.5611039644486024,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 10949
    },
    {
      "epoch": 2.5613378552216117,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 10950
    },
    {
      "epoch": 2.5615717459946206,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5388,
      "step": 10951
    },
    {
      "epoch": 2.5618056367676294,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 10952
    },
    {
      "epoch": 2.5620395275406387,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 10953
    },
    {
      "epoch": 2.5622734183136475,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 10954
    },
    {
      "epoch": 2.562507309086657,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 10955
    },
    {
      "epoch": 2.5627411998596656,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5184,
      "step": 10956
    },
    {
      "epoch": 2.5629750906326745,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1004,
      "step": 10957
    },
    {
      "epoch": 2.5632089814056833,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 10958
    },
    {
      "epoch": 2.5634428721786926,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 10959
    },
    {
      "epoch": 2.5636767629517014,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 10960
    },
    {
      "epoch": 2.5639106537247107,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 10961
    },
    {
      "epoch": 2.5641445444977196,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.267,
      "step": 10962
    },
    {
      "epoch": 2.5643784352707284,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 10963
    },
    {
      "epoch": 2.5646123260437377,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 10964
    },
    {
      "epoch": 2.5648462168167465,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 10965
    },
    {
      "epoch": 2.565080107589756,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 10966
    },
    {
      "epoch": 2.5653139983627646,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 10967
    },
    {
      "epoch": 2.5655478891357735,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 10968
    },
    {
      "epoch": 2.5657817799087828,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4302,
      "step": 10969
    },
    {
      "epoch": 2.5660156706817916,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 10970
    },
    {
      "epoch": 2.566249561454801,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 10971
    },
    {
      "epoch": 2.5664834522278097,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 10972
    },
    {
      "epoch": 2.5667173430008186,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 10973
    },
    {
      "epoch": 2.5669512337738274,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8456,
      "step": 10974
    },
    {
      "epoch": 2.5671851245468367,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 10975
    },
    {
      "epoch": 2.5674190153198455,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 10976
    },
    {
      "epoch": 2.567652906092855,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 10977
    },
    {
      "epoch": 2.5678867968658636,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 10978
    },
    {
      "epoch": 2.5681206876388725,
      "grad_norm": 12.0625,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 10979
    },
    {
      "epoch": 2.5683545784118818,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 10980
    },
    {
      "epoch": 2.5685884691848906,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 10981
    },
    {
      "epoch": 2.5688223599579,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 10982
    },
    {
      "epoch": 2.5690562507309087,
      "grad_norm": 8.375,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 10983
    },
    {
      "epoch": 2.5692901415039175,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 10984
    },
    {
      "epoch": 2.569524032276927,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 10985
    },
    {
      "epoch": 2.5697579230499357,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5705,
      "step": 10986
    },
    {
      "epoch": 2.569991813822945,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3345,
      "step": 10987
    },
    {
      "epoch": 2.570225704595954,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 10988
    },
    {
      "epoch": 2.5704595953689626,
      "grad_norm": 8.1875,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 10989
    },
    {
      "epoch": 2.5706934861419715,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 10990
    },
    {
      "epoch": 2.5709273769149807,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 10991
    },
    {
      "epoch": 2.5711612676879896,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6372,
      "step": 10992
    },
    {
      "epoch": 2.571395158460999,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6825,
      "step": 10993
    },
    {
      "epoch": 2.5716290492340077,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7284,
      "step": 10994
    },
    {
      "epoch": 2.5718629400070165,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1198,
      "step": 10995
    },
    {
      "epoch": 2.572096830780026,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8905,
      "step": 10996
    },
    {
      "epoch": 2.5723307215530347,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 10997
    },
    {
      "epoch": 2.572564612326044,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 10998
    },
    {
      "epoch": 2.572798503099053,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 10999
    },
    {
      "epoch": 2.5730323938720616,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 11000
    },
    {
      "epoch": 2.5730323938720616,
      "eval_runtime": 4.5967,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 11000
    },
    {
      "epoch": 2.573266284645071,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 11001
    },
    {
      "epoch": 2.5735001754180797,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 11002
    },
    {
      "epoch": 2.573734066191089,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 11003
    },
    {
      "epoch": 2.573967956964098,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 11004
    },
    {
      "epoch": 2.5742018477371067,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 11005
    },
    {
      "epoch": 2.5744357385101155,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 11006
    },
    {
      "epoch": 2.574669629283125,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7579,
      "step": 11007
    },
    {
      "epoch": 2.5749035200561337,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 11008
    },
    {
      "epoch": 2.575137410829143,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.599,
      "step": 11009
    },
    {
      "epoch": 2.5753713016021518,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 11010
    },
    {
      "epoch": 2.5756051923751606,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 11011
    },
    {
      "epoch": 2.57583908314817,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 11012
    },
    {
      "epoch": 2.5760729739211787,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.77,
      "step": 11013
    },
    {
      "epoch": 2.576306864694188,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 11014
    },
    {
      "epoch": 2.576540755467197,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 11015
    },
    {
      "epoch": 2.5767746462402057,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 11016
    },
    {
      "epoch": 2.577008537013215,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 11017
    },
    {
      "epoch": 2.577242427786224,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8411,
      "step": 11018
    },
    {
      "epoch": 2.577476318559233,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 11019
    },
    {
      "epoch": 2.577710209332242,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 11020
    },
    {
      "epoch": 2.5779441001052508,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 11021
    },
    {
      "epoch": 2.5781779908782596,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 11022
    },
    {
      "epoch": 2.578411881651269,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 11023
    },
    {
      "epoch": 2.5786457724242777,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 11024
    },
    {
      "epoch": 2.578879663197287,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 11025
    },
    {
      "epoch": 2.579113553970296,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 11026
    },
    {
      "epoch": 2.5793474447433047,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 11027
    },
    {
      "epoch": 2.579581335516314,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 11028
    },
    {
      "epoch": 2.579815226289323,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 11029
    },
    {
      "epoch": 2.580049117062332,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8714,
      "step": 11030
    },
    {
      "epoch": 2.580283007835341,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 11031
    },
    {
      "epoch": 2.5805168986083498,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 11032
    },
    {
      "epoch": 2.580750789381359,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 11033
    },
    {
      "epoch": 2.580984680154368,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 11034
    },
    {
      "epoch": 2.581218570927377,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 11035
    },
    {
      "epoch": 2.581452461700386,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0806,
      "step": 11036
    },
    {
      "epoch": 2.581686352473395,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 11037
    },
    {
      "epoch": 2.5819202432464037,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 11038
    },
    {
      "epoch": 2.582154134019413,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 11039
    },
    {
      "epoch": 2.582388024792422,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 11040
    },
    {
      "epoch": 2.582621915565431,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 11041
    },
    {
      "epoch": 2.58285580633844,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 11042
    },
    {
      "epoch": 2.5830896971114488,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 11043
    },
    {
      "epoch": 2.583323587884458,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9103,
      "step": 11044
    },
    {
      "epoch": 2.583557478657467,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 11045
    },
    {
      "epoch": 2.583791369430476,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 11046
    },
    {
      "epoch": 2.584025260203485,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 11047
    },
    {
      "epoch": 2.584259150976494,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 11048
    },
    {
      "epoch": 2.584493041749503,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 11049
    },
    {
      "epoch": 2.584726932522512,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5497,
      "step": 11050
    },
    {
      "epoch": 2.5849608232955212,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 11051
    },
    {
      "epoch": 2.58519471406853,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 11052
    },
    {
      "epoch": 2.585428604841539,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7175,
      "step": 11053
    },
    {
      "epoch": 2.5856624956145478,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 11054
    },
    {
      "epoch": 2.585896386387557,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 11055
    },
    {
      "epoch": 2.586130277160566,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 11056
    },
    {
      "epoch": 2.586364167933575,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 11057
    },
    {
      "epoch": 2.586598058706584,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 11058
    },
    {
      "epoch": 2.586831949479593,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 11059
    },
    {
      "epoch": 2.587065840252602,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 11060
    },
    {
      "epoch": 2.587299731025611,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 11061
    },
    {
      "epoch": 2.5875336217986202,
      "grad_norm": 9.8125,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 11062
    },
    {
      "epoch": 2.587767512571629,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 11063
    },
    {
      "epoch": 2.588001403344638,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5033,
      "step": 11064
    },
    {
      "epoch": 2.588235294117647,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 11065
    },
    {
      "epoch": 2.588469184890656,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 11066
    },
    {
      "epoch": 2.5887030756636653,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 11067
    },
    {
      "epoch": 2.588936966436674,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 11068
    },
    {
      "epoch": 2.589170857209683,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7416,
      "step": 11069
    },
    {
      "epoch": 2.589404747982692,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 11070
    },
    {
      "epoch": 2.589638638755701,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 11071
    },
    {
      "epoch": 2.58987252952871,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 11072
    },
    {
      "epoch": 2.5901064203017192,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 11073
    },
    {
      "epoch": 2.590340311074728,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 11074
    },
    {
      "epoch": 2.590574201847737,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 11075
    },
    {
      "epoch": 2.590808092620746,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6099,
      "step": 11076
    },
    {
      "epoch": 2.591041983393755,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 11077
    },
    {
      "epoch": 2.5912758741667643,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5592,
      "step": 11078
    },
    {
      "epoch": 2.591509764939773,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 11079
    },
    {
      "epoch": 2.591743655712782,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 11080
    },
    {
      "epoch": 2.5919775464857913,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 11081
    },
    {
      "epoch": 2.5922114372588,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 11082
    },
    {
      "epoch": 2.5924453280318094,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6109,
      "step": 11083
    },
    {
      "epoch": 2.592679218804818,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 11084
    },
    {
      "epoch": 2.592913109577827,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 11085
    },
    {
      "epoch": 2.593147000350836,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 11086
    },
    {
      "epoch": 2.593380891123845,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 11087
    },
    {
      "epoch": 2.593614781896854,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.5058,
      "step": 11088
    },
    {
      "epoch": 2.5938486726698633,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 11089
    },
    {
      "epoch": 2.594082563442872,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 11090
    },
    {
      "epoch": 2.594316454215881,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 11091
    },
    {
      "epoch": 2.5945503449888903,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 11092
    },
    {
      "epoch": 2.594784235761899,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 11093
    },
    {
      "epoch": 2.5950181265349084,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5212,
      "step": 11094
    },
    {
      "epoch": 2.595252017307917,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 11095
    },
    {
      "epoch": 2.595485908080926,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 11096
    },
    {
      "epoch": 2.5957197988539353,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 11097
    },
    {
      "epoch": 2.595953689626944,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6967,
      "step": 11098
    },
    {
      "epoch": 2.5961875803999535,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 11099
    },
    {
      "epoch": 2.5964214711729623,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 11100
    },
    {
      "epoch": 2.5964214711729623,
      "eval_runtime": 4.6401,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 11100
    },
    {
      "epoch": 2.596655361945971,
      "grad_norm": 7.8125,
      "learning_rate": 3e-05,
      "loss": 2.3797,
      "step": 11101
    },
    {
      "epoch": 2.59688925271898,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 11102
    },
    {
      "epoch": 2.5971231434919893,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 11103
    },
    {
      "epoch": 2.597357034264998,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 11104
    },
    {
      "epoch": 2.5975909250380074,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 11105
    },
    {
      "epoch": 2.597824815811016,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2968,
      "step": 11106
    },
    {
      "epoch": 2.598058706584025,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7437,
      "step": 11107
    },
    {
      "epoch": 2.5982925973570343,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 11108
    },
    {
      "epoch": 2.598526488130043,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 11109
    },
    {
      "epoch": 2.5987603789030524,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 11110
    },
    {
      "epoch": 2.5989942696760613,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 11111
    },
    {
      "epoch": 2.59922816044907,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 11112
    },
    {
      "epoch": 2.5994620512220794,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 11113
    },
    {
      "epoch": 2.5996959419950882,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 11114
    },
    {
      "epoch": 2.5999298327680975,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 11115
    },
    {
      "epoch": 2.6001637235411064,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 11116
    },
    {
      "epoch": 2.600397614314115,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6658,
      "step": 11117
    },
    {
      "epoch": 2.600631505087124,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 11118
    },
    {
      "epoch": 2.6008653958601333,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7463,
      "step": 11119
    },
    {
      "epoch": 2.601099286633142,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2647,
      "step": 11120
    },
    {
      "epoch": 2.6013331774061514,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.5107,
      "step": 11121
    },
    {
      "epoch": 2.6015670681791603,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 11122
    },
    {
      "epoch": 2.601800958952169,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 11123
    },
    {
      "epoch": 2.6020348497251784,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 11124
    },
    {
      "epoch": 2.6022687404981872,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 11125
    },
    {
      "epoch": 2.6025026312711965,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 11126
    },
    {
      "epoch": 2.6027365220442054,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 11127
    },
    {
      "epoch": 2.602970412817214,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 11128
    },
    {
      "epoch": 2.6032043035902235,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6921,
      "step": 11129
    },
    {
      "epoch": 2.6034381943632323,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 11130
    },
    {
      "epoch": 2.6036720851362416,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 11131
    },
    {
      "epoch": 2.6039059759092504,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 11132
    },
    {
      "epoch": 2.6041398666822593,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.663,
      "step": 11133
    },
    {
      "epoch": 2.604373757455268,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7987,
      "step": 11134
    },
    {
      "epoch": 2.6046076482282774,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 11135
    },
    {
      "epoch": 2.6048415390012867,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 11136
    },
    {
      "epoch": 2.6050754297742955,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 11137
    },
    {
      "epoch": 2.6053093205473044,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 11138
    },
    {
      "epoch": 2.605543211320313,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 11139
    },
    {
      "epoch": 2.6057771020933225,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 11140
    },
    {
      "epoch": 2.6060109928663313,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3151,
      "step": 11141
    },
    {
      "epoch": 2.6062448836393406,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0799,
      "step": 11142
    },
    {
      "epoch": 2.6064787744123494,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 11143
    },
    {
      "epoch": 2.6067126651853583,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6095,
      "step": 11144
    },
    {
      "epoch": 2.6069465559583676,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.5748,
      "step": 11145
    },
    {
      "epoch": 2.6071804467313764,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 11146
    },
    {
      "epoch": 2.6074143375043857,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 11147
    },
    {
      "epoch": 2.6076482282773945,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 11148
    },
    {
      "epoch": 2.6078821190504033,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 11149
    },
    {
      "epoch": 2.6081160098234126,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 11150
    },
    {
      "epoch": 2.6083499005964215,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 11151
    },
    {
      "epoch": 2.6085837913694307,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 11152
    },
    {
      "epoch": 2.6088176821424396,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7838,
      "step": 11153
    },
    {
      "epoch": 2.6090515729154484,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 11154
    },
    {
      "epoch": 2.6092854636884573,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.71,
      "step": 11155
    },
    {
      "epoch": 2.6095193544614665,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 11156
    },
    {
      "epoch": 2.6097532452344754,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 11157
    },
    {
      "epoch": 2.6099871360074847,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 11158
    },
    {
      "epoch": 2.6102210267804935,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 11159
    },
    {
      "epoch": 2.6104549175535023,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 11160
    },
    {
      "epoch": 2.6106888083265116,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 11161
    },
    {
      "epoch": 2.6109226990995205,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 11162
    },
    {
      "epoch": 2.6111565898725297,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 11163
    },
    {
      "epoch": 2.6113904806455386,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 11164
    },
    {
      "epoch": 2.6116243714185474,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 11165
    },
    {
      "epoch": 2.6118582621915567,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 11166
    },
    {
      "epoch": 2.6120921529645655,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 11167
    },
    {
      "epoch": 2.612326043737575,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 11168
    },
    {
      "epoch": 2.6125599345105837,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5867,
      "step": 11169
    },
    {
      "epoch": 2.6127938252835925,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7524,
      "step": 11170
    },
    {
      "epoch": 2.6130277160566013,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 11171
    },
    {
      "epoch": 2.6132616068296106,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 11172
    },
    {
      "epoch": 2.6134954976026195,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 11173
    },
    {
      "epoch": 2.6137293883756287,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 11174
    },
    {
      "epoch": 2.6139632791486376,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 11175
    },
    {
      "epoch": 2.6141971699216464,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 11176
    },
    {
      "epoch": 2.6144310606946557,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 11177
    },
    {
      "epoch": 2.6146649514676645,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 11178
    },
    {
      "epoch": 2.614898842240674,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 11179
    },
    {
      "epoch": 2.6151327330136827,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7204,
      "step": 11180
    },
    {
      "epoch": 2.6153666237866915,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 11181
    },
    {
      "epoch": 2.6156005145597008,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.68,
      "step": 11182
    },
    {
      "epoch": 2.6158344053327096,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 11183
    },
    {
      "epoch": 2.616068296105719,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6894,
      "step": 11184
    },
    {
      "epoch": 2.6163021868787277,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 11185
    },
    {
      "epoch": 2.6165360776517366,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 11186
    },
    {
      "epoch": 2.6167699684247454,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 11187
    },
    {
      "epoch": 2.6170038591977547,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 11188
    },
    {
      "epoch": 2.6172377499707635,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 11189
    },
    {
      "epoch": 2.617471640743773,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 11190
    },
    {
      "epoch": 2.6177055315167816,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.793,
      "step": 11191
    },
    {
      "epoch": 2.6179394222897905,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6642,
      "step": 11192
    },
    {
      "epoch": 2.6181733130627998,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 11193
    },
    {
      "epoch": 2.6184072038358086,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 11194
    },
    {
      "epoch": 2.618641094608818,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 11195
    },
    {
      "epoch": 2.6188749853818267,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5669,
      "step": 11196
    },
    {
      "epoch": 2.6191088761548356,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 11197
    },
    {
      "epoch": 2.619342766927845,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 11198
    },
    {
      "epoch": 2.6195766577008537,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 11199
    },
    {
      "epoch": 2.619810548473863,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 11200
    },
    {
      "epoch": 2.619810548473863,
      "eval_runtime": 4.6371,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 11200
    },
    {
      "epoch": 2.620044439246872,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 11201
    },
    {
      "epoch": 2.6202783300198806,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 11202
    },
    {
      "epoch": 2.6205122207928895,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 11203
    },
    {
      "epoch": 2.6207461115658988,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 11204
    },
    {
      "epoch": 2.6209800023389076,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 11205
    },
    {
      "epoch": 2.621213893111917,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6099,
      "step": 11206
    },
    {
      "epoch": 2.6214477838849257,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 11207
    },
    {
      "epoch": 2.6216816746579346,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 11208
    },
    {
      "epoch": 2.621915565430944,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 11209
    },
    {
      "epoch": 2.6221494562039527,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 11210
    },
    {
      "epoch": 2.622383346976962,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 11211
    },
    {
      "epoch": 2.622617237749971,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 11212
    },
    {
      "epoch": 2.6228511285229796,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6587,
      "step": 11213
    },
    {
      "epoch": 2.623085019295989,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2616,
      "step": 11214
    },
    {
      "epoch": 2.6233189100689978,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 11215
    },
    {
      "epoch": 2.623552800842007,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 11216
    },
    {
      "epoch": 2.623786691615016,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 11217
    },
    {
      "epoch": 2.6240205823880247,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 11218
    },
    {
      "epoch": 2.6242544731610336,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 11219
    },
    {
      "epoch": 2.624488363934043,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 11220
    },
    {
      "epoch": 2.6247222547070517,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 11221
    },
    {
      "epoch": 2.624956145480061,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 11222
    },
    {
      "epoch": 2.62519003625307,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 11223
    },
    {
      "epoch": 2.6254239270260786,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 11224
    },
    {
      "epoch": 2.625657817799088,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 11225
    },
    {
      "epoch": 2.6258917085720967,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 11226
    },
    {
      "epoch": 2.626125599345106,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 11227
    },
    {
      "epoch": 2.626359490118115,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 11228
    },
    {
      "epoch": 2.6265933808911237,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 11229
    },
    {
      "epoch": 2.626827271664133,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8305,
      "step": 11230
    },
    {
      "epoch": 2.627061162437142,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 11231
    },
    {
      "epoch": 2.627295053210151,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5107,
      "step": 11232
    },
    {
      "epoch": 2.62752894398316,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 11233
    },
    {
      "epoch": 2.627762834756169,
      "grad_norm": 13.25,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 11234
    },
    {
      "epoch": 2.6279967255291776,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 11235
    },
    {
      "epoch": 2.628230616302187,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 11236
    },
    {
      "epoch": 2.6284645070751957,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6833,
      "step": 11237
    },
    {
      "epoch": 2.628698397848205,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 11238
    },
    {
      "epoch": 2.628932288621214,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 11239
    },
    {
      "epoch": 2.6291661793942227,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 11240
    },
    {
      "epoch": 2.629400070167232,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 11241
    },
    {
      "epoch": 2.629633960940241,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6315,
      "step": 11242
    },
    {
      "epoch": 2.62986785171325,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 11243
    },
    {
      "epoch": 2.630101742486259,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7188,
      "step": 11244
    },
    {
      "epoch": 2.630335633259268,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 11245
    },
    {
      "epoch": 2.630569524032277,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 11246
    },
    {
      "epoch": 2.630803414805286,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6856,
      "step": 11247
    },
    {
      "epoch": 2.631037305578295,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7162,
      "step": 11248
    },
    {
      "epoch": 2.631271196351304,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 11249
    },
    {
      "epoch": 2.631505087124313,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 11250
    },
    {
      "epoch": 2.6317389778973217,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 11251
    },
    {
      "epoch": 2.631972868670331,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 11252
    },
    {
      "epoch": 2.63220675944334,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 11253
    },
    {
      "epoch": 2.632440650216349,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4245,
      "step": 11254
    },
    {
      "epoch": 2.632674540989358,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 11255
    },
    {
      "epoch": 2.6329084317623668,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 11256
    },
    {
      "epoch": 2.633142322535376,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5329,
      "step": 11257
    },
    {
      "epoch": 2.633376213308385,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 11258
    },
    {
      "epoch": 2.633610104081394,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 11259
    },
    {
      "epoch": 2.633843994854403,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 11260
    },
    {
      "epoch": 2.634077885627412,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 11261
    },
    {
      "epoch": 2.634311776400421,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 11262
    },
    {
      "epoch": 2.63454566717343,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 11263
    },
    {
      "epoch": 2.6347795579464393,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 11264
    },
    {
      "epoch": 2.635013448719448,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 11265
    },
    {
      "epoch": 2.635247339492457,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 11266
    },
    {
      "epoch": 2.6354812302654658,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 11267
    },
    {
      "epoch": 2.635715121038475,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 11268
    },
    {
      "epoch": 2.635949011811484,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 11269
    },
    {
      "epoch": 2.636182902584493,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 11270
    },
    {
      "epoch": 2.636416793357502,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0649,
      "step": 11271
    },
    {
      "epoch": 2.636650684130511,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 11272
    },
    {
      "epoch": 2.63688457490352,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 11273
    },
    {
      "epoch": 2.637118465676529,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 11274
    },
    {
      "epoch": 2.6373523564495382,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 11275
    },
    {
      "epoch": 2.637586247222547,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 11276
    },
    {
      "epoch": 2.637820137995556,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 11277
    },
    {
      "epoch": 2.638054028768565,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 11278
    },
    {
      "epoch": 2.638287919541574,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 11279
    },
    {
      "epoch": 2.6385218103145833,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7088,
      "step": 11280
    },
    {
      "epoch": 2.638755701087592,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 11281
    },
    {
      "epoch": 2.638989591860601,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 11282
    },
    {
      "epoch": 2.63922348263361,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6644,
      "step": 11283
    },
    {
      "epoch": 2.639457373406619,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 11284
    },
    {
      "epoch": 2.639691264179628,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 11285
    },
    {
      "epoch": 2.6399251549526372,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 11286
    },
    {
      "epoch": 2.640159045725646,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 11287
    },
    {
      "epoch": 2.640392936498655,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 11288
    },
    {
      "epoch": 2.640626827271664,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 11289
    },
    {
      "epoch": 2.640860718044673,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 11290
    },
    {
      "epoch": 2.6410946088176823,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 11291
    },
    {
      "epoch": 2.641328499590691,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 11292
    },
    {
      "epoch": 2.6415623903637,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 11293
    },
    {
      "epoch": 2.6417962811367093,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 11294
    },
    {
      "epoch": 2.642030171909718,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 11295
    },
    {
      "epoch": 2.6422640626827274,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 11296
    },
    {
      "epoch": 2.6424979534557362,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 11297
    },
    {
      "epoch": 2.642731844228745,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2302,
      "step": 11298
    },
    {
      "epoch": 2.642965735001754,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 11299
    },
    {
      "epoch": 2.643199625774763,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 11300
    },
    {
      "epoch": 2.643199625774763,
      "eval_runtime": 4.6611,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 11300
    },
    {
      "epoch": 2.643433516547772,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 11301
    },
    {
      "epoch": 2.6436674073207813,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4429,
      "step": 11302
    },
    {
      "epoch": 2.64390129809379,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 11303
    },
    {
      "epoch": 2.644135188866799,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3921,
      "step": 11304
    },
    {
      "epoch": 2.6443690796398083,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 11305
    },
    {
      "epoch": 2.644602970412817,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 11306
    },
    {
      "epoch": 2.6448368611858264,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 11307
    },
    {
      "epoch": 2.6450707519588352,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 11308
    },
    {
      "epoch": 2.645304642731844,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 11309
    },
    {
      "epoch": 2.6455385335048534,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 11310
    },
    {
      "epoch": 2.645772424277862,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7872,
      "step": 11311
    },
    {
      "epoch": 2.6460063150508715,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 11312
    },
    {
      "epoch": 2.6462402058238803,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 11313
    },
    {
      "epoch": 2.646474096596889,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 11314
    },
    {
      "epoch": 2.646707987369898,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 11315
    },
    {
      "epoch": 2.6469418781429073,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 11316
    },
    {
      "epoch": 2.647175768915916,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 11317
    },
    {
      "epoch": 2.6474096596889254,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 11318
    },
    {
      "epoch": 2.6476435504619342,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 11319
    },
    {
      "epoch": 2.647877441234943,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 11320
    },
    {
      "epoch": 2.6481113320079523,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 11321
    },
    {
      "epoch": 2.648345222780961,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 11322
    },
    {
      "epoch": 2.6485791135539705,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.196,
      "step": 11323
    },
    {
      "epoch": 2.6488130043269793,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 11324
    },
    {
      "epoch": 2.649046895099988,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 11325
    },
    {
      "epoch": 2.6492807858729974,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 11326
    },
    {
      "epoch": 2.6495146766460063,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 11327
    },
    {
      "epoch": 2.6497485674190155,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 11328
    },
    {
      "epoch": 2.6499824581920244,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 11329
    },
    {
      "epoch": 2.650216348965033,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 11330
    },
    {
      "epoch": 2.650450239738042,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 11331
    },
    {
      "epoch": 2.6506841305110513,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 11332
    },
    {
      "epoch": 2.65091802128406,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 11333
    },
    {
      "epoch": 2.6511519120570695,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 11334
    },
    {
      "epoch": 2.6513858028300783,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7156,
      "step": 11335
    },
    {
      "epoch": 2.651619693603087,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9255,
      "step": 11336
    },
    {
      "epoch": 2.6518535843760964,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9001,
      "step": 11337
    },
    {
      "epoch": 2.6520874751491053,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5364,
      "step": 11338
    },
    {
      "epoch": 2.6523213659221145,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 11339
    },
    {
      "epoch": 2.6525552566951234,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 11340
    },
    {
      "epoch": 2.652789147468132,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.901,
      "step": 11341
    },
    {
      "epoch": 2.6530230382411415,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 11342
    },
    {
      "epoch": 2.6532569290141503,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 11343
    },
    {
      "epoch": 2.6534908197871596,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 11344
    },
    {
      "epoch": 2.6537247105601685,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 11345
    },
    {
      "epoch": 2.6539586013331773,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 11346
    },
    {
      "epoch": 2.6541924921061866,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 11347
    },
    {
      "epoch": 2.6544263828791954,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 11348
    },
    {
      "epoch": 2.6546602736522047,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7045,
      "step": 11349
    },
    {
      "epoch": 2.6548941644252135,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6626,
      "step": 11350
    },
    {
      "epoch": 2.6551280551982224,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 11351
    },
    {
      "epoch": 2.655361945971231,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 11352
    },
    {
      "epoch": 2.6555958367442405,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 11353
    },
    {
      "epoch": 2.6558297275172493,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 11354
    },
    {
      "epoch": 2.6560636182902586,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7794,
      "step": 11355
    },
    {
      "epoch": 2.6562975090632674,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 11356
    },
    {
      "epoch": 2.6565313998362763,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 11357
    },
    {
      "epoch": 2.6567652906092856,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 11358
    },
    {
      "epoch": 2.6569991813822944,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5354,
      "step": 11359
    },
    {
      "epoch": 2.6572330721553037,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 11360
    },
    {
      "epoch": 2.6574669629283125,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 11361
    },
    {
      "epoch": 2.6577008537013214,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 11362
    },
    {
      "epoch": 2.6579347444743306,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 11363
    },
    {
      "epoch": 2.6581686352473395,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 11364
    },
    {
      "epoch": 2.6584025260203488,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1601,
      "step": 11365
    },
    {
      "epoch": 2.6586364167933576,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4664,
      "step": 11366
    },
    {
      "epoch": 2.6588703075663664,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 11367
    },
    {
      "epoch": 2.6591041983393753,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8347,
      "step": 11368
    },
    {
      "epoch": 2.6593380891123846,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9751,
      "step": 11369
    },
    {
      "epoch": 2.6595719798853934,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 11370
    },
    {
      "epoch": 2.6598058706584027,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 11371
    },
    {
      "epoch": 2.6600397614314115,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 11372
    },
    {
      "epoch": 2.6602736522044204,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 11373
    },
    {
      "epoch": 2.6605075429774296,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7454,
      "step": 11374
    },
    {
      "epoch": 2.6607414337504385,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 11375
    },
    {
      "epoch": 2.6609753245234478,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2136,
      "step": 11376
    },
    {
      "epoch": 2.6612092152964566,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 11377
    },
    {
      "epoch": 2.6614431060694654,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5272,
      "step": 11378
    },
    {
      "epoch": 2.6616769968424747,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7007,
      "step": 11379
    },
    {
      "epoch": 2.6619108876154836,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 11380
    },
    {
      "epoch": 2.662144778388493,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 11381
    },
    {
      "epoch": 2.6623786691615017,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 11382
    },
    {
      "epoch": 2.6626125599345105,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 11383
    },
    {
      "epoch": 2.6628464507075194,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 11384
    },
    {
      "epoch": 2.6630803414805286,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 11385
    },
    {
      "epoch": 2.6633142322535375,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 11386
    },
    {
      "epoch": 2.6635481230265468,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 11387
    },
    {
      "epoch": 2.6637820137995556,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 11388
    },
    {
      "epoch": 2.6640159045725644,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9759,
      "step": 11389
    },
    {
      "epoch": 2.6642497953455737,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 11390
    },
    {
      "epoch": 2.6644836861185826,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 11391
    },
    {
      "epoch": 2.664717576891592,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 11392
    },
    {
      "epoch": 2.6649514676646007,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 11393
    },
    {
      "epoch": 2.6651853584376095,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 11394
    },
    {
      "epoch": 2.665419249210619,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8524,
      "step": 11395
    },
    {
      "epoch": 2.6656531399836276,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.701,
      "step": 11396
    },
    {
      "epoch": 2.665887030756637,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.193,
      "step": 11397
    },
    {
      "epoch": 2.6661209215296457,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 11398
    },
    {
      "epoch": 2.6663548123026546,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 11399
    },
    {
      "epoch": 2.6665887030756634,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3596,
      "step": 11400
    },
    {
      "epoch": 2.6665887030756634,
      "eval_runtime": 4.5754,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 11400
    },
    {
      "epoch": 2.6668225938486727,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 11401
    },
    {
      "epoch": 2.6670564846216815,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 11402
    },
    {
      "epoch": 2.667290375394691,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 11403
    },
    {
      "epoch": 2.6675242661676997,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 11404
    },
    {
      "epoch": 2.6677581569407085,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 11405
    },
    {
      "epoch": 2.667992047713718,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 11406
    },
    {
      "epoch": 2.6682259384867266,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2697,
      "step": 11407
    },
    {
      "epoch": 2.668459829259736,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 11408
    },
    {
      "epoch": 2.6686937200327447,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7625,
      "step": 11409
    },
    {
      "epoch": 2.6689276108057536,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 11410
    },
    {
      "epoch": 2.669161501578763,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 11411
    },
    {
      "epoch": 2.6693953923517717,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 11412
    },
    {
      "epoch": 2.669629283124781,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 11413
    },
    {
      "epoch": 2.66986317389779,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 11414
    },
    {
      "epoch": 2.6700970646707987,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 11415
    },
    {
      "epoch": 2.6703309554438075,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 11416
    },
    {
      "epoch": 2.670564846216817,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 11417
    },
    {
      "epoch": 2.6707987369898256,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7411,
      "step": 11418
    },
    {
      "epoch": 2.671032627762835,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 11419
    },
    {
      "epoch": 2.6712665185358437,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8711,
      "step": 11420
    },
    {
      "epoch": 2.6715004093088526,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 11421
    },
    {
      "epoch": 2.671734300081862,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6327,
      "step": 11422
    },
    {
      "epoch": 2.6719681908548707,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7625,
      "step": 11423
    },
    {
      "epoch": 2.67220208162788,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 11424
    },
    {
      "epoch": 2.672435972400889,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 11425
    },
    {
      "epoch": 2.6726698631738977,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 11426
    },
    {
      "epoch": 2.672903753946907,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 11427
    },
    {
      "epoch": 2.6731376447199158,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 11428
    },
    {
      "epoch": 2.673371535492925,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 11429
    },
    {
      "epoch": 2.673605426265934,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 11430
    },
    {
      "epoch": 2.6738393170389427,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 11431
    },
    {
      "epoch": 2.6740732078119516,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 11432
    },
    {
      "epoch": 2.674307098584961,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 11433
    },
    {
      "epoch": 2.6745409893579697,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 11434
    },
    {
      "epoch": 2.674774880130979,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7954,
      "step": 11435
    },
    {
      "epoch": 2.675008770903988,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 11436
    },
    {
      "epoch": 2.6752426616769966,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 11437
    },
    {
      "epoch": 2.675476552450006,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 11438
    },
    {
      "epoch": 2.6757104432230148,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4711,
      "step": 11439
    },
    {
      "epoch": 2.675944333996024,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 11440
    },
    {
      "epoch": 2.676178224769033,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 11441
    },
    {
      "epoch": 2.6764121155420417,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 11442
    },
    {
      "epoch": 2.676646006315051,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 11443
    },
    {
      "epoch": 2.67687989708806,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.4982,
      "step": 11444
    },
    {
      "epoch": 2.677113787861069,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 11445
    },
    {
      "epoch": 2.677347678634078,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 11446
    },
    {
      "epoch": 2.677581569407087,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.59,
      "step": 11447
    },
    {
      "epoch": 2.6778154601800956,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5637,
      "step": 11448
    },
    {
      "epoch": 2.678049350953105,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 11449
    },
    {
      "epoch": 2.6782832417261138,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 11450
    },
    {
      "epoch": 2.678517132499123,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6784,
      "step": 11451
    },
    {
      "epoch": 2.678751023272132,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 11452
    },
    {
      "epoch": 2.6789849140451407,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 11453
    },
    {
      "epoch": 2.67921880481815,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6847,
      "step": 11454
    },
    {
      "epoch": 2.679452695591159,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6547,
      "step": 11455
    },
    {
      "epoch": 2.679686586364168,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 11456
    },
    {
      "epoch": 2.679920477137177,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 11457
    },
    {
      "epoch": 2.680154367910186,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 11458
    },
    {
      "epoch": 2.680388258683195,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7876,
      "step": 11459
    },
    {
      "epoch": 2.680622149456204,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 11460
    },
    {
      "epoch": 2.680856040229213,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2907,
      "step": 11461
    },
    {
      "epoch": 2.681089931002222,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 11462
    },
    {
      "epoch": 2.681323821775231,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 11463
    },
    {
      "epoch": 2.6815577125482397,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 11464
    },
    {
      "epoch": 2.681791603321249,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 11465
    },
    {
      "epoch": 2.682025494094258,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 11466
    },
    {
      "epoch": 2.682259384867267,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 11467
    },
    {
      "epoch": 2.682493275640276,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 11468
    },
    {
      "epoch": 2.682727166413285,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6459,
      "step": 11469
    },
    {
      "epoch": 2.682961057186294,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 11470
    },
    {
      "epoch": 2.683194947959303,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 11471
    },
    {
      "epoch": 2.683428838732312,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 11472
    },
    {
      "epoch": 2.683662729505321,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 11473
    },
    {
      "epoch": 2.68389662027833,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 11474
    },
    {
      "epoch": 2.684130511051339,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 11475
    },
    {
      "epoch": 2.684364401824348,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.454,
      "step": 11476
    },
    {
      "epoch": 2.6845982925973573,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 11477
    },
    {
      "epoch": 2.684832183370366,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 11478
    },
    {
      "epoch": 2.685066074143375,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 11479
    },
    {
      "epoch": 2.685299964916384,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 11480
    },
    {
      "epoch": 2.685533855689393,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 11481
    },
    {
      "epoch": 2.685767746462402,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8361,
      "step": 11482
    },
    {
      "epoch": 2.686001637235411,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 11483
    },
    {
      "epoch": 2.68623552800842,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 11484
    },
    {
      "epoch": 2.686469418781429,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 11485
    },
    {
      "epoch": 2.686703309554438,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 11486
    },
    {
      "epoch": 2.686937200327447,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 11487
    },
    {
      "epoch": 2.6871710911004563,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 11488
    },
    {
      "epoch": 2.687404981873465,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 11489
    },
    {
      "epoch": 2.687638872646474,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7552,
      "step": 11490
    },
    {
      "epoch": 2.6878727634194832,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6037,
      "step": 11491
    },
    {
      "epoch": 2.688106654192492,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 11492
    },
    {
      "epoch": 2.6883405449655013,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 11493
    },
    {
      "epoch": 2.68857443573851,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 11494
    },
    {
      "epoch": 2.688808326511519,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 11495
    },
    {
      "epoch": 2.689042217284528,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9196,
      "step": 11496
    },
    {
      "epoch": 2.689276108057537,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6819,
      "step": 11497
    },
    {
      "epoch": 2.689509998830546,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 11498
    },
    {
      "epoch": 2.6897438896035553,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 11499
    },
    {
      "epoch": 2.689977780376564,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 11500
    },
    {
      "epoch": 2.689977780376564,
      "eval_runtime": 4.6308,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 11500
    },
    {
      "epoch": 2.690211671149573,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6495,
      "step": 11501
    },
    {
      "epoch": 2.690445561922582,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 11502
    },
    {
      "epoch": 2.690679452695591,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 11503
    },
    {
      "epoch": 2.6909133434686003,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 11504
    },
    {
      "epoch": 2.691147234241609,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 11505
    },
    {
      "epoch": 2.691381125014618,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7154,
      "step": 11506
    },
    {
      "epoch": 2.6916150157876273,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 11507
    },
    {
      "epoch": 2.691848906560636,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 11508
    },
    {
      "epoch": 2.6920827973336454,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 11509
    },
    {
      "epoch": 2.6923166881066543,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6573,
      "step": 11510
    },
    {
      "epoch": 2.692550578879663,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 11511
    },
    {
      "epoch": 2.692784469652672,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 11512
    },
    {
      "epoch": 2.693018360425681,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 11513
    },
    {
      "epoch": 2.69325225119869,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 11514
    },
    {
      "epoch": 2.6934861419716993,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8503,
      "step": 11515
    },
    {
      "epoch": 2.693720032744708,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 11516
    },
    {
      "epoch": 2.693953923517717,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 11517
    },
    {
      "epoch": 2.6941878142907263,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 11518
    },
    {
      "epoch": 2.694421705063735,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0122,
      "step": 11519
    },
    {
      "epoch": 2.6946555958367444,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 11520
    },
    {
      "epoch": 2.6948894866097532,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7651,
      "step": 11521
    },
    {
      "epoch": 2.695123377382762,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 11522
    },
    {
      "epoch": 2.6953572681557714,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 11523
    },
    {
      "epoch": 2.69559115892878,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 11524
    },
    {
      "epoch": 2.6958250497017895,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 11525
    },
    {
      "epoch": 2.6960589404747983,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 11526
    },
    {
      "epoch": 2.696292831247807,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6909,
      "step": 11527
    },
    {
      "epoch": 2.696526722020816,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 11528
    },
    {
      "epoch": 2.6967606127938253,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7224,
      "step": 11529
    },
    {
      "epoch": 2.696994503566834,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 11530
    },
    {
      "epoch": 2.6972283943398434,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 11531
    },
    {
      "epoch": 2.6974622851128522,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 11532
    },
    {
      "epoch": 2.697696175885861,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6235,
      "step": 11533
    },
    {
      "epoch": 2.6979300666588704,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7732,
      "step": 11534
    },
    {
      "epoch": 2.698163957431879,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 11535
    },
    {
      "epoch": 2.6983978482048885,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 11536
    },
    {
      "epoch": 2.6986317389778973,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 11537
    },
    {
      "epoch": 2.698865629750906,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 11538
    },
    {
      "epoch": 2.6990995205239154,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 11539
    },
    {
      "epoch": 2.6993334112969243,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 11540
    },
    {
      "epoch": 2.6995673020699336,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 11541
    },
    {
      "epoch": 2.6998011928429424,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 11542
    },
    {
      "epoch": 2.7000350836159512,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 11543
    },
    {
      "epoch": 2.7002689743889605,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 11544
    },
    {
      "epoch": 2.7005028651619694,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 11545
    },
    {
      "epoch": 2.7007367559349786,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 11546
    },
    {
      "epoch": 2.7009706467079875,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 11547
    },
    {
      "epoch": 2.7012045374809963,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 11548
    },
    {
      "epoch": 2.701438428254005,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 11549
    },
    {
      "epoch": 2.7016723190270144,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 11550
    },
    {
      "epoch": 2.7019062098000233,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6195,
      "step": 11551
    },
    {
      "epoch": 2.7021401005730326,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 11552
    },
    {
      "epoch": 2.7023739913460414,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 11553
    },
    {
      "epoch": 2.7026078821190502,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 11554
    },
    {
      "epoch": 2.7028417728920595,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 11555
    },
    {
      "epoch": 2.7030756636650684,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 11556
    },
    {
      "epoch": 2.7033095544380776,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 11557
    },
    {
      "epoch": 2.7035434452110865,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 11558
    },
    {
      "epoch": 2.7037773359840953,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 11559
    },
    {
      "epoch": 2.7040112267571046,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5163,
      "step": 11560
    },
    {
      "epoch": 2.7042451175301134,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 11561
    },
    {
      "epoch": 2.7044790083031227,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 11562
    },
    {
      "epoch": 2.7047128990761315,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 11563
    },
    {
      "epoch": 2.7049467898491404,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 11564
    },
    {
      "epoch": 2.7051806806221492,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 11565
    },
    {
      "epoch": 2.7054145713951585,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 11566
    },
    {
      "epoch": 2.7056484621681673,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 11567
    },
    {
      "epoch": 2.7058823529411766,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 11568
    },
    {
      "epoch": 2.7061162437141855,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2165,
      "step": 11569
    },
    {
      "epoch": 2.7063501344871943,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 11570
    },
    {
      "epoch": 2.7065840252602036,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 11571
    },
    {
      "epoch": 2.7068179160332124,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 11572
    },
    {
      "epoch": 2.7070518068062217,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 11573
    },
    {
      "epoch": 2.7072856975792305,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 11574
    },
    {
      "epoch": 2.7075195883522394,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 11575
    },
    {
      "epoch": 2.7077534791252487,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.177,
      "step": 11576
    },
    {
      "epoch": 2.7079873698982575,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 11577
    },
    {
      "epoch": 2.708221260671267,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 11578
    },
    {
      "epoch": 2.7084551514442756,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 11579
    },
    {
      "epoch": 2.7086890422172845,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 11580
    },
    {
      "epoch": 2.7089229329902933,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2205,
      "step": 11581
    },
    {
      "epoch": 2.7091568237633026,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 11582
    },
    {
      "epoch": 2.7093907145363114,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 11583
    },
    {
      "epoch": 2.7096246053093207,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8608,
      "step": 11584
    },
    {
      "epoch": 2.7098584960823295,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 11585
    },
    {
      "epoch": 2.7100923868553384,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 11586
    },
    {
      "epoch": 2.7103262776283477,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 11587
    },
    {
      "epoch": 2.7105601684013565,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 11588
    },
    {
      "epoch": 2.7107940591743658,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 11589
    },
    {
      "epoch": 2.7110279499473746,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 11590
    },
    {
      "epoch": 2.7112618407203835,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 11591
    },
    {
      "epoch": 2.7114957314933927,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6219,
      "step": 11592
    },
    {
      "epoch": 2.7117296222664016,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 11593
    },
    {
      "epoch": 2.711963513039411,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 11594
    },
    {
      "epoch": 2.7121974038124197,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 11595
    },
    {
      "epoch": 2.7124312945854285,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 11596
    },
    {
      "epoch": 2.7126651853584374,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 11597
    },
    {
      "epoch": 2.7128990761314467,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 11598
    },
    {
      "epoch": 2.7131329669044555,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 11599
    },
    {
      "epoch": 2.7133668576774648,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 11600
    },
    {
      "epoch": 2.7133668576774648,
      "eval_runtime": 4.6665,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 11600
    },
    {
      "epoch": 2.7136007484504736,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 11601
    },
    {
      "epoch": 2.7138346392234824,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.037,
      "step": 11602
    },
    {
      "epoch": 2.7140685299964917,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 11603
    },
    {
      "epoch": 2.7143024207695006,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 11604
    },
    {
      "epoch": 2.71453631154251,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 11605
    },
    {
      "epoch": 2.7147702023155187,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 11606
    },
    {
      "epoch": 2.7150040930885275,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 11607
    },
    {
      "epoch": 2.715237983861537,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 11608
    },
    {
      "epoch": 2.7154718746345456,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 11609
    },
    {
      "epoch": 2.715705765407555,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 11610
    },
    {
      "epoch": 2.7159396561805638,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 11611
    },
    {
      "epoch": 2.7161735469535726,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 11612
    },
    {
      "epoch": 2.7164074377265814,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.9255,
      "step": 11613
    },
    {
      "epoch": 2.7166413284995907,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 11614
    },
    {
      "epoch": 2.7168752192725996,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 11615
    },
    {
      "epoch": 2.717109110045609,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6347,
      "step": 11616
    },
    {
      "epoch": 2.7173430008186177,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 11617
    },
    {
      "epoch": 2.7175768915916265,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 11618
    },
    {
      "epoch": 2.717810782364636,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 11619
    },
    {
      "epoch": 2.7180446731376446,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6892,
      "step": 11620
    },
    {
      "epoch": 2.718278563910654,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 11621
    },
    {
      "epoch": 2.7185124546836628,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 11622
    },
    {
      "epoch": 2.7187463454566716,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 11623
    },
    {
      "epoch": 2.718980236229681,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 11624
    },
    {
      "epoch": 2.7192141270026897,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 11625
    },
    {
      "epoch": 2.719448017775699,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7209,
      "step": 11626
    },
    {
      "epoch": 2.719681908548708,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6914,
      "step": 11627
    },
    {
      "epoch": 2.7199157993217167,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 11628
    },
    {
      "epoch": 2.7201496900947255,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 11629
    },
    {
      "epoch": 2.720383580867735,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3126,
      "step": 11630
    },
    {
      "epoch": 2.7206174716407436,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 11631
    },
    {
      "epoch": 2.720851362413753,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 11632
    },
    {
      "epoch": 2.7210852531867618,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 11633
    },
    {
      "epoch": 2.7213191439597706,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 11634
    },
    {
      "epoch": 2.72155303473278,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 11635
    },
    {
      "epoch": 2.7217869255057887,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 11636
    },
    {
      "epoch": 2.722020816278798,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6996,
      "step": 11637
    },
    {
      "epoch": 2.722254707051807,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 11638
    },
    {
      "epoch": 2.7224885978248157,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2929,
      "step": 11639
    },
    {
      "epoch": 2.722722488597825,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5999,
      "step": 11640
    },
    {
      "epoch": 2.722956379370834,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 11641
    },
    {
      "epoch": 2.723190270143843,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 11642
    },
    {
      "epoch": 2.723424160916852,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 11643
    },
    {
      "epoch": 2.7236580516898607,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7061,
      "step": 11644
    },
    {
      "epoch": 2.7238919424628696,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 11645
    },
    {
      "epoch": 2.724125833235879,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 11646
    },
    {
      "epoch": 2.7243597240088877,
      "grad_norm": 10.375,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 11647
    },
    {
      "epoch": 2.724593614781897,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8234,
      "step": 11648
    },
    {
      "epoch": 2.724827505554906,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2778,
      "step": 11649
    },
    {
      "epoch": 2.7250613963279147,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 11650
    },
    {
      "epoch": 2.725295287100924,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 11651
    },
    {
      "epoch": 2.725529177873933,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 11652
    },
    {
      "epoch": 2.725763068646942,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 11653
    },
    {
      "epoch": 2.725996959419951,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 11654
    },
    {
      "epoch": 2.7262308501929597,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 11655
    },
    {
      "epoch": 2.726464740965969,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 11656
    },
    {
      "epoch": 2.726698631738978,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 11657
    },
    {
      "epoch": 2.726932522511987,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 11658
    },
    {
      "epoch": 2.727166413284996,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 11659
    },
    {
      "epoch": 2.727400304058005,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 11660
    },
    {
      "epoch": 2.7276341948310137,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 11661
    },
    {
      "epoch": 2.727868085604023,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 11662
    },
    {
      "epoch": 2.728101976377032,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6554,
      "step": 11663
    },
    {
      "epoch": 2.728335867150041,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 11664
    },
    {
      "epoch": 2.72856975792305,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2707,
      "step": 11665
    },
    {
      "epoch": 2.7288036486960587,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 11666
    },
    {
      "epoch": 2.729037539469068,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 11667
    },
    {
      "epoch": 2.729271430242077,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 11668
    },
    {
      "epoch": 2.729505321015086,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 11669
    },
    {
      "epoch": 2.729739211788095,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 11670
    },
    {
      "epoch": 2.729973102561104,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 11671
    },
    {
      "epoch": 2.730206993334113,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6402,
      "step": 11672
    },
    {
      "epoch": 2.730440884107122,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 11673
    },
    {
      "epoch": 2.730674774880131,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 11674
    },
    {
      "epoch": 2.73090866565314,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 11675
    },
    {
      "epoch": 2.731142556426149,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 11676
    },
    {
      "epoch": 2.7313764471991577,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 11677
    },
    {
      "epoch": 2.731610337972167,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 11678
    },
    {
      "epoch": 2.731844228745176,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 11679
    },
    {
      "epoch": 2.732078119518185,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1767,
      "step": 11680
    },
    {
      "epoch": 2.732312010291194,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 11681
    },
    {
      "epoch": 2.732545901064203,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 11682
    },
    {
      "epoch": 2.732779791837212,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 11683
    },
    {
      "epoch": 2.733013682610221,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 11684
    },
    {
      "epoch": 2.73324757338323,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6914,
      "step": 11685
    },
    {
      "epoch": 2.733481464156239,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 11686
    },
    {
      "epoch": 2.733715354929248,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 11687
    },
    {
      "epoch": 2.733949245702257,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 11688
    },
    {
      "epoch": 2.734183136475266,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 11689
    },
    {
      "epoch": 2.7344170272482753,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 11690
    },
    {
      "epoch": 2.734650918021284,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 11691
    },
    {
      "epoch": 2.734884808794293,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 11692
    },
    {
      "epoch": 2.735118699567302,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 11693
    },
    {
      "epoch": 2.735352590340311,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 11694
    },
    {
      "epoch": 2.73558648111332,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5732,
      "step": 11695
    },
    {
      "epoch": 2.735820371886329,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 11696
    },
    {
      "epoch": 2.736054262659338,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 11697
    },
    {
      "epoch": 2.736288153432347,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 11698
    },
    {
      "epoch": 2.736522044205356,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 11699
    },
    {
      "epoch": 2.736755934978365,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6063,
      "step": 11700
    },
    {
      "epoch": 2.736755934978365,
      "eval_runtime": 4.6741,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 11700
    },
    {
      "epoch": 2.7369898257513743,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 11701
    },
    {
      "epoch": 2.737223716524383,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 11702
    },
    {
      "epoch": 2.737457607297392,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 11703
    },
    {
      "epoch": 2.7376914980704012,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0299,
      "step": 11704
    },
    {
      "epoch": 2.73792538884341,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 11705
    },
    {
      "epoch": 2.7381592796164194,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 11706
    },
    {
      "epoch": 2.738393170389428,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 11707
    },
    {
      "epoch": 2.738627061162437,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8329,
      "step": 11708
    },
    {
      "epoch": 2.738860951935446,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 11709
    },
    {
      "epoch": 2.739094842708455,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3475,
      "step": 11710
    },
    {
      "epoch": 2.739328733481464,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 11711
    },
    {
      "epoch": 2.7395626242544733,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8707,
      "step": 11712
    },
    {
      "epoch": 2.739796515027482,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2103,
      "step": 11713
    },
    {
      "epoch": 2.740030405800491,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 11714
    },
    {
      "epoch": 2.7402642965735002,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 11715
    },
    {
      "epoch": 2.740498187346509,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 11716
    },
    {
      "epoch": 2.7407320781195184,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 11717
    },
    {
      "epoch": 2.740965968892527,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 11718
    },
    {
      "epoch": 2.741199859665536,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 11719
    },
    {
      "epoch": 2.7414337504385453,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 11720
    },
    {
      "epoch": 2.741667641211554,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 11721
    },
    {
      "epoch": 2.7419015319845634,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 11722
    },
    {
      "epoch": 2.7421354227575723,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 11723
    },
    {
      "epoch": 2.742369313530581,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 11724
    },
    {
      "epoch": 2.74260320430359,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3481,
      "step": 11725
    },
    {
      "epoch": 2.7428370950765992,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 11726
    },
    {
      "epoch": 2.743070985849608,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8366,
      "step": 11727
    },
    {
      "epoch": 2.7433048766226173,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6852,
      "step": 11728
    },
    {
      "epoch": 2.743538767395626,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 11729
    },
    {
      "epoch": 2.743772658168635,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 11730
    },
    {
      "epoch": 2.7440065489416443,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6986,
      "step": 11731
    },
    {
      "epoch": 2.744240439714653,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 11732
    },
    {
      "epoch": 2.7444743304876624,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3596,
      "step": 11733
    },
    {
      "epoch": 2.7447082212606713,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 11734
    },
    {
      "epoch": 2.74494211203368,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 11735
    },
    {
      "epoch": 2.7451760028066894,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 11736
    },
    {
      "epoch": 2.7454098935796982,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 11737
    },
    {
      "epoch": 2.7456437843527075,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 11738
    },
    {
      "epoch": 2.7458776751257163,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 11739
    },
    {
      "epoch": 2.746111565898725,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 11740
    },
    {
      "epoch": 2.7463454566717345,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 11741
    },
    {
      "epoch": 2.7465793474447433,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 11742
    },
    {
      "epoch": 2.7468132382177526,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 11743
    },
    {
      "epoch": 2.7470471289907614,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 11744
    },
    {
      "epoch": 2.7472810197637703,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 11745
    },
    {
      "epoch": 2.747514910536779,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 11746
    },
    {
      "epoch": 2.7477488013097884,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6962,
      "step": 11747
    },
    {
      "epoch": 2.747982692082797,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6763,
      "step": 11748
    },
    {
      "epoch": 2.7482165828558065,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 11749
    },
    {
      "epoch": 2.7484504736288153,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 11750
    },
    {
      "epoch": 2.748684364401824,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 11751
    },
    {
      "epoch": 2.7489182551748335,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 11752
    },
    {
      "epoch": 2.7491521459478423,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 11753
    },
    {
      "epoch": 2.7493860367208516,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 11754
    },
    {
      "epoch": 2.7496199274938604,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5512,
      "step": 11755
    },
    {
      "epoch": 2.7498538182668693,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 11756
    },
    {
      "epoch": 2.7500877090398785,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 11757
    },
    {
      "epoch": 2.7503215998128874,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 1.6904,
      "step": 11758
    },
    {
      "epoch": 2.7505554905858967,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6858,
      "step": 11759
    },
    {
      "epoch": 2.7507893813589055,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 11760
    },
    {
      "epoch": 2.7510232721319143,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 11761
    },
    {
      "epoch": 2.751257162904923,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 11762
    },
    {
      "epoch": 2.7514910536779325,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 11763
    },
    {
      "epoch": 2.7517249444509413,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6638,
      "step": 11764
    },
    {
      "epoch": 2.7519588352239506,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 11765
    },
    {
      "epoch": 2.7521927259969594,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 11766
    },
    {
      "epoch": 2.7524266167699682,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 11767
    },
    {
      "epoch": 2.7526605075429775,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 11768
    },
    {
      "epoch": 2.7528943983159864,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7267,
      "step": 11769
    },
    {
      "epoch": 2.7531282890889957,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.819,
      "step": 11770
    },
    {
      "epoch": 2.7533621798620045,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 11771
    },
    {
      "epoch": 2.7535960706350133,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 11772
    },
    {
      "epoch": 2.7538299614080226,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 11773
    },
    {
      "epoch": 2.7540638521810314,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 11774
    },
    {
      "epoch": 2.7542977429540407,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8058,
      "step": 11775
    },
    {
      "epoch": 2.7545316337270496,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 11776
    },
    {
      "epoch": 2.7547655245000584,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 11777
    },
    {
      "epoch": 2.7549994152730672,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6173,
      "step": 11778
    },
    {
      "epoch": 2.7552333060460765,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 11779
    },
    {
      "epoch": 2.7554671968190854,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 11780
    },
    {
      "epoch": 2.7557010875920946,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 11781
    },
    {
      "epoch": 2.7559349783651035,
      "grad_norm": 2.34375,
      "learning_rate": 3e-05,
      "loss": 1.542,
      "step": 11782
    },
    {
      "epoch": 2.7561688691381123,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 11783
    },
    {
      "epoch": 2.7564027599111216,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 11784
    },
    {
      "epoch": 2.7566366506841304,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4563,
      "step": 11785
    },
    {
      "epoch": 2.7568705414571397,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 11786
    },
    {
      "epoch": 2.7571044322301486,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 11787
    },
    {
      "epoch": 2.7573383230031574,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 11788
    },
    {
      "epoch": 2.7575722137761667,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 11789
    },
    {
      "epoch": 2.7578061045491755,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 11790
    },
    {
      "epoch": 2.758039995322185,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 11791
    },
    {
      "epoch": 2.7582738860951936,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 11792
    },
    {
      "epoch": 2.7585077768682025,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 11793
    },
    {
      "epoch": 2.7587416676412113,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 11794
    },
    {
      "epoch": 2.7589755584142206,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5864,
      "step": 11795
    },
    {
      "epoch": 2.7592094491872294,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 11796
    },
    {
      "epoch": 2.7594433399602387,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.274,
      "step": 11797
    },
    {
      "epoch": 2.7596772307332476,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 11798
    },
    {
      "epoch": 2.7599111215062564,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 11799
    },
    {
      "epoch": 2.7601450122792657,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 11800
    },
    {
      "epoch": 2.7601450122792657,
      "eval_runtime": 4.6262,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 11800
    },
    {
      "epoch": 2.7603789030522745,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 11801
    },
    {
      "epoch": 2.760612793825284,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 11802
    },
    {
      "epoch": 2.7608466845982926,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 11803
    },
    {
      "epoch": 2.7610805753713015,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 11804
    },
    {
      "epoch": 2.7613144661443108,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 11805
    },
    {
      "epoch": 2.7615483569173196,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 11806
    },
    {
      "epoch": 2.761782247690329,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 11807
    },
    {
      "epoch": 2.7620161384633377,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 11808
    },
    {
      "epoch": 2.7622500292363465,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 11809
    },
    {
      "epoch": 2.7624839200093554,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3069,
      "step": 11810
    },
    {
      "epoch": 2.7627178107823647,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7838,
      "step": 11811
    },
    {
      "epoch": 2.7629517015553735,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6593,
      "step": 11812
    },
    {
      "epoch": 2.763185592328383,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 11813
    },
    {
      "epoch": 2.7634194831013916,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4358,
      "step": 11814
    },
    {
      "epoch": 2.7636533738744005,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 11815
    },
    {
      "epoch": 2.7638872646474097,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 11816
    },
    {
      "epoch": 2.7641211554204186,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 11817
    },
    {
      "epoch": 2.764355046193428,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 11818
    },
    {
      "epoch": 2.7645889369664367,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 11819
    },
    {
      "epoch": 2.7648228277394455,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 11820
    },
    {
      "epoch": 2.765056718512455,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 11821
    },
    {
      "epoch": 2.7652906092854637,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 11822
    },
    {
      "epoch": 2.765524500058473,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 11823
    },
    {
      "epoch": 2.765758390831482,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 11824
    },
    {
      "epoch": 2.7659922816044906,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 11825
    },
    {
      "epoch": 2.7662261723774995,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 11826
    },
    {
      "epoch": 2.7664600631505087,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 11827
    },
    {
      "epoch": 2.7666939539235176,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 11828
    },
    {
      "epoch": 2.766927844696527,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 11829
    },
    {
      "epoch": 2.7671617354695357,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 11830
    },
    {
      "epoch": 2.7673956262425445,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 11831
    },
    {
      "epoch": 2.767629517015554,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 11832
    },
    {
      "epoch": 2.7678634077885627,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8601,
      "step": 11833
    },
    {
      "epoch": 2.768097298561572,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 11834
    },
    {
      "epoch": 2.7683311893345808,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 11835
    },
    {
      "epoch": 2.7685650801075896,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 11836
    },
    {
      "epoch": 2.768798970880599,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 11837
    },
    {
      "epoch": 2.7690328616536077,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 11838
    },
    {
      "epoch": 2.769266752426617,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6576,
      "step": 11839
    },
    {
      "epoch": 2.769500643199626,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 11840
    },
    {
      "epoch": 2.7697345339726347,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6575,
      "step": 11841
    },
    {
      "epoch": 2.7699684247456435,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 11842
    },
    {
      "epoch": 2.770202315518653,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 11843
    },
    {
      "epoch": 2.7704362062916617,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 11844
    },
    {
      "epoch": 2.770670097064671,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 11845
    },
    {
      "epoch": 2.7709039878376798,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 11846
    },
    {
      "epoch": 2.7711378786106886,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 11847
    },
    {
      "epoch": 2.771371769383698,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 11848
    },
    {
      "epoch": 2.7716056601567067,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.037,
      "step": 11849
    },
    {
      "epoch": 2.771839550929716,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 11850
    },
    {
      "epoch": 2.772073441702725,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 11851
    },
    {
      "epoch": 2.7723073324757337,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6258,
      "step": 11852
    },
    {
      "epoch": 2.772541223248743,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 11853
    },
    {
      "epoch": 2.772775114021752,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 11854
    },
    {
      "epoch": 2.773009004794761,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6779,
      "step": 11855
    },
    {
      "epoch": 2.77324289556777,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 11856
    },
    {
      "epoch": 2.7734767863407788,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 11857
    },
    {
      "epoch": 2.7737106771137876,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9756,
      "step": 11858
    },
    {
      "epoch": 2.773944567886797,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7855,
      "step": 11859
    },
    {
      "epoch": 2.7741784586598057,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 11860
    },
    {
      "epoch": 2.774412349432815,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 11861
    },
    {
      "epoch": 2.774646240205824,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5858,
      "step": 11862
    },
    {
      "epoch": 2.7748801309788327,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 11863
    },
    {
      "epoch": 2.775114021751842,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 11864
    },
    {
      "epoch": 2.775347912524851,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 11865
    },
    {
      "epoch": 2.77558180329786,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 11866
    },
    {
      "epoch": 2.775815694070869,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9569,
      "step": 11867
    },
    {
      "epoch": 2.7760495848438778,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 11868
    },
    {
      "epoch": 2.776283475616887,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 11869
    },
    {
      "epoch": 2.776517366389896,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 11870
    },
    {
      "epoch": 2.776751257162905,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 11871
    },
    {
      "epoch": 2.776985147935914,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6941,
      "step": 11872
    },
    {
      "epoch": 2.777219038708923,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 11873
    },
    {
      "epoch": 2.7774529294819317,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 11874
    },
    {
      "epoch": 2.777686820254941,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 11875
    },
    {
      "epoch": 2.77792071102795,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 11876
    },
    {
      "epoch": 2.778154601800959,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 11877
    },
    {
      "epoch": 2.778388492573968,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0472,
      "step": 11878
    },
    {
      "epoch": 2.7786223833469768,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 11879
    },
    {
      "epoch": 2.778856274119986,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 11880
    },
    {
      "epoch": 2.779090164892995,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6451,
      "step": 11881
    },
    {
      "epoch": 2.779324055666004,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6523,
      "step": 11882
    },
    {
      "epoch": 2.779557946439013,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 11883
    },
    {
      "epoch": 2.779791837212022,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 11884
    },
    {
      "epoch": 2.780025727985031,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 11885
    },
    {
      "epoch": 2.78025961875804,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8315,
      "step": 11886
    },
    {
      "epoch": 2.7804935095310492,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.4538,
      "step": 11887
    },
    {
      "epoch": 2.780727400304058,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5797,
      "step": 11888
    },
    {
      "epoch": 2.780961291077067,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 11889
    },
    {
      "epoch": 2.7811951818500757,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.4541,
      "step": 11890
    },
    {
      "epoch": 2.781429072623085,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 11891
    },
    {
      "epoch": 2.781662963396094,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 11892
    },
    {
      "epoch": 2.781896854169103,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 11893
    },
    {
      "epoch": 2.782130744942112,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 11894
    },
    {
      "epoch": 2.782364635715121,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 11895
    },
    {
      "epoch": 2.78259852648813,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 11896
    },
    {
      "epoch": 2.782832417261139,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 11897
    },
    {
      "epoch": 2.7830663080341482,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 11898
    },
    {
      "epoch": 2.783300198807157,
      "grad_norm": 15.75,
      "learning_rate": 3e-05,
      "loss": 1.8337,
      "step": 11899
    },
    {
      "epoch": 2.783534089580166,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 11900
    },
    {
      "epoch": 2.783534089580166,
      "eval_runtime": 4.6777,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 11900
    },
    {
      "epoch": 2.783767980353175,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8349,
      "step": 11901
    },
    {
      "epoch": 2.784001871126184,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 11902
    },
    {
      "epoch": 2.7842357618991933,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 11903
    },
    {
      "epoch": 2.784469652672202,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4814,
      "step": 11904
    },
    {
      "epoch": 2.784703543445211,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 11905
    },
    {
      "epoch": 2.78493743421822,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 11906
    },
    {
      "epoch": 2.785171324991229,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 11907
    },
    {
      "epoch": 2.785405215764238,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 11908
    },
    {
      "epoch": 2.7856391065372472,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2359,
      "step": 11909
    },
    {
      "epoch": 2.785872997310256,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 11910
    },
    {
      "epoch": 2.786106888083265,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 11911
    },
    {
      "epoch": 2.786340778856274,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 11912
    },
    {
      "epoch": 2.786574669629283,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 11913
    },
    {
      "epoch": 2.7868085604022923,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 11914
    },
    {
      "epoch": 2.787042451175301,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 11915
    },
    {
      "epoch": 2.78727634194831,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 11916
    },
    {
      "epoch": 2.7875102327213193,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 11917
    },
    {
      "epoch": 2.787744123494328,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6293,
      "step": 11918
    },
    {
      "epoch": 2.7879780142673374,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 11919
    },
    {
      "epoch": 2.788211905040346,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 11920
    },
    {
      "epoch": 2.788445795813355,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 11921
    },
    {
      "epoch": 2.788679686586364,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 11922
    },
    {
      "epoch": 2.788913577359373,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 11923
    },
    {
      "epoch": 2.789147468132382,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 11924
    },
    {
      "epoch": 2.7893813589053913,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6564,
      "step": 11925
    },
    {
      "epoch": 2.7896152496784,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3948,
      "step": 11926
    },
    {
      "epoch": 2.789849140451409,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 11927
    },
    {
      "epoch": 2.7900830312244183,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 11928
    },
    {
      "epoch": 2.790316921997427,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7984,
      "step": 11929
    },
    {
      "epoch": 2.7905508127704364,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 11930
    },
    {
      "epoch": 2.790784703543445,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 11931
    },
    {
      "epoch": 2.791018594316454,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.58,
      "step": 11932
    },
    {
      "epoch": 2.7912524850894633,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 11933
    },
    {
      "epoch": 2.791486375862472,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 11934
    },
    {
      "epoch": 2.7917202666354815,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 11935
    },
    {
      "epoch": 2.7919541574084903,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 11936
    },
    {
      "epoch": 2.792188048181499,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 11937
    },
    {
      "epoch": 2.7924219389545084,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.7803,
      "step": 11938
    },
    {
      "epoch": 2.7926558297275172,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 11939
    },
    {
      "epoch": 2.7928897205005265,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 11940
    },
    {
      "epoch": 2.7931236112735354,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 11941
    },
    {
      "epoch": 2.793357502046544,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 11942
    },
    {
      "epoch": 2.793591392819553,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7401,
      "step": 11943
    },
    {
      "epoch": 2.7938252835925623,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 11944
    },
    {
      "epoch": 2.794059174365571,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5025,
      "step": 11945
    },
    {
      "epoch": 2.7942930651385804,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 11946
    },
    {
      "epoch": 2.7945269559115893,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 11947
    },
    {
      "epoch": 2.794760846684598,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 11948
    },
    {
      "epoch": 2.7949947374576074,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 11949
    },
    {
      "epoch": 2.7952286282306162,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 11950
    },
    {
      "epoch": 2.7954625190036255,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 11951
    },
    {
      "epoch": 2.7956964097766344,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 11952
    },
    {
      "epoch": 2.795930300549643,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6535,
      "step": 11953
    },
    {
      "epoch": 2.7961641913226525,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8165,
      "step": 11954
    },
    {
      "epoch": 2.7963980820956613,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 11955
    },
    {
      "epoch": 2.7966319728686706,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 11956
    },
    {
      "epoch": 2.7968658636416794,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 11957
    },
    {
      "epoch": 2.7970997544146883,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6152,
      "step": 11958
    },
    {
      "epoch": 2.797333645187697,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 11959
    },
    {
      "epoch": 2.7975675359607064,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 11960
    },
    {
      "epoch": 2.7978014267337152,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 11961
    },
    {
      "epoch": 2.7980353175067245,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 11962
    },
    {
      "epoch": 2.7982692082797334,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 11963
    },
    {
      "epoch": 2.798503099052742,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 11964
    },
    {
      "epoch": 2.7987369898257515,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 11965
    },
    {
      "epoch": 2.7989708805987603,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 11966
    },
    {
      "epoch": 2.7992047713717696,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6823,
      "step": 11967
    },
    {
      "epoch": 2.7994386621447784,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 11968
    },
    {
      "epoch": 2.7996725529177873,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 11969
    },
    {
      "epoch": 2.7999064436907966,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 11970
    },
    {
      "epoch": 2.8001403344638054,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 11971
    },
    {
      "epoch": 2.8003742252368147,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 11972
    },
    {
      "epoch": 2.8006081160098235,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 11973
    },
    {
      "epoch": 2.8008420067828323,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7946,
      "step": 11974
    },
    {
      "epoch": 2.801075897555841,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 11975
    },
    {
      "epoch": 2.8013097883288505,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7826,
      "step": 11976
    },
    {
      "epoch": 2.8015436791018593,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 11977
    },
    {
      "epoch": 2.8017775698748686,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 11978
    },
    {
      "epoch": 2.8020114606478774,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 11979
    },
    {
      "epoch": 2.8022453514208863,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6651,
      "step": 11980
    },
    {
      "epoch": 2.8024792421938955,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 11981
    },
    {
      "epoch": 2.8027131329669044,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 11982
    },
    {
      "epoch": 2.8029470237399137,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 11983
    },
    {
      "epoch": 2.8031809145129225,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 11984
    },
    {
      "epoch": 2.8034148052859313,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 11985
    },
    {
      "epoch": 2.8036486960589406,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 11986
    },
    {
      "epoch": 2.8038825868319495,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 11987
    },
    {
      "epoch": 2.8041164776049587,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 11988
    },
    {
      "epoch": 2.8043503683779676,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5034,
      "step": 11989
    },
    {
      "epoch": 2.8045842591509764,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5527,
      "step": 11990
    },
    {
      "epoch": 2.8048181499239853,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 11991
    },
    {
      "epoch": 2.8050520406969945,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 11992
    },
    {
      "epoch": 2.8052859314700034,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1241,
      "step": 11993
    },
    {
      "epoch": 2.8055198222430127,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 11994
    },
    {
      "epoch": 2.8057537130160215,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 11995
    },
    {
      "epoch": 2.8059876037890303,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 11996
    },
    {
      "epoch": 2.8062214945620396,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 11997
    },
    {
      "epoch": 2.8064553853350485,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 11998
    },
    {
      "epoch": 2.8066892761080577,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 11999
    },
    {
      "epoch": 2.8069231668810666,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 12000
    },
    {
      "epoch": 2.8069231668810666,
      "eval_runtime": 4.5957,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 12000
    },
    {
      "epoch": 2.8071570576540754,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 12001
    },
    {
      "epoch": 2.8073909484270847,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 12002
    },
    {
      "epoch": 2.8076248392000935,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 12003
    },
    {
      "epoch": 2.807858729973103,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 12004
    },
    {
      "epoch": 2.8080926207461117,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 12005
    },
    {
      "epoch": 2.8083265115191205,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 12006
    },
    {
      "epoch": 2.8085604022921293,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 12007
    },
    {
      "epoch": 2.8087942930651386,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 12008
    },
    {
      "epoch": 2.8090281838381475,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 12009
    },
    {
      "epoch": 2.8092620746111567,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 12010
    },
    {
      "epoch": 2.8094959653841656,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 12011
    },
    {
      "epoch": 2.8097298561571744,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6287,
      "step": 12012
    },
    {
      "epoch": 2.8099637469301837,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 12013
    },
    {
      "epoch": 2.8101976377031925,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 12014
    },
    {
      "epoch": 2.810431528476202,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 12015
    },
    {
      "epoch": 2.8106654192492107,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 12016
    },
    {
      "epoch": 2.8108993100222195,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 12017
    },
    {
      "epoch": 2.8111332007952288,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 12018
    },
    {
      "epoch": 2.8113670915682376,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 12019
    },
    {
      "epoch": 2.811600982341247,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 12020
    },
    {
      "epoch": 2.8118348731142557,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 12021
    },
    {
      "epoch": 2.8120687638872646,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 12022
    },
    {
      "epoch": 2.8123026546602734,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 12023
    },
    {
      "epoch": 2.8125365454332827,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 12024
    },
    {
      "epoch": 2.8127704362062915,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8318,
      "step": 12025
    },
    {
      "epoch": 2.813004326979301,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6829,
      "step": 12026
    },
    {
      "epoch": 2.8132382177523096,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 12027
    },
    {
      "epoch": 2.8134721085253185,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 12028
    },
    {
      "epoch": 2.8137059992983278,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 12029
    },
    {
      "epoch": 2.8139398900713366,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7649,
      "step": 12030
    },
    {
      "epoch": 2.814173780844346,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 12031
    },
    {
      "epoch": 2.8144076716173547,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 12032
    },
    {
      "epoch": 2.8146415623903636,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4397,
      "step": 12033
    },
    {
      "epoch": 2.814875453163373,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6497,
      "step": 12034
    },
    {
      "epoch": 2.8151093439363817,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 12035
    },
    {
      "epoch": 2.815343234709391,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 12036
    },
    {
      "epoch": 2.8155771254824,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 12037
    },
    {
      "epoch": 2.8158110162554086,
      "grad_norm": 13.3125,
      "learning_rate": 3e-05,
      "loss": 2.6379,
      "step": 12038
    },
    {
      "epoch": 2.8160449070284175,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 12039
    },
    {
      "epoch": 2.8162787978014268,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7579,
      "step": 12040
    },
    {
      "epoch": 2.8165126885744356,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 12041
    },
    {
      "epoch": 2.816746579347445,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4607,
      "step": 12042
    },
    {
      "epoch": 2.8169804701204537,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 12043
    },
    {
      "epoch": 2.8172143608934626,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 12044
    },
    {
      "epoch": 2.817448251666472,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 12045
    },
    {
      "epoch": 2.8176821424394807,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 12046
    },
    {
      "epoch": 2.81791603321249,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 12047
    },
    {
      "epoch": 2.818149923985499,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 12048
    },
    {
      "epoch": 2.8183838147585076,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 12049
    },
    {
      "epoch": 2.818617705531517,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9811,
      "step": 12050
    },
    {
      "epoch": 2.8188515963045258,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6034,
      "step": 12051
    },
    {
      "epoch": 2.819085487077535,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 12052
    },
    {
      "epoch": 2.819319377850544,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 12053
    },
    {
      "epoch": 2.8195532686235527,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.4112,
      "step": 12054
    },
    {
      "epoch": 2.8197871593965615,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.658,
      "step": 12055
    },
    {
      "epoch": 2.820021050169571,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 12056
    },
    {
      "epoch": 2.8202549409425797,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 12057
    },
    {
      "epoch": 2.820488831715589,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 12058
    },
    {
      "epoch": 2.820722722488598,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 12059
    },
    {
      "epoch": 2.8209566132616066,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 12060
    },
    {
      "epoch": 2.821190504034616,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 12061
    },
    {
      "epoch": 2.8214243948076247,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1726,
      "step": 12062
    },
    {
      "epoch": 2.821658285580634,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 12063
    },
    {
      "epoch": 2.821892176353643,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 12064
    },
    {
      "epoch": 2.8221260671266517,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 12065
    },
    {
      "epoch": 2.822359957899661,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 12066
    },
    {
      "epoch": 2.82259384867267,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 12067
    },
    {
      "epoch": 2.822827739445679,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 12068
    },
    {
      "epoch": 2.823061630218688,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 12069
    },
    {
      "epoch": 2.823295520991697,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 12070
    },
    {
      "epoch": 2.8235294117647056,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 12071
    },
    {
      "epoch": 2.823763302537715,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 12072
    },
    {
      "epoch": 2.8239971933107237,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 12073
    },
    {
      "epoch": 2.824231084083733,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 12074
    },
    {
      "epoch": 2.824464974856742,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5174,
      "step": 12075
    },
    {
      "epoch": 2.8246988656297507,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6894,
      "step": 12076
    },
    {
      "epoch": 2.82493275640276,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 12077
    },
    {
      "epoch": 2.825166647175769,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 12078
    },
    {
      "epoch": 2.825400537948778,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8289,
      "step": 12079
    },
    {
      "epoch": 2.825634428721787,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 12080
    },
    {
      "epoch": 2.8258683194947958,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 12081
    },
    {
      "epoch": 2.826102210267805,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 12082
    },
    {
      "epoch": 2.826336101040814,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 12083
    },
    {
      "epoch": 2.826569991813823,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 12084
    },
    {
      "epoch": 2.826803882586832,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 12085
    },
    {
      "epoch": 2.827037773359841,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 12086
    },
    {
      "epoch": 2.8272716641328497,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 12087
    },
    {
      "epoch": 2.827505554905859,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 12088
    },
    {
      "epoch": 2.827739445678868,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 12089
    },
    {
      "epoch": 2.827973336451877,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 12090
    },
    {
      "epoch": 2.828207227224886,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 12091
    },
    {
      "epoch": 2.8284411179978948,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 12092
    },
    {
      "epoch": 2.828675008770904,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6902,
      "step": 12093
    },
    {
      "epoch": 2.828908899543913,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 12094
    },
    {
      "epoch": 2.829142790316922,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8549,
      "step": 12095
    },
    {
      "epoch": 2.829376681089931,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 12096
    },
    {
      "epoch": 2.82961057186294,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 12097
    },
    {
      "epoch": 2.829844462635949,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 12098
    },
    {
      "epoch": 2.830078353408958,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 12099
    },
    {
      "epoch": 2.8303122441819673,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7525,
      "step": 12100
    },
    {
      "epoch": 2.8303122441819673,
      "eval_runtime": 4.6059,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 12100
    },
    {
      "epoch": 2.830546134954976,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 12101
    },
    {
      "epoch": 2.830780025727985,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 12102
    },
    {
      "epoch": 2.8310139165009938,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6369,
      "step": 12103
    },
    {
      "epoch": 2.831247807274003,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 12104
    },
    {
      "epoch": 2.831481698047012,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 12105
    },
    {
      "epoch": 2.831715588820021,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 12106
    },
    {
      "epoch": 2.83194947959303,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 12107
    },
    {
      "epoch": 2.832183370366039,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 12108
    },
    {
      "epoch": 2.832417261139048,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 12109
    },
    {
      "epoch": 2.832651151912057,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 12110
    },
    {
      "epoch": 2.8328850426850662,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 12111
    },
    {
      "epoch": 2.833118933458075,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 12112
    },
    {
      "epoch": 2.833352824231084,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 12113
    },
    {
      "epoch": 2.833586715004093,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4894,
      "step": 12114
    },
    {
      "epoch": 2.833820605777102,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 12115
    },
    {
      "epoch": 2.8340544965501113,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 12116
    },
    {
      "epoch": 2.83428838732312,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 12117
    },
    {
      "epoch": 2.834522278096129,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 12118
    },
    {
      "epoch": 2.834756168869138,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 12119
    },
    {
      "epoch": 2.834990059642147,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 12120
    },
    {
      "epoch": 2.835223950415156,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 12121
    },
    {
      "epoch": 2.8354578411881652,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 12122
    },
    {
      "epoch": 2.835691731961174,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 12123
    },
    {
      "epoch": 2.835925622734183,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 12124
    },
    {
      "epoch": 2.836159513507192,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 12125
    },
    {
      "epoch": 2.836393404280201,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 12126
    },
    {
      "epoch": 2.8366272950532103,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.612,
      "step": 12127
    },
    {
      "epoch": 2.836861185826219,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 12128
    },
    {
      "epoch": 2.837095076599228,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 12129
    },
    {
      "epoch": 2.8373289673722373,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 12130
    },
    {
      "epoch": 2.837562858145246,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8029,
      "step": 12131
    },
    {
      "epoch": 2.8377967489182554,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 12132
    },
    {
      "epoch": 2.8380306396912642,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 12133
    },
    {
      "epoch": 2.838264530464273,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6688,
      "step": 12134
    },
    {
      "epoch": 2.8384984212372824,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 12135
    },
    {
      "epoch": 2.838732312010291,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 12136
    },
    {
      "epoch": 2.8389662027833005,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 12137
    },
    {
      "epoch": 2.8392000935563093,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 12138
    },
    {
      "epoch": 2.839433984329318,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2789,
      "step": 12139
    },
    {
      "epoch": 2.839667875102327,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 12140
    },
    {
      "epoch": 2.8399017658753363,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 12141
    },
    {
      "epoch": 2.840135656648345,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 12142
    },
    {
      "epoch": 2.8403695474213544,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2927,
      "step": 12143
    },
    {
      "epoch": 2.8406034381943632,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 12144
    },
    {
      "epoch": 2.840837328967372,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 12145
    },
    {
      "epoch": 2.8410712197403813,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 12146
    },
    {
      "epoch": 2.84130511051339,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6103,
      "step": 12147
    },
    {
      "epoch": 2.8415390012863995,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 12148
    },
    {
      "epoch": 2.8417728920594083,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5625,
      "step": 12149
    },
    {
      "epoch": 2.842006782832417,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 12150
    },
    {
      "epoch": 2.8422406736054264,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 12151
    },
    {
      "epoch": 2.8424745643784353,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 12152
    },
    {
      "epoch": 2.8427084551514445,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 12153
    },
    {
      "epoch": 2.8429423459244534,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 12154
    },
    {
      "epoch": 2.8431762366974622,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4154,
      "step": 12155
    },
    {
      "epoch": 2.843410127470471,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 12156
    },
    {
      "epoch": 2.8436440182434803,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 12157
    },
    {
      "epoch": 2.843877909016489,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7257,
      "step": 12158
    },
    {
      "epoch": 2.8441117997894985,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6946,
      "step": 12159
    },
    {
      "epoch": 2.8443456905625073,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 12160
    },
    {
      "epoch": 2.844579581335516,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7984,
      "step": 12161
    },
    {
      "epoch": 2.8448134721085254,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 12162
    },
    {
      "epoch": 2.8450473628815343,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 12163
    },
    {
      "epoch": 2.8452812536545435,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 12164
    },
    {
      "epoch": 2.8455151444275524,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 12165
    },
    {
      "epoch": 2.845749035200561,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 12166
    },
    {
      "epoch": 2.8459829259735705,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 12167
    },
    {
      "epoch": 2.8462168167465793,
      "grad_norm": 12.4375,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 12168
    },
    {
      "epoch": 2.8464507075195886,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 12169
    },
    {
      "epoch": 2.8466845982925975,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7259,
      "step": 12170
    },
    {
      "epoch": 2.8469184890656063,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 12171
    },
    {
      "epoch": 2.847152379838615,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 12172
    },
    {
      "epoch": 2.8473862706116244,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 12173
    },
    {
      "epoch": 2.8476201613846333,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7454,
      "step": 12174
    },
    {
      "epoch": 2.8478540521576425,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 12175
    },
    {
      "epoch": 2.8480879429306514,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 12176
    },
    {
      "epoch": 2.84832183370366,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 12177
    },
    {
      "epoch": 2.8485557244766695,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 12178
    },
    {
      "epoch": 2.8487896152496783,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 12179
    },
    {
      "epoch": 2.8490235060226876,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 12180
    },
    {
      "epoch": 2.8492573967956965,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.406,
      "step": 12181
    },
    {
      "epoch": 2.8494912875687053,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 12182
    },
    {
      "epoch": 2.8497251783417146,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 12183
    },
    {
      "epoch": 2.8499590691147234,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 12184
    },
    {
      "epoch": 2.8501929598877327,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 12185
    },
    {
      "epoch": 2.8504268506607415,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5977,
      "step": 12186
    },
    {
      "epoch": 2.8506607414337504,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 12187
    },
    {
      "epoch": 2.850894632206759,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 12188
    },
    {
      "epoch": 2.8511285229797685,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 12189
    },
    {
      "epoch": 2.8513624137527773,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 12190
    },
    {
      "epoch": 2.8515963045257866,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 12191
    },
    {
      "epoch": 2.8518301952987954,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 12192
    },
    {
      "epoch": 2.8520640860718043,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 12193
    },
    {
      "epoch": 2.8522979768448136,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 12194
    },
    {
      "epoch": 2.8525318676178224,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 12195
    },
    {
      "epoch": 2.8527657583908317,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6832,
      "step": 12196
    },
    {
      "epoch": 2.8529996491638405,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 12197
    },
    {
      "epoch": 2.8532335399368494,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 12198
    },
    {
      "epoch": 2.8534674307098586,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 12199
    },
    {
      "epoch": 2.8537013214828675,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8416,
      "step": 12200
    },
    {
      "epoch": 2.8537013214828675,
      "eval_runtime": 4.5923,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 12200
    },
    {
      "epoch": 2.8539352122558768,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 12201
    },
    {
      "epoch": 2.8541691030288856,
      "grad_norm": 9.0,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 12202
    },
    {
      "epoch": 2.8544029938018944,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 12203
    },
    {
      "epoch": 2.8546368845749033,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 12204
    },
    {
      "epoch": 2.8548707753479126,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 12205
    },
    {
      "epoch": 2.8551046661209214,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 12206
    },
    {
      "epoch": 2.8553385568939307,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 12207
    },
    {
      "epoch": 2.8555724476669395,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 12208
    },
    {
      "epoch": 2.8558063384399484,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 12209
    },
    {
      "epoch": 2.8560402292129576,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 12210
    },
    {
      "epoch": 2.8562741199859665,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 12211
    },
    {
      "epoch": 2.8565080107589758,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6522,
      "step": 12212
    },
    {
      "epoch": 2.8567419015319846,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 12213
    },
    {
      "epoch": 2.8569757923049934,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 12214
    },
    {
      "epoch": 2.8572096830780027,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 12215
    },
    {
      "epoch": 2.8574435738510116,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.7081,
      "step": 12216
    },
    {
      "epoch": 2.857677464624021,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 12217
    },
    {
      "epoch": 2.8579113553970297,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 12218
    },
    {
      "epoch": 2.8581452461700385,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 12219
    },
    {
      "epoch": 2.8583791369430473,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 12220
    },
    {
      "epoch": 2.8586130277160566,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 12221
    },
    {
      "epoch": 2.8588469184890655,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4542,
      "step": 12222
    },
    {
      "epoch": 2.8590808092620748,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 12223
    },
    {
      "epoch": 2.8593147000350836,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 12224
    },
    {
      "epoch": 2.8595485908080924,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 12225
    },
    {
      "epoch": 2.8597824815811017,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 12226
    },
    {
      "epoch": 2.8600163723541105,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7152,
      "step": 12227
    },
    {
      "epoch": 2.86025026312712,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 12228
    },
    {
      "epoch": 2.8604841539001287,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5278,
      "step": 12229
    },
    {
      "epoch": 2.8607180446731375,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 12230
    },
    {
      "epoch": 2.860951935446147,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 12231
    },
    {
      "epoch": 2.8611858262191556,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5704,
      "step": 12232
    },
    {
      "epoch": 2.861419716992165,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 12233
    },
    {
      "epoch": 2.8616536077651737,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 12234
    },
    {
      "epoch": 2.8618874985381826,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6857,
      "step": 12235
    },
    {
      "epoch": 2.8621213893111914,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 12236
    },
    {
      "epoch": 2.8623552800842007,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 12237
    },
    {
      "epoch": 2.8625891708572095,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 12238
    },
    {
      "epoch": 2.862823061630219,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 12239
    },
    {
      "epoch": 2.8630569524032277,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 12240
    },
    {
      "epoch": 2.8632908431762365,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.901,
      "step": 12241
    },
    {
      "epoch": 2.863524733949246,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 12242
    },
    {
      "epoch": 2.8637586247222546,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8416,
      "step": 12243
    },
    {
      "epoch": 2.863992515495264,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 12244
    },
    {
      "epoch": 2.8642264062682727,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 12245
    },
    {
      "epoch": 2.8644602970412816,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 12246
    },
    {
      "epoch": 2.864694187814291,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 12247
    },
    {
      "epoch": 2.8649280785872997,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 12248
    },
    {
      "epoch": 2.865161969360309,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 12249
    },
    {
      "epoch": 2.865395860133318,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 12250
    },
    {
      "epoch": 2.8656297509063267,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6785,
      "step": 12251
    },
    {
      "epoch": 2.8658636416793355,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5312,
      "step": 12252
    },
    {
      "epoch": 2.8660975324523448,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 12253
    },
    {
      "epoch": 2.8663314232253536,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 12254
    },
    {
      "epoch": 2.866565313998363,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 12255
    },
    {
      "epoch": 2.8667992047713717,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 12256
    },
    {
      "epoch": 2.8670330955443806,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 12257
    },
    {
      "epoch": 2.86726698631739,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 12258
    },
    {
      "epoch": 2.8675008770903987,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 12259
    },
    {
      "epoch": 2.867734767863408,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6093,
      "step": 12260
    },
    {
      "epoch": 2.867968658636417,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 12261
    },
    {
      "epoch": 2.8682025494094257,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 12262
    },
    {
      "epoch": 2.868436440182435,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 12263
    },
    {
      "epoch": 2.8686703309554438,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 12264
    },
    {
      "epoch": 2.868904221728453,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 12265
    },
    {
      "epoch": 2.869138112501462,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 12266
    },
    {
      "epoch": 2.8693720032744707,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 12267
    },
    {
      "epoch": 2.8696058940474796,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 12268
    },
    {
      "epoch": 2.869839784820489,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 12269
    },
    {
      "epoch": 2.8700736755934977,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 12270
    },
    {
      "epoch": 2.870307566366507,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 12271
    },
    {
      "epoch": 2.870541457139516,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 12272
    },
    {
      "epoch": 2.8707753479125246,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 12273
    },
    {
      "epoch": 2.871009238685534,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 12274
    },
    {
      "epoch": 2.8712431294585428,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 12275
    },
    {
      "epoch": 2.871477020231552,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 12276
    },
    {
      "epoch": 2.871710911004561,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 12277
    },
    {
      "epoch": 2.8719448017775697,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 12278
    },
    {
      "epoch": 2.872178692550579,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 12279
    },
    {
      "epoch": 2.872412583323588,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 12280
    },
    {
      "epoch": 2.872646474096597,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 12281
    },
    {
      "epoch": 2.872880364869606,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8153,
      "step": 12282
    },
    {
      "epoch": 2.873114255642615,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 12283
    },
    {
      "epoch": 2.8733481464156236,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8748,
      "step": 12284
    },
    {
      "epoch": 2.873582037188633,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 12285
    },
    {
      "epoch": 2.8738159279616418,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 12286
    },
    {
      "epoch": 2.874049818734651,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 12287
    },
    {
      "epoch": 2.87428370950766,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6961,
      "step": 12288
    },
    {
      "epoch": 2.8745176002806687,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 12289
    },
    {
      "epoch": 2.874751491053678,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 12290
    },
    {
      "epoch": 2.874985381826687,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 12291
    },
    {
      "epoch": 2.875219272599696,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 12292
    },
    {
      "epoch": 2.875453163372705,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 12293
    },
    {
      "epoch": 2.875687054145714,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 12294
    },
    {
      "epoch": 2.875920944918723,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 12295
    },
    {
      "epoch": 2.876154835691732,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 12296
    },
    {
      "epoch": 2.876388726464741,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 12297
    },
    {
      "epoch": 2.87662261723775,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 12298
    },
    {
      "epoch": 2.876856508010759,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 12299
    },
    {
      "epoch": 2.8770903987837677,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 12300
    },
    {
      "epoch": 2.8770903987837677,
      "eval_runtime": 4.6964,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 12300
    },
    {
      "epoch": 2.877324289556777,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 12301
    },
    {
      "epoch": 2.877558180329786,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 12302
    },
    {
      "epoch": 2.877792071102795,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 12303
    },
    {
      "epoch": 2.878025961875804,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 12304
    },
    {
      "epoch": 2.878259852648813,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 12305
    },
    {
      "epoch": 2.878493743421822,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9724,
      "step": 12306
    },
    {
      "epoch": 2.878727634194831,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 12307
    },
    {
      "epoch": 2.87896152496784,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 12308
    },
    {
      "epoch": 2.879195415740849,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 12309
    },
    {
      "epoch": 2.879429306513858,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 12310
    },
    {
      "epoch": 2.879663197286867,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 12311
    },
    {
      "epoch": 2.879897088059876,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8175,
      "step": 12312
    },
    {
      "epoch": 2.8801309788328853,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9756,
      "step": 12313
    },
    {
      "epoch": 2.880364869605894,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 12314
    },
    {
      "epoch": 2.880598760378903,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 12315
    },
    {
      "epoch": 2.880832651151912,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7449,
      "step": 12316
    },
    {
      "epoch": 2.881066541924921,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5919,
      "step": 12317
    },
    {
      "epoch": 2.88130043269793,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 12318
    },
    {
      "epoch": 2.881534323470939,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 12319
    },
    {
      "epoch": 2.881768214243948,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 12320
    },
    {
      "epoch": 2.882002105016957,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1773,
      "step": 12321
    },
    {
      "epoch": 2.882235995789966,
      "grad_norm": 7.21875,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 12322
    },
    {
      "epoch": 2.882469886562975,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6946,
      "step": 12323
    },
    {
      "epoch": 2.8827037773359843,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9506,
      "step": 12324
    },
    {
      "epoch": 2.882937668108993,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 12325
    },
    {
      "epoch": 2.883171558882002,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 12326
    },
    {
      "epoch": 2.883405449655011,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 12327
    },
    {
      "epoch": 2.88363934042802,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 12328
    },
    {
      "epoch": 2.8838732312010293,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 12329
    },
    {
      "epoch": 2.884107121974038,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 12330
    },
    {
      "epoch": 2.884341012747047,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 12331
    },
    {
      "epoch": 2.8845749035200563,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 12332
    },
    {
      "epoch": 2.884808794293065,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 12333
    },
    {
      "epoch": 2.8850426850660744,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 12334
    },
    {
      "epoch": 2.8852765758390833,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2854,
      "step": 12335
    },
    {
      "epoch": 2.885510466612092,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 12336
    },
    {
      "epoch": 2.885744357385101,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 12337
    },
    {
      "epoch": 2.88597824815811,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 12338
    },
    {
      "epoch": 2.886212138931119,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 12339
    },
    {
      "epoch": 2.8864460297041283,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 12340
    },
    {
      "epoch": 2.886679920477137,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 12341
    },
    {
      "epoch": 2.886913811250146,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 12342
    },
    {
      "epoch": 2.8871477020231553,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 12343
    },
    {
      "epoch": 2.887381592796164,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3792,
      "step": 12344
    },
    {
      "epoch": 2.8876154835691734,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2867,
      "step": 12345
    },
    {
      "epoch": 2.8878493743421823,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6775,
      "step": 12346
    },
    {
      "epoch": 2.888083265115191,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 12347
    },
    {
      "epoch": 2.8883171558882004,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 12348
    },
    {
      "epoch": 2.888551046661209,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 12349
    },
    {
      "epoch": 2.8887849374342185,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 12350
    },
    {
      "epoch": 2.8890188282072273,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 12351
    },
    {
      "epoch": 2.889252718980236,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 12352
    },
    {
      "epoch": 2.889486609753245,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 12353
    },
    {
      "epoch": 2.8897205005262543,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 12354
    },
    {
      "epoch": 2.889954391299263,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 12355
    },
    {
      "epoch": 2.8901882820722724,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 12356
    },
    {
      "epoch": 2.8904221728452812,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 12357
    },
    {
      "epoch": 2.89065606361829,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7209,
      "step": 12358
    },
    {
      "epoch": 2.8908899543912994,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 12359
    },
    {
      "epoch": 2.891123845164308,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 12360
    },
    {
      "epoch": 2.8913577359373175,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 12361
    },
    {
      "epoch": 2.8915916267103263,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 12362
    },
    {
      "epoch": 2.891825517483335,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 12363
    },
    {
      "epoch": 2.8920594082563444,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 12364
    },
    {
      "epoch": 2.8922932990293533,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 12365
    },
    {
      "epoch": 2.8925271898023626,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 12366
    },
    {
      "epoch": 2.8927610805753714,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8846,
      "step": 12367
    },
    {
      "epoch": 2.8929949713483802,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 12368
    },
    {
      "epoch": 2.893228862121389,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7724,
      "step": 12369
    },
    {
      "epoch": 2.8934627528943984,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 12370
    },
    {
      "epoch": 2.893696643667407,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7757,
      "step": 12371
    },
    {
      "epoch": 2.8939305344404165,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 12372
    },
    {
      "epoch": 2.8941644252134253,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 12373
    },
    {
      "epoch": 2.894398315986434,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 12374
    },
    {
      "epoch": 2.8946322067594434,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6426,
      "step": 12375
    },
    {
      "epoch": 2.8948660975324523,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 12376
    },
    {
      "epoch": 2.8950999883054616,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 12377
    },
    {
      "epoch": 2.8953338790784704,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 12378
    },
    {
      "epoch": 2.8955677698514792,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 12379
    },
    {
      "epoch": 2.8958016606244885,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 12380
    },
    {
      "epoch": 2.8960355513974974,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 12381
    },
    {
      "epoch": 2.8962694421705066,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 12382
    },
    {
      "epoch": 2.8965033329435155,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 12383
    },
    {
      "epoch": 2.8967372237165243,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 12384
    },
    {
      "epoch": 2.896971114489533,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6686,
      "step": 12385
    },
    {
      "epoch": 2.8972050052625424,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2643,
      "step": 12386
    },
    {
      "epoch": 2.8974388960355513,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 12387
    },
    {
      "epoch": 2.8976727868085606,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 12388
    },
    {
      "epoch": 2.8979066775815694,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 12389
    },
    {
      "epoch": 2.8981405683545782,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 12390
    },
    {
      "epoch": 2.8983744591275875,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 12391
    },
    {
      "epoch": 2.8986083499005963,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 12392
    },
    {
      "epoch": 2.8988422406736056,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 12393
    },
    {
      "epoch": 2.8990761314466145,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7354,
      "step": 12394
    },
    {
      "epoch": 2.8993100222196233,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 12395
    },
    {
      "epoch": 2.8995439129926326,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 12396
    },
    {
      "epoch": 2.8997778037656414,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 12397
    },
    {
      "epoch": 2.9000116945386507,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 12398
    },
    {
      "epoch": 2.9002455853116595,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 12399
    },
    {
      "epoch": 2.9004794760846684,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4759,
      "step": 12400
    },
    {
      "epoch": 2.9004794760846684,
      "eval_runtime": 4.6414,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 12400
    },
    {
      "epoch": 2.9007133668576772,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5658,
      "step": 12401
    },
    {
      "epoch": 2.9009472576306865,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8305,
      "step": 12402
    },
    {
      "epoch": 2.9011811484036953,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 12403
    },
    {
      "epoch": 2.9014150391767046,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 12404
    },
    {
      "epoch": 2.9016489299497135,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 12405
    },
    {
      "epoch": 2.9018828207227223,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 12406
    },
    {
      "epoch": 2.9021167114957316,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 12407
    },
    {
      "epoch": 2.9023506022687404,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 12408
    },
    {
      "epoch": 2.9025844930417497,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5796,
      "step": 12409
    },
    {
      "epoch": 2.9028183838147585,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 12410
    },
    {
      "epoch": 2.9030522745877674,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 12411
    },
    {
      "epoch": 2.9032861653607767,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 12412
    },
    {
      "epoch": 2.9035200561337855,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 12413
    },
    {
      "epoch": 2.903753946906795,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 12414
    },
    {
      "epoch": 2.9039878376798036,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 12415
    },
    {
      "epoch": 2.9042217284528125,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 12416
    },
    {
      "epoch": 2.9044556192258213,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7443,
      "step": 12417
    },
    {
      "epoch": 2.9046895099988306,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8645,
      "step": 12418
    },
    {
      "epoch": 2.9049234007718394,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 12419
    },
    {
      "epoch": 2.9051572915448487,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 12420
    },
    {
      "epoch": 2.9053911823178575,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7979,
      "step": 12421
    },
    {
      "epoch": 2.9056250730908664,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 12422
    },
    {
      "epoch": 2.9058589638638757,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.916,
      "step": 12423
    },
    {
      "epoch": 2.9060928546368845,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7105,
      "step": 12424
    },
    {
      "epoch": 2.9063267454098938,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7699,
      "step": 12425
    },
    {
      "epoch": 2.9065606361829026,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6145,
      "step": 12426
    },
    {
      "epoch": 2.9067945269559115,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 12427
    },
    {
      "epoch": 2.9070284177289207,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 12428
    },
    {
      "epoch": 2.9072623085019296,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5781,
      "step": 12429
    },
    {
      "epoch": 2.907496199274939,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 12430
    },
    {
      "epoch": 2.9077300900479477,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7089,
      "step": 12431
    },
    {
      "epoch": 2.9079639808209565,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 12432
    },
    {
      "epoch": 2.9081978715939654,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 12433
    },
    {
      "epoch": 2.9084317623669746,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 12434
    },
    {
      "epoch": 2.9086656531399835,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 12435
    },
    {
      "epoch": 2.9088995439129928,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 12436
    },
    {
      "epoch": 2.9091334346860016,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 12437
    },
    {
      "epoch": 2.9093673254590104,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1483,
      "step": 12438
    },
    {
      "epoch": 2.9096012162320197,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 12439
    },
    {
      "epoch": 2.9098351070050286,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7261,
      "step": 12440
    },
    {
      "epoch": 2.910068997778038,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 12441
    },
    {
      "epoch": 2.9103028885510467,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6725,
      "step": 12442
    },
    {
      "epoch": 2.9105367793240555,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.5545,
      "step": 12443
    },
    {
      "epoch": 2.910770670097065,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 12444
    },
    {
      "epoch": 2.9110045608700736,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 12445
    },
    {
      "epoch": 2.911238451643083,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 12446
    },
    {
      "epoch": 2.9114723424160918,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 12447
    },
    {
      "epoch": 2.9117062331891006,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 12448
    },
    {
      "epoch": 2.9119401239621094,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 12449
    },
    {
      "epoch": 2.9121740147351187,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7474,
      "step": 12450
    },
    {
      "epoch": 2.9124079055081276,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9204,
      "step": 12451
    },
    {
      "epoch": 2.912641796281137,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 12452
    },
    {
      "epoch": 2.9128756870541457,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 12453
    },
    {
      "epoch": 2.9131095778271545,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4643,
      "step": 12454
    },
    {
      "epoch": 2.913343468600164,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 12455
    },
    {
      "epoch": 2.9135773593731726,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 12456
    },
    {
      "epoch": 2.913811250146182,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.7997,
      "step": 12457
    },
    {
      "epoch": 2.9140451409191908,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 12458
    },
    {
      "epoch": 2.9142790316921996,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 12459
    },
    {
      "epoch": 2.914512922465209,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 12460
    },
    {
      "epoch": 2.9147468132382177,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.26,
      "step": 12461
    },
    {
      "epoch": 2.914980704011227,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 12462
    },
    {
      "epoch": 2.915214594784236,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 12463
    },
    {
      "epoch": 2.9154484855572447,
      "grad_norm": 9.25,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 12464
    },
    {
      "epoch": 2.9156823763302535,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 12465
    },
    {
      "epoch": 2.915916267103263,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 12466
    },
    {
      "epoch": 2.9161501578762716,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 12467
    },
    {
      "epoch": 2.916384048649281,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 12468
    },
    {
      "epoch": 2.9166179394222898,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 12469
    },
    {
      "epoch": 2.9168518301952986,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 12470
    },
    {
      "epoch": 2.917085720968308,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 12471
    },
    {
      "epoch": 2.9173196117413167,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 12472
    },
    {
      "epoch": 2.917553502514326,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 12473
    },
    {
      "epoch": 2.917787393287335,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6186,
      "step": 12474
    },
    {
      "epoch": 2.9180212840603437,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 12475
    },
    {
      "epoch": 2.918255174833353,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 12476
    },
    {
      "epoch": 2.918489065606362,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 12477
    },
    {
      "epoch": 2.918722956379371,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 12478
    },
    {
      "epoch": 2.91895684715238,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4362,
      "step": 12479
    },
    {
      "epoch": 2.9191907379253887,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7979,
      "step": 12480
    },
    {
      "epoch": 2.9194246286983976,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 12481
    },
    {
      "epoch": 2.919658519471407,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 12482
    },
    {
      "epoch": 2.9198924102444157,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 12483
    },
    {
      "epoch": 2.920126301017425,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 12484
    },
    {
      "epoch": 2.920360191790434,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5687,
      "step": 12485
    },
    {
      "epoch": 2.9205940825634427,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 12486
    },
    {
      "epoch": 2.920827973336452,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 12487
    },
    {
      "epoch": 2.921061864109461,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 12488
    },
    {
      "epoch": 2.92129575488247,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5682,
      "step": 12489
    },
    {
      "epoch": 2.921529645655479,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 12490
    },
    {
      "epoch": 2.9217635364284877,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 12491
    },
    {
      "epoch": 2.921997427201497,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.5479,
      "step": 12492
    },
    {
      "epoch": 2.922231317974506,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 12493
    },
    {
      "epoch": 2.922465208747515,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 12494
    },
    {
      "epoch": 2.922699099520524,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 12495
    },
    {
      "epoch": 2.922932990293533,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 12496
    },
    {
      "epoch": 2.9231668810665417,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 12497
    },
    {
      "epoch": 2.923400771839551,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 12498
    },
    {
      "epoch": 2.9236346626125598,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8026,
      "step": 12499
    },
    {
      "epoch": 2.923868553385569,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 12500
    },
    {
      "epoch": 2.923868553385569,
      "eval_runtime": 4.6368,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 12500
    },
    {
      "epoch": 2.924102444158578,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.66,
      "step": 12501
    },
    {
      "epoch": 2.9243363349315867,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 12502
    },
    {
      "epoch": 2.924570225704596,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2843,
      "step": 12503
    },
    {
      "epoch": 2.924804116477605,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 12504
    },
    {
      "epoch": 2.925038007250614,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 12505
    },
    {
      "epoch": 2.925271898023623,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6832,
      "step": 12506
    },
    {
      "epoch": 2.925505788796632,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 12507
    },
    {
      "epoch": 2.925739679569641,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 12508
    },
    {
      "epoch": 2.92597357034265,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 12509
    },
    {
      "epoch": 2.926207461115659,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 12510
    },
    {
      "epoch": 2.926441351888668,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 12511
    },
    {
      "epoch": 2.926675242661677,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 12512
    },
    {
      "epoch": 2.9269091334346857,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 12513
    },
    {
      "epoch": 2.927143024207695,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 12514
    },
    {
      "epoch": 2.927376914980704,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 12515
    },
    {
      "epoch": 2.927610805753713,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9759,
      "step": 12516
    },
    {
      "epoch": 2.927844696526722,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 12517
    },
    {
      "epoch": 2.928078587299731,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 12518
    },
    {
      "epoch": 2.92831247807274,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 12519
    },
    {
      "epoch": 2.928546368845749,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 12520
    },
    {
      "epoch": 2.928780259618758,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 12521
    },
    {
      "epoch": 2.929014150391767,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.3578,
      "step": 12522
    },
    {
      "epoch": 2.929248041164776,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 12523
    },
    {
      "epoch": 2.929481931937785,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 12524
    },
    {
      "epoch": 2.929715822710794,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 12525
    },
    {
      "epoch": 2.9299497134838033,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 12526
    },
    {
      "epoch": 2.930183604256812,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 12527
    },
    {
      "epoch": 2.930417495029821,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 12528
    },
    {
      "epoch": 2.9306513858028302,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 12529
    },
    {
      "epoch": 2.930885276575839,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 12530
    },
    {
      "epoch": 2.9311191673488484,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 12531
    },
    {
      "epoch": 2.931353058121857,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7266,
      "step": 12532
    },
    {
      "epoch": 2.931586948894866,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 12533
    },
    {
      "epoch": 2.931820839667875,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 12534
    },
    {
      "epoch": 2.932054730440884,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 12535
    },
    {
      "epoch": 2.932288621213893,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 12536
    },
    {
      "epoch": 2.9325225119869023,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 12537
    },
    {
      "epoch": 2.932756402759911,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 12538
    },
    {
      "epoch": 2.93299029353292,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 12539
    },
    {
      "epoch": 2.9332241843059292,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 12540
    },
    {
      "epoch": 2.933458075078938,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 12541
    },
    {
      "epoch": 2.9336919658519474,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 12542
    },
    {
      "epoch": 2.933925856624956,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 12543
    },
    {
      "epoch": 2.934159747397965,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 12544
    },
    {
      "epoch": 2.9343936381709743,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 12545
    },
    {
      "epoch": 2.934627528943983,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 12546
    },
    {
      "epoch": 2.9348614197169924,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6601,
      "step": 12547
    },
    {
      "epoch": 2.9350953104900013,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 12548
    },
    {
      "epoch": 2.93532920126301,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 12549
    },
    {
      "epoch": 2.935563092036019,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6315,
      "step": 12550
    },
    {
      "epoch": 2.9357969828090282,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 12551
    },
    {
      "epoch": 2.936030873582037,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6907,
      "step": 12552
    },
    {
      "epoch": 2.9362647643550464,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 12553
    },
    {
      "epoch": 2.936498655128055,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 12554
    },
    {
      "epoch": 2.936732545901064,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8153,
      "step": 12555
    },
    {
      "epoch": 2.9369664366740733,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 12556
    },
    {
      "epoch": 2.937200327447082,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 12557
    },
    {
      "epoch": 2.9374342182200914,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7364,
      "step": 12558
    },
    {
      "epoch": 2.9376681089931003,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 12559
    },
    {
      "epoch": 2.937901999766109,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 12560
    },
    {
      "epoch": 2.9381358905391184,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7068,
      "step": 12561
    },
    {
      "epoch": 2.9383697813121272,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 12562
    },
    {
      "epoch": 2.9386036720851365,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 12563
    },
    {
      "epoch": 2.9388375628581453,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 12564
    },
    {
      "epoch": 2.939071453631154,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 12565
    },
    {
      "epoch": 2.939305344404163,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 12566
    },
    {
      "epoch": 2.9395392351771723,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 12567
    },
    {
      "epoch": 2.939773125950181,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 12568
    },
    {
      "epoch": 2.9400070167231904,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.701,
      "step": 12569
    },
    {
      "epoch": 2.9402409074961993,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 12570
    },
    {
      "epoch": 2.940474798269208,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 12571
    },
    {
      "epoch": 2.9407086890422174,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 12572
    },
    {
      "epoch": 2.940942579815226,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 12573
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 12574
    },
    {
      "epoch": 2.9414103613612443,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 12575
    },
    {
      "epoch": 2.941644252134253,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 12576
    },
    {
      "epoch": 2.9418781429072625,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 12577
    },
    {
      "epoch": 2.9421120336802713,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 12578
    },
    {
      "epoch": 2.9423459244532806,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8043,
      "step": 12579
    },
    {
      "epoch": 2.9425798152262894,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 12580
    },
    {
      "epoch": 2.9428137059992983,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 12581
    },
    {
      "epoch": 2.943047596772307,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3832,
      "step": 12582
    },
    {
      "epoch": 2.9432814875453164,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5745,
      "step": 12583
    },
    {
      "epoch": 2.943515378318325,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8645,
      "step": 12584
    },
    {
      "epoch": 2.9437492690913345,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5454,
      "step": 12585
    },
    {
      "epoch": 2.9439831598643433,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 12586
    },
    {
      "epoch": 2.944217050637352,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5113,
      "step": 12587
    },
    {
      "epoch": 2.9444509414103615,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 12588
    },
    {
      "epoch": 2.9446848321833703,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 12589
    },
    {
      "epoch": 2.9449187229563796,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 12590
    },
    {
      "epoch": 2.9451526137293884,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 12591
    },
    {
      "epoch": 2.9453865045023973,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 12592
    },
    {
      "epoch": 2.9456203952754065,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 12593
    },
    {
      "epoch": 2.9458542860484154,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 12594
    },
    {
      "epoch": 2.9460881768214247,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5638,
      "step": 12595
    },
    {
      "epoch": 2.9463220675944335,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 12596
    },
    {
      "epoch": 2.9465559583674423,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 12597
    },
    {
      "epoch": 2.946789849140451,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 12598
    },
    {
      "epoch": 2.9470237399134604,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 12599
    },
    {
      "epoch": 2.9472576306864693,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 12600
    },
    {
      "epoch": 2.9472576306864693,
      "eval_runtime": 4.6235,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 12600
    },
    {
      "epoch": 2.9474915214594786,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 12601
    },
    {
      "epoch": 2.9477254122324874,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 12602
    },
    {
      "epoch": 2.9479593030054962,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 12603
    },
    {
      "epoch": 2.9481931937785055,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 12604
    },
    {
      "epoch": 2.9484270845515144,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 12605
    },
    {
      "epoch": 2.9486609753245236,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 12606
    },
    {
      "epoch": 2.9488948660975325,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 12607
    },
    {
      "epoch": 2.9491287568705413,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7432,
      "step": 12608
    },
    {
      "epoch": 2.9493626476435506,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 12609
    },
    {
      "epoch": 2.9495965384165594,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 12610
    },
    {
      "epoch": 2.9498304291895687,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 12611
    },
    {
      "epoch": 2.9500643199625776,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 12612
    },
    {
      "epoch": 2.9502982107355864,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2334,
      "step": 12613
    },
    {
      "epoch": 2.9505321015085952,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 12614
    },
    {
      "epoch": 2.9507659922816045,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 12615
    },
    {
      "epoch": 2.9509998830546134,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 12616
    },
    {
      "epoch": 2.9512337738276226,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 12617
    },
    {
      "epoch": 2.9514676646006315,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.684,
      "step": 12618
    },
    {
      "epoch": 2.9517015553736403,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 12619
    },
    {
      "epoch": 2.9519354461466496,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 12620
    },
    {
      "epoch": 2.9521693369196584,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 12621
    },
    {
      "epoch": 2.9524032276926677,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 12622
    },
    {
      "epoch": 2.9526371184656766,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 12623
    },
    {
      "epoch": 2.9528710092386854,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 12624
    },
    {
      "epoch": 2.9531049000116947,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3095,
      "step": 12625
    },
    {
      "epoch": 2.9533387907847035,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3269,
      "step": 12626
    },
    {
      "epoch": 2.953572681557713,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 12627
    },
    {
      "epoch": 2.9538065723307216,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 12628
    },
    {
      "epoch": 2.9540404631037305,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 12629
    },
    {
      "epoch": 2.9542743538767393,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 12630
    },
    {
      "epoch": 2.9545082446497486,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 12631
    },
    {
      "epoch": 2.9547421354227574,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 12632
    },
    {
      "epoch": 2.9549760261957667,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.4663,
      "step": 12633
    },
    {
      "epoch": 2.9552099169687756,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 12634
    },
    {
      "epoch": 2.9554438077417844,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3415,
      "step": 12635
    },
    {
      "epoch": 2.9556776985147937,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 12636
    },
    {
      "epoch": 2.9559115892878025,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 12637
    },
    {
      "epoch": 2.956145480060812,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 12638
    },
    {
      "epoch": 2.9563793708338206,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 12639
    },
    {
      "epoch": 2.9566132616068295,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 12640
    },
    {
      "epoch": 2.9568471523798388,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 12641
    },
    {
      "epoch": 2.9570810431528476,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 12642
    },
    {
      "epoch": 2.957314933925857,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 12643
    },
    {
      "epoch": 2.9575488246988657,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 12644
    },
    {
      "epoch": 2.9577827154718745,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.192,
      "step": 12645
    },
    {
      "epoch": 2.9580166062448834,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 12646
    },
    {
      "epoch": 2.9582504970178927,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 12647
    },
    {
      "epoch": 2.9584843877909015,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5295,
      "step": 12648
    },
    {
      "epoch": 2.958718278563911,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 12649
    },
    {
      "epoch": 2.9589521693369196,
      "grad_norm": 2.515625,
      "learning_rate": 3e-05,
      "loss": 1.7209,
      "step": 12650
    },
    {
      "epoch": 2.9591860601099285,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 12651
    },
    {
      "epoch": 2.9594199508829377,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.637,
      "step": 12652
    },
    {
      "epoch": 2.9596538416559466,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 12653
    },
    {
      "epoch": 2.959887732428956,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.7194,
      "step": 12654
    },
    {
      "epoch": 2.9601216232019647,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8988,
      "step": 12655
    },
    {
      "epoch": 2.9603555139749735,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 12656
    },
    {
      "epoch": 2.960589404747983,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 12657
    },
    {
      "epoch": 2.9608232955209917,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 12658
    },
    {
      "epoch": 2.961057186294001,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 12659
    },
    {
      "epoch": 2.96129107706701,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 12660
    },
    {
      "epoch": 2.9615249678400186,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.675,
      "step": 12661
    },
    {
      "epoch": 2.9617588586130275,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 12662
    },
    {
      "epoch": 2.9619927493860367,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 12663
    },
    {
      "epoch": 2.9622266401590456,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 12664
    },
    {
      "epoch": 2.962460530932055,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 12665
    },
    {
      "epoch": 2.9626944217050637,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5217,
      "step": 12666
    },
    {
      "epoch": 2.9629283124780725,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 12667
    },
    {
      "epoch": 2.963162203251082,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 12668
    },
    {
      "epoch": 2.9633960940240907,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 12669
    },
    {
      "epoch": 2.9636299847971,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7561,
      "step": 12670
    },
    {
      "epoch": 2.9638638755701088,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7214,
      "step": 12671
    },
    {
      "epoch": 2.9640977663431176,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 12672
    },
    {
      "epoch": 2.964331657116127,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 12673
    },
    {
      "epoch": 2.9645655478891357,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 12674
    },
    {
      "epoch": 2.964799438662145,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 12675
    },
    {
      "epoch": 2.965033329435154,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 12676
    },
    {
      "epoch": 2.9652672202081627,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.743,
      "step": 12677
    },
    {
      "epoch": 2.9655011109811715,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3301,
      "step": 12678
    },
    {
      "epoch": 2.965735001754181,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 12679
    },
    {
      "epoch": 2.9659688925271896,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 12680
    },
    {
      "epoch": 2.966202783300199,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 12681
    },
    {
      "epoch": 2.9664366740732078,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4533,
      "step": 12682
    },
    {
      "epoch": 2.9666705648462166,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 12683
    },
    {
      "epoch": 2.966904455619226,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 12684
    },
    {
      "epoch": 2.9671383463922347,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 12685
    },
    {
      "epoch": 2.967372237165244,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 12686
    },
    {
      "epoch": 2.967606127938253,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 12687
    },
    {
      "epoch": 2.9678400187112617,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 12688
    },
    {
      "epoch": 2.968073909484271,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 12689
    },
    {
      "epoch": 2.96830780025728,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 12690
    },
    {
      "epoch": 2.968541691030289,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 12691
    },
    {
      "epoch": 2.968775581803298,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 12692
    },
    {
      "epoch": 2.9690094725763068,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 12693
    },
    {
      "epoch": 2.9692433633493156,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 12694
    },
    {
      "epoch": 2.969477254122325,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 12695
    },
    {
      "epoch": 2.9697111448953337,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 12696
    },
    {
      "epoch": 2.969945035668343,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 12697
    },
    {
      "epoch": 2.970178926441352,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 12698
    },
    {
      "epoch": 2.9704128172143607,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 12699
    },
    {
      "epoch": 2.97064670798737,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5698,
      "step": 12700
    },
    {
      "epoch": 2.97064670798737,
      "eval_runtime": 4.5999,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 12700
    },
    {
      "epoch": 2.970880598760379,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 12701
    },
    {
      "epoch": 2.971114489533388,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 12702
    },
    {
      "epoch": 2.971348380306397,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 12703
    },
    {
      "epoch": 2.9715822710794058,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 12704
    },
    {
      "epoch": 2.971816161852415,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2327,
      "step": 12705
    },
    {
      "epoch": 2.972050052625424,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 12706
    },
    {
      "epoch": 2.972283943398433,
      "grad_norm": 7.21875,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 12707
    },
    {
      "epoch": 2.972517834171442,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 12708
    },
    {
      "epoch": 2.972751724944451,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7084,
      "step": 12709
    },
    {
      "epoch": 2.9729856157174597,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 12710
    },
    {
      "epoch": 2.973219506490469,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 12711
    },
    {
      "epoch": 2.973453397263478,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 12712
    },
    {
      "epoch": 2.973687288036487,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 12713
    },
    {
      "epoch": 2.973921178809496,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5305,
      "step": 12714
    },
    {
      "epoch": 2.9741550695825048,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 12715
    },
    {
      "epoch": 2.974388960355514,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7803,
      "step": 12716
    },
    {
      "epoch": 2.974622851128523,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.759,
      "step": 12717
    },
    {
      "epoch": 2.974856741901532,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5805,
      "step": 12718
    },
    {
      "epoch": 2.975090632674541,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 12719
    },
    {
      "epoch": 2.97532452344755,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 12720
    },
    {
      "epoch": 2.975558414220559,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7058,
      "step": 12721
    },
    {
      "epoch": 2.975792304993568,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 12722
    },
    {
      "epoch": 2.9760261957665772,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 12723
    },
    {
      "epoch": 2.976260086539586,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 12724
    },
    {
      "epoch": 2.976493977312595,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8473,
      "step": 12725
    },
    {
      "epoch": 2.9767278680856037,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 12726
    },
    {
      "epoch": 2.976961758858613,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 12727
    },
    {
      "epoch": 2.9771956496316223,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 12728
    },
    {
      "epoch": 2.977429540404631,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 12729
    },
    {
      "epoch": 2.97766343117764,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 12730
    },
    {
      "epoch": 2.977897321950649,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 12731
    },
    {
      "epoch": 2.978131212723658,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 12732
    },
    {
      "epoch": 2.978365103496667,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 12733
    },
    {
      "epoch": 2.9785989942696762,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.662,
      "step": 12734
    },
    {
      "epoch": 2.978832885042685,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 12735
    },
    {
      "epoch": 2.979066775815694,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 12736
    },
    {
      "epoch": 2.979300666588703,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.701,
      "step": 12737
    },
    {
      "epoch": 2.979534557361712,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2808,
      "step": 12738
    },
    {
      "epoch": 2.9797684481347213,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 12739
    },
    {
      "epoch": 2.98000233890773,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 12740
    },
    {
      "epoch": 2.980236229680739,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 12741
    },
    {
      "epoch": 2.9804701204537483,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.8349,
      "step": 12742
    },
    {
      "epoch": 2.980704011226757,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 12743
    },
    {
      "epoch": 2.9809379019997664,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 12744
    },
    {
      "epoch": 2.981171792772775,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 12745
    },
    {
      "epoch": 2.981405683545784,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 12746
    },
    {
      "epoch": 2.981639574318793,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 12747
    },
    {
      "epoch": 2.981873465091802,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7447,
      "step": 12748
    },
    {
      "epoch": 2.982107355864811,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5642,
      "step": 12749
    },
    {
      "epoch": 2.9823412466378203,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 12750
    },
    {
      "epoch": 2.982575137410829,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 12751
    },
    {
      "epoch": 2.982809028183838,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 12752
    },
    {
      "epoch": 2.9830429189568473,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 12753
    },
    {
      "epoch": 2.983276809729856,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 12754
    },
    {
      "epoch": 2.9835107005028654,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.3906,
      "step": 12755
    },
    {
      "epoch": 2.983744591275874,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 12756
    },
    {
      "epoch": 2.983978482048883,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6344,
      "step": 12757
    },
    {
      "epoch": 2.9842123728218923,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 12758
    },
    {
      "epoch": 2.984446263594901,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 12759
    },
    {
      "epoch": 2.9846801543679105,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 12760
    },
    {
      "epoch": 2.9849140451409193,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 12761
    },
    {
      "epoch": 2.985147935913928,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.2573,
      "step": 12762
    },
    {
      "epoch": 2.985381826686937,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 12763
    },
    {
      "epoch": 2.9856157174599463,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 12764
    },
    {
      "epoch": 2.985849608232955,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 12765
    },
    {
      "epoch": 2.9860834990059644,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 12766
    },
    {
      "epoch": 2.986317389778973,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 12767
    },
    {
      "epoch": 2.986551280551982,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 12768
    },
    {
      "epoch": 2.9867851713249913,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8905,
      "step": 12769
    },
    {
      "epoch": 2.987019062098,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 12770
    },
    {
      "epoch": 2.9872529528710094,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 12771
    },
    {
      "epoch": 2.9874868436440183,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 12772
    },
    {
      "epoch": 2.987720734417027,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 12773
    },
    {
      "epoch": 2.9879546251900364,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 12774
    },
    {
      "epoch": 2.9881885159630452,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 12775
    },
    {
      "epoch": 2.9884224067360545,
      "grad_norm": 10.375,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 12776
    },
    {
      "epoch": 2.9886562975090634,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 12777
    },
    {
      "epoch": 2.988890188282072,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 12778
    },
    {
      "epoch": 2.989124079055081,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 12779
    },
    {
      "epoch": 2.9893579698280903,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7729,
      "step": 12780
    },
    {
      "epoch": 2.989591860601099,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 12781
    },
    {
      "epoch": 2.9898257513741084,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 12782
    },
    {
      "epoch": 2.9900596421471173,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5144,
      "step": 12783
    },
    {
      "epoch": 2.990293532920126,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 12784
    },
    {
      "epoch": 2.9905274236931354,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 12785
    },
    {
      "epoch": 2.9907613144661442,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 12786
    },
    {
      "epoch": 2.9909952052391535,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6752,
      "step": 12787
    },
    {
      "epoch": 2.9912290960121624,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 12788
    },
    {
      "epoch": 2.991462986785171,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 12789
    },
    {
      "epoch": 2.9916968775581805,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 12790
    },
    {
      "epoch": 2.9919307683311893,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 12791
    },
    {
      "epoch": 2.9921646591041986,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 12792
    },
    {
      "epoch": 2.9923985498772074,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 12793
    },
    {
      "epoch": 2.9926324406502163,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 12794
    },
    {
      "epoch": 2.992866331423225,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 12795
    },
    {
      "epoch": 2.9931002221962344,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 12796
    },
    {
      "epoch": 2.9933341129692432,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 12797
    },
    {
      "epoch": 2.9935680037422525,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 12798
    },
    {
      "epoch": 2.9938018945152614,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 12799
    },
    {
      "epoch": 2.99403578528827,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 12800
    },
    {
      "epoch": 2.99403578528827,
      "eval_runtime": 4.593,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 12800
    },
    {
      "epoch": 2.9942696760612795,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 12801
    },
    {
      "epoch": 2.9945035668342883,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 12802
    },
    {
      "epoch": 2.9947374576072976,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 12803
    },
    {
      "epoch": 2.9949713483803064,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7155,
      "step": 12804
    },
    {
      "epoch": 2.9952052391533153,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 12805
    },
    {
      "epoch": 2.9954391299263246,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 12806
    },
    {
      "epoch": 2.9956730206993334,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 12807
    },
    {
      "epoch": 2.9959069114723427,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 12808
    },
    {
      "epoch": 2.9961408022453515,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 12809
    },
    {
      "epoch": 2.9963746930183603,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 12810
    },
    {
      "epoch": 2.996608583791369,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 12811
    },
    {
      "epoch": 2.9968424745643785,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 12812
    },
    {
      "epoch": 2.9970763653373873,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 12813
    },
    {
      "epoch": 2.9973102561103966,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.554,
      "step": 12814
    },
    {
      "epoch": 2.9975441468834054,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2294,
      "step": 12815
    },
    {
      "epoch": 2.9977780376564143,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8536,
      "step": 12816
    },
    {
      "epoch": 2.9980119284294235,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 12817
    },
    {
      "epoch": 2.9982458192024324,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8714,
      "step": 12818
    },
    {
      "epoch": 2.9984797099754417,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 12819
    },
    {
      "epoch": 2.9987136007484505,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 12820
    },
    {
      "epoch": 2.9989474915214593,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0266,
      "step": 12821
    },
    {
      "epoch": 2.9991813822944686,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 12822
    },
    {
      "epoch": 2.9994152730674775,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 12823
    },
    {
      "epoch": 2.9996491638404867,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5233,
      "step": 12824
    },
    {
      "epoch": 2.9998830546134956,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 12825
    },
    {
      "epoch": 3.000233890773009,
      "grad_norm": 11.0625,
      "learning_rate": 3e-05,
      "loss": 4.723,
      "step": 12826
    },
    {
      "epoch": 3.000467781546018,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 12827
    },
    {
      "epoch": 3.000701672319027,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 12828
    },
    {
      "epoch": 3.0009355630920362,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8165,
      "step": 12829
    },
    {
      "epoch": 3.001169453865045,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 12830
    },
    {
      "epoch": 3.001403344638054,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 12831
    },
    {
      "epoch": 3.001637235411063,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 12832
    },
    {
      "epoch": 3.001871126184072,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 12833
    },
    {
      "epoch": 3.002105016957081,
      "grad_norm": 7.78125,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 12834
    },
    {
      "epoch": 3.00233890773009,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.686,
      "step": 12835
    },
    {
      "epoch": 3.002572798503099,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6816,
      "step": 12836
    },
    {
      "epoch": 3.0028066892761083,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 12837
    },
    {
      "epoch": 3.003040580049117,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 12838
    },
    {
      "epoch": 3.003274470822126,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 12839
    },
    {
      "epoch": 3.0035083615951352,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 12840
    },
    {
      "epoch": 3.003742252368144,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 12841
    },
    {
      "epoch": 3.003976143141153,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7092,
      "step": 12842
    },
    {
      "epoch": 3.004210033914162,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 12843
    },
    {
      "epoch": 3.004443924687171,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6064,
      "step": 12844
    },
    {
      "epoch": 3.0046778154601803,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6755,
      "step": 12845
    },
    {
      "epoch": 3.004911706233189,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 12846
    },
    {
      "epoch": 3.005145597006198,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 12847
    },
    {
      "epoch": 3.0053794877792073,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 12848
    },
    {
      "epoch": 3.005613378552216,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.6053,
      "step": 12849
    },
    {
      "epoch": 3.005847269325225,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 12850
    },
    {
      "epoch": 3.0060811600982342,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 12851
    },
    {
      "epoch": 3.006315050871243,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6653,
      "step": 12852
    },
    {
      "epoch": 3.0065489416442523,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 12853
    },
    {
      "epoch": 3.006782832417261,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.47,
      "step": 12854
    },
    {
      "epoch": 3.00701672319027,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7058,
      "step": 12855
    },
    {
      "epoch": 3.0072506139632793,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 12856
    },
    {
      "epoch": 3.007484504736288,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 12857
    },
    {
      "epoch": 3.007718395509297,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 12858
    },
    {
      "epoch": 3.0079522862823063,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 12859
    },
    {
      "epoch": 3.008186177055315,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 12860
    },
    {
      "epoch": 3.0084200678283244,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 12861
    },
    {
      "epoch": 3.0086539586013332,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 12862
    },
    {
      "epoch": 3.008887849374342,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 12863
    },
    {
      "epoch": 3.0091217401473513,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 12864
    },
    {
      "epoch": 3.00935563092036,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 12865
    },
    {
      "epoch": 3.009589521693369,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7285,
      "step": 12866
    },
    {
      "epoch": 3.0098234124663783,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 12867
    },
    {
      "epoch": 3.010057303239387,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6657,
      "step": 12868
    },
    {
      "epoch": 3.0102911940123964,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1038,
      "step": 12869
    },
    {
      "epoch": 3.0105250847854053,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 12870
    },
    {
      "epoch": 3.010758975558414,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6179,
      "step": 12871
    },
    {
      "epoch": 3.0109928663314234,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 12872
    },
    {
      "epoch": 3.011226757104432,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.788,
      "step": 12873
    },
    {
      "epoch": 3.011460647877441,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 12874
    },
    {
      "epoch": 3.0116945386504503,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 12875
    },
    {
      "epoch": 3.011928429423459,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 12876
    },
    {
      "epoch": 3.0121623201964685,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 12877
    },
    {
      "epoch": 3.0123962109694773,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 12878
    },
    {
      "epoch": 3.012630101742486,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 12879
    },
    {
      "epoch": 3.0128639925154954,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 12880
    },
    {
      "epoch": 3.0130978832885043,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 12881
    },
    {
      "epoch": 3.013331774061513,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 12882
    },
    {
      "epoch": 3.0135656648345224,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 12883
    },
    {
      "epoch": 3.013799555607531,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7037,
      "step": 12884
    },
    {
      "epoch": 3.0140334463805405,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 12885
    },
    {
      "epoch": 3.0142673371535493,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 12886
    },
    {
      "epoch": 3.014501227926558,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 12887
    },
    {
      "epoch": 3.0147351186995675,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 12888
    },
    {
      "epoch": 3.0149690094725763,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 12889
    },
    {
      "epoch": 3.015202900245585,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6365,
      "step": 12890
    },
    {
      "epoch": 3.0154367910185944,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5216,
      "step": 12891
    },
    {
      "epoch": 3.0156706817916032,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 12892
    },
    {
      "epoch": 3.0159045725646125,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 12893
    },
    {
      "epoch": 3.0161384633376214,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.686,
      "step": 12894
    },
    {
      "epoch": 3.01637235411063,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 12895
    },
    {
      "epoch": 3.0166062448836395,
      "grad_norm": 13.9375,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 12896
    },
    {
      "epoch": 3.0168401356566483,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 12897
    },
    {
      "epoch": 3.017074026429657,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 12898
    },
    {
      "epoch": 3.0173079172026664,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 12899
    },
    {
      "epoch": 3.0175418079756753,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 12900
    },
    {
      "epoch": 3.0175418079756753,
      "eval_runtime": 4.6255,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 12900
    },
    {
      "epoch": 3.0177756987486846,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 12901
    },
    {
      "epoch": 3.0180095895216934,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 12902
    },
    {
      "epoch": 3.0182434802947022,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 12903
    },
    {
      "epoch": 3.0184773710677115,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 12904
    },
    {
      "epoch": 3.0187112618407204,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 12905
    },
    {
      "epoch": 3.018945152613729,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 12906
    },
    {
      "epoch": 3.0191790433867385,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 12907
    },
    {
      "epoch": 3.0194129341597473,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5671,
      "step": 12908
    },
    {
      "epoch": 3.0196468249327566,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 12909
    },
    {
      "epoch": 3.0198807157057654,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 12910
    },
    {
      "epoch": 3.0201146064787743,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 12911
    },
    {
      "epoch": 3.0203484972517836,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 12912
    },
    {
      "epoch": 3.0205823880247924,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 12913
    },
    {
      "epoch": 3.0208162787978012,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 12914
    },
    {
      "epoch": 3.0210501695708105,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 12915
    },
    {
      "epoch": 3.0212840603438194,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 12916
    },
    {
      "epoch": 3.0215179511168286,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 12917
    },
    {
      "epoch": 3.0217518418898375,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 12918
    },
    {
      "epoch": 3.0219857326628463,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4884,
      "step": 12919
    },
    {
      "epoch": 3.0222196234358556,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.743,
      "step": 12920
    },
    {
      "epoch": 3.0224535142088644,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6547,
      "step": 12921
    },
    {
      "epoch": 3.0226874049818733,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 12922
    },
    {
      "epoch": 3.0229212957548826,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 12923
    },
    {
      "epoch": 3.0231551865278914,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7728,
      "step": 12924
    },
    {
      "epoch": 3.0233890773009007,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 12925
    },
    {
      "epoch": 3.0236229680739095,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 12926
    },
    {
      "epoch": 3.0238568588469183,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 12927
    },
    {
      "epoch": 3.0240907496199276,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.886,
      "step": 12928
    },
    {
      "epoch": 3.0243246403929365,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 12929
    },
    {
      "epoch": 3.0245585311659453,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 12930
    },
    {
      "epoch": 3.0247924219389546,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 12931
    },
    {
      "epoch": 3.0250263127119634,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 12932
    },
    {
      "epoch": 3.0252602034849727,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 12933
    },
    {
      "epoch": 3.0254940942579815,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6354,
      "step": 12934
    },
    {
      "epoch": 3.0257279850309904,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7421,
      "step": 12935
    },
    {
      "epoch": 3.0259618758039997,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 12936
    },
    {
      "epoch": 3.0261957665770085,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 12937
    },
    {
      "epoch": 3.0264296573500173,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6449,
      "step": 12938
    },
    {
      "epoch": 3.0266635481230266,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6727,
      "step": 12939
    },
    {
      "epoch": 3.0268974388960355,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 12940
    },
    {
      "epoch": 3.0271313296690447,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6852,
      "step": 12941
    },
    {
      "epoch": 3.0273652204420536,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 12942
    },
    {
      "epoch": 3.0275991112150624,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 12943
    },
    {
      "epoch": 3.0278330019880717,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 12944
    },
    {
      "epoch": 3.0280668927610805,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 12945
    },
    {
      "epoch": 3.0283007835340894,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 12946
    },
    {
      "epoch": 3.0285346743070987,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 12947
    },
    {
      "epoch": 3.0287685650801075,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 12948
    },
    {
      "epoch": 3.029002455853117,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6168,
      "step": 12949
    },
    {
      "epoch": 3.0292363466261256,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 12950
    },
    {
      "epoch": 3.0294702373991345,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 12951
    },
    {
      "epoch": 3.0297041281721437,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 12952
    },
    {
      "epoch": 3.0299380189451526,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4954,
      "step": 12953
    },
    {
      "epoch": 3.0301719097181614,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 12954
    },
    {
      "epoch": 3.0304058004911707,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 12955
    },
    {
      "epoch": 3.0306396912641795,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 12956
    },
    {
      "epoch": 3.030873582037189,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5909,
      "step": 12957
    },
    {
      "epoch": 3.0311074728101977,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 12958
    },
    {
      "epoch": 3.0313413635832065,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 12959
    },
    {
      "epoch": 3.0315752543562158,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5917,
      "step": 12960
    },
    {
      "epoch": 3.0318091451292246,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 12961
    },
    {
      "epoch": 3.0320430359022335,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5394,
      "step": 12962
    },
    {
      "epoch": 3.0322769266752427,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 12963
    },
    {
      "epoch": 3.0325108174482516,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6634,
      "step": 12964
    },
    {
      "epoch": 3.032744708221261,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 12965
    },
    {
      "epoch": 3.0329785989942697,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 12966
    },
    {
      "epoch": 3.0332124897672785,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 12967
    },
    {
      "epoch": 3.033446380540288,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 12968
    },
    {
      "epoch": 3.0336802713132967,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 12969
    },
    {
      "epoch": 3.0339141620863055,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6671,
      "step": 12970
    },
    {
      "epoch": 3.0341480528593148,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 12971
    },
    {
      "epoch": 3.0343819436323236,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 12972
    },
    {
      "epoch": 3.034615834405333,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 12973
    },
    {
      "epoch": 3.0348497251783417,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 12974
    },
    {
      "epoch": 3.0350836159513506,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 12975
    },
    {
      "epoch": 3.03531750672436,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3635,
      "step": 12976
    },
    {
      "epoch": 3.0355513974973687,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 12977
    },
    {
      "epoch": 3.0357852882703775,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 12978
    },
    {
      "epoch": 3.036019179043387,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 12979
    },
    {
      "epoch": 3.0362530698163956,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.4283,
      "step": 12980
    },
    {
      "epoch": 3.036486960589405,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 12981
    },
    {
      "epoch": 3.0367208513624138,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 12982
    },
    {
      "epoch": 3.0369547421354226,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5325,
      "step": 12983
    },
    {
      "epoch": 3.037188632908432,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 12984
    },
    {
      "epoch": 3.0374225236814407,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 12985
    },
    {
      "epoch": 3.0376564144544496,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 12986
    },
    {
      "epoch": 3.037890305227459,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 12987
    },
    {
      "epoch": 3.0381241960004677,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.5042,
      "step": 12988
    },
    {
      "epoch": 3.038358086773477,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6692,
      "step": 12989
    },
    {
      "epoch": 3.038591977546486,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8815,
      "step": 12990
    },
    {
      "epoch": 3.0388258683194946,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6035,
      "step": 12991
    },
    {
      "epoch": 3.039059759092504,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 12992
    },
    {
      "epoch": 3.0392936498655128,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0266,
      "step": 12993
    },
    {
      "epoch": 3.0395275406385216,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7292,
      "step": 12994
    },
    {
      "epoch": 3.039761431411531,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 12995
    },
    {
      "epoch": 3.0399953221845397,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 12996
    },
    {
      "epoch": 3.040229212957549,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 12997
    },
    {
      "epoch": 3.040463103730558,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.654,
      "step": 12998
    },
    {
      "epoch": 3.0406969945035667,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 12999
    },
    {
      "epoch": 3.040930885276576,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8846,
      "step": 13000
    },
    {
      "epoch": 3.040930885276576,
      "eval_runtime": 4.62,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 13000
    },
    {
      "epoch": 3.041164776049585,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7737,
      "step": 13001
    },
    {
      "epoch": 3.0413986668225936,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.568,
      "step": 13002
    },
    {
      "epoch": 3.041632557595603,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 13003
    },
    {
      "epoch": 3.0418664483686118,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 13004
    },
    {
      "epoch": 3.042100339141621,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 13005
    },
    {
      "epoch": 3.04233422991463,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 13006
    },
    {
      "epoch": 3.0425681206876387,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 13007
    },
    {
      "epoch": 3.042802011460648,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6581,
      "step": 13008
    },
    {
      "epoch": 3.043035902233657,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 13009
    },
    {
      "epoch": 3.0432697930066657,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 13010
    },
    {
      "epoch": 3.043503683779675,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 13011
    },
    {
      "epoch": 3.043737574552684,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 13012
    },
    {
      "epoch": 3.043971465325693,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 13013
    },
    {
      "epoch": 3.044205356098702,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 13014
    },
    {
      "epoch": 3.0444392468717107,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 13015
    },
    {
      "epoch": 3.04467313764472,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 13016
    },
    {
      "epoch": 3.044907028417729,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 13017
    },
    {
      "epoch": 3.0451409191907377,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 13018
    },
    {
      "epoch": 3.045374809963747,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5348,
      "step": 13019
    },
    {
      "epoch": 3.045608700736756,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8751,
      "step": 13020
    },
    {
      "epoch": 3.045842591509765,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 13021
    },
    {
      "epoch": 3.046076482282774,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5786,
      "step": 13022
    },
    {
      "epoch": 3.046310373055783,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 13023
    },
    {
      "epoch": 3.046544263828792,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 13024
    },
    {
      "epoch": 3.046778154601801,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 13025
    },
    {
      "epoch": 3.04701204537481,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 13026
    },
    {
      "epoch": 3.047245936147819,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 13027
    },
    {
      "epoch": 3.047479826920828,
      "grad_norm": 8.75,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 13028
    },
    {
      "epoch": 3.047713717693837,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6866,
      "step": 13029
    },
    {
      "epoch": 3.047947608466846,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5238,
      "step": 13030
    },
    {
      "epoch": 3.048181499239855,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.6626,
      "step": 13031
    },
    {
      "epoch": 3.048415390012864,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 13032
    },
    {
      "epoch": 3.048649280785873,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 13033
    },
    {
      "epoch": 3.0488831715588818,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 13034
    },
    {
      "epoch": 3.049117062331891,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 13035
    },
    {
      "epoch": 3.0493509531049,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 13036
    },
    {
      "epoch": 3.049584843877909,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 13037
    },
    {
      "epoch": 3.049818734650918,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5854,
      "step": 13038
    },
    {
      "epoch": 3.050052625423927,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 13039
    },
    {
      "epoch": 3.050286516196936,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 13040
    },
    {
      "epoch": 3.050520406969945,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6814,
      "step": 13041
    },
    {
      "epoch": 3.0507542977429543,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 13042
    },
    {
      "epoch": 3.050988188515963,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6792,
      "step": 13043
    },
    {
      "epoch": 3.051222079288972,
      "grad_norm": 9.1875,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 13044
    },
    {
      "epoch": 3.051455970061981,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7764,
      "step": 13045
    },
    {
      "epoch": 3.05168986083499,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7173,
      "step": 13046
    },
    {
      "epoch": 3.051923751607999,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 13047
    },
    {
      "epoch": 3.052157642381008,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7424,
      "step": 13048
    },
    {
      "epoch": 3.052391533154017,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 13049
    },
    {
      "epoch": 3.0526254239270263,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 13050
    },
    {
      "epoch": 3.052859314700035,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 13051
    },
    {
      "epoch": 3.053093205473044,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 13052
    },
    {
      "epoch": 3.0533270962460533,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.7718,
      "step": 13053
    },
    {
      "epoch": 3.053560987019062,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7201,
      "step": 13054
    },
    {
      "epoch": 3.053794877792071,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 13055
    },
    {
      "epoch": 3.05402876856508,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 13056
    },
    {
      "epoch": 3.054262659338089,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 13057
    },
    {
      "epoch": 3.0544965501110983,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.669,
      "step": 13058
    },
    {
      "epoch": 3.054730440884107,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 13059
    },
    {
      "epoch": 3.054964331657116,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 13060
    },
    {
      "epoch": 3.0551982224301253,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 13061
    },
    {
      "epoch": 3.055432113203134,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 13062
    },
    {
      "epoch": 3.055666003976143,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5678,
      "step": 13063
    },
    {
      "epoch": 3.0558998947491522,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 13064
    },
    {
      "epoch": 3.056133785522161,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8058,
      "step": 13065
    },
    {
      "epoch": 3.0563676762951704,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 13066
    },
    {
      "epoch": 3.056601567068179,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 13067
    },
    {
      "epoch": 3.056835457841188,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 13068
    },
    {
      "epoch": 3.0570693486141973,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 13069
    },
    {
      "epoch": 3.057303239387206,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 13070
    },
    {
      "epoch": 3.057537130160215,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8087,
      "step": 13071
    },
    {
      "epoch": 3.0577710209332243,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6183,
      "step": 13072
    },
    {
      "epoch": 3.058004911706233,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6701,
      "step": 13073
    },
    {
      "epoch": 3.0582388024792424,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 13074
    },
    {
      "epoch": 3.0584726932522512,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 13075
    },
    {
      "epoch": 3.05870658402526,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 13076
    },
    {
      "epoch": 3.0589404747982694,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6603,
      "step": 13077
    },
    {
      "epoch": 3.059174365571278,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 13078
    },
    {
      "epoch": 3.059408256344287,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 13079
    },
    {
      "epoch": 3.0596421471172963,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 13080
    },
    {
      "epoch": 3.059876037890305,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 13081
    },
    {
      "epoch": 3.0601099286633144,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 13082
    },
    {
      "epoch": 3.0603438194363233,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 13083
    },
    {
      "epoch": 3.060577710209332,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 13084
    },
    {
      "epoch": 3.0608116009823414,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 13085
    },
    {
      "epoch": 3.0610454917553502,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 13086
    },
    {
      "epoch": 3.061279382528359,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6971,
      "step": 13087
    },
    {
      "epoch": 3.0615132733013684,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4604,
      "step": 13088
    },
    {
      "epoch": 3.061747164074377,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7069,
      "step": 13089
    },
    {
      "epoch": 3.0619810548473865,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 13090
    },
    {
      "epoch": 3.0622149456203953,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 13091
    },
    {
      "epoch": 3.062448836393404,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 13092
    },
    {
      "epoch": 3.0626827271664134,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 13093
    },
    {
      "epoch": 3.0629166179394223,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 13094
    },
    {
      "epoch": 3.063150508712431,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 13095
    },
    {
      "epoch": 3.0633843994854404,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6632,
      "step": 13096
    },
    {
      "epoch": 3.0636182902584492,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 13097
    },
    {
      "epoch": 3.0638521810314585,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 13098
    },
    {
      "epoch": 3.0640860718044673,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5677,
      "step": 13099
    },
    {
      "epoch": 3.064319962577476,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 13100
    },
    {
      "epoch": 3.064319962577476,
      "eval_runtime": 4.5984,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 13100
    },
    {
      "epoch": 3.0645538533504855,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 13101
    },
    {
      "epoch": 3.0647877441234943,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 13102
    },
    {
      "epoch": 3.065021634896503,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 13103
    },
    {
      "epoch": 3.0652555256695124,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 13104
    },
    {
      "epoch": 3.0654894164425213,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 13105
    },
    {
      "epoch": 3.0657233072155305,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 13106
    },
    {
      "epoch": 3.0659571979885394,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 13107
    },
    {
      "epoch": 3.0661910887615482,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 13108
    },
    {
      "epoch": 3.0664249795345575,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 13109
    },
    {
      "epoch": 3.0666588703075663,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 13110
    },
    {
      "epoch": 3.066892761080575,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 13111
    },
    {
      "epoch": 3.0671266518535845,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6317,
      "step": 13112
    },
    {
      "epoch": 3.0673605426265933,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 13113
    },
    {
      "epoch": 3.0675944333996026,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 13114
    },
    {
      "epoch": 3.0678283241726114,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 13115
    },
    {
      "epoch": 3.0680622149456203,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6421,
      "step": 13116
    },
    {
      "epoch": 3.0682961057186295,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 13117
    },
    {
      "epoch": 3.0685299964916384,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 13118
    },
    {
      "epoch": 3.068763887264647,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 13119
    },
    {
      "epoch": 3.0689977780376565,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 13120
    },
    {
      "epoch": 3.0692316688106653,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8176,
      "step": 13121
    },
    {
      "epoch": 3.0694655595836746,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6486,
      "step": 13122
    },
    {
      "epoch": 3.0696994503566835,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 13123
    },
    {
      "epoch": 3.0699333411296923,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5679,
      "step": 13124
    },
    {
      "epoch": 3.0701672319027016,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 13125
    },
    {
      "epoch": 3.0704011226757104,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 13126
    },
    {
      "epoch": 3.0706350134487193,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 13127
    },
    {
      "epoch": 3.0708689042217285,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 13128
    },
    {
      "epoch": 3.0711027949947374,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 13129
    },
    {
      "epoch": 3.0713366857677467,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 13130
    },
    {
      "epoch": 3.0715705765407555,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 13131
    },
    {
      "epoch": 3.0718044673137643,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 13132
    },
    {
      "epoch": 3.0720383580867736,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6735,
      "step": 13133
    },
    {
      "epoch": 3.0722722488597825,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 13134
    },
    {
      "epoch": 3.0725061396327913,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 13135
    },
    {
      "epoch": 3.0727400304058006,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 13136
    },
    {
      "epoch": 3.0729739211788094,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 13137
    },
    {
      "epoch": 3.0732078119518187,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.709,
      "step": 13138
    },
    {
      "epoch": 3.0734417027248275,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 13139
    },
    {
      "epoch": 3.0736755934978364,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7367,
      "step": 13140
    },
    {
      "epoch": 3.0739094842708456,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 13141
    },
    {
      "epoch": 3.0741433750438545,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2054,
      "step": 13142
    },
    {
      "epoch": 3.0743772658168633,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 13143
    },
    {
      "epoch": 3.0746111565898726,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 13144
    },
    {
      "epoch": 3.0748450473628814,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 13145
    },
    {
      "epoch": 3.0750789381358907,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.7769,
      "step": 13146
    },
    {
      "epoch": 3.0753128289088996,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 13147
    },
    {
      "epoch": 3.0755467196819084,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 13148
    },
    {
      "epoch": 3.0757806104549177,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 13149
    },
    {
      "epoch": 3.0760145012279265,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 13150
    },
    {
      "epoch": 3.0762483920009354,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 13151
    },
    {
      "epoch": 3.0764822827739446,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6333,
      "step": 13152
    },
    {
      "epoch": 3.0767161735469535,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 13153
    },
    {
      "epoch": 3.0769500643199628,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 13154
    },
    {
      "epoch": 3.0771839550929716,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 13155
    },
    {
      "epoch": 3.0774178458659804,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 13156
    },
    {
      "epoch": 3.0776517366389897,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7052,
      "step": 13157
    },
    {
      "epoch": 3.0778856274119986,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 13158
    },
    {
      "epoch": 3.0781195181850074,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6828,
      "step": 13159
    },
    {
      "epoch": 3.0783534089580167,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 13160
    },
    {
      "epoch": 3.0785872997310255,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 13161
    },
    {
      "epoch": 3.078821190504035,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1878,
      "step": 13162
    },
    {
      "epoch": 3.0790550812770436,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 13163
    },
    {
      "epoch": 3.0792889720500525,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 13164
    },
    {
      "epoch": 3.0795228628230618,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 13165
    },
    {
      "epoch": 3.0797567535960706,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 13166
    },
    {
      "epoch": 3.0799906443690794,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.629,
      "step": 13167
    },
    {
      "epoch": 3.0802245351420887,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 13168
    },
    {
      "epoch": 3.0804584259150976,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 13169
    },
    {
      "epoch": 3.080692316688107,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6246,
      "step": 13170
    },
    {
      "epoch": 3.0809262074611157,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 13171
    },
    {
      "epoch": 3.0811600982341245,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6341,
      "step": 13172
    },
    {
      "epoch": 3.081393989007134,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 13173
    },
    {
      "epoch": 3.0816278797801426,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5321,
      "step": 13174
    },
    {
      "epoch": 3.0818617705531515,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 13175
    },
    {
      "epoch": 3.0820956613261608,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 13176
    },
    {
      "epoch": 3.0823295520991696,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 13177
    },
    {
      "epoch": 3.082563442872179,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6249,
      "step": 13178
    },
    {
      "epoch": 3.0827973336451877,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 13179
    },
    {
      "epoch": 3.0830312244181965,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6482,
      "step": 13180
    },
    {
      "epoch": 3.083265115191206,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5279,
      "step": 13181
    },
    {
      "epoch": 3.0834990059642147,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 13182
    },
    {
      "epoch": 3.0837328967372235,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 13183
    },
    {
      "epoch": 3.083966787510233,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.3765,
      "step": 13184
    },
    {
      "epoch": 3.0842006782832416,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5626,
      "step": 13185
    },
    {
      "epoch": 3.084434569056251,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 13186
    },
    {
      "epoch": 3.0846684598292597,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 13187
    },
    {
      "epoch": 3.0849023506022686,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 13188
    },
    {
      "epoch": 3.085136241375278,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 13189
    },
    {
      "epoch": 3.0853701321482867,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 13190
    },
    {
      "epoch": 3.0856040229212955,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 13191
    },
    {
      "epoch": 3.085837913694305,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 13192
    },
    {
      "epoch": 3.0860718044673137,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 13193
    },
    {
      "epoch": 3.086305695240323,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 13194
    },
    {
      "epoch": 3.086539586013332,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 13195
    },
    {
      "epoch": 3.0867734767863406,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 13196
    },
    {
      "epoch": 3.08700736755935,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 13197
    },
    {
      "epoch": 3.0872412583323587,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 13198
    },
    {
      "epoch": 3.0874751491053676,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 13199
    },
    {
      "epoch": 3.087709039878377,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 13200
    },
    {
      "epoch": 3.087709039878377,
      "eval_runtime": 4.6633,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 13200
    },
    {
      "epoch": 3.0879429306513857,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8634,
      "step": 13201
    },
    {
      "epoch": 3.088176821424395,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.718,
      "step": 13202
    },
    {
      "epoch": 3.088410712197404,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 13203
    },
    {
      "epoch": 3.0886446029704127,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 13204
    },
    {
      "epoch": 3.088878493743422,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 13205
    },
    {
      "epoch": 3.0891123845164308,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 13206
    },
    {
      "epoch": 3.08934627528944,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 13207
    },
    {
      "epoch": 3.089580166062449,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 13208
    },
    {
      "epoch": 3.0898140568354577,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 13209
    },
    {
      "epoch": 3.090047947608467,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 13210
    },
    {
      "epoch": 3.090281838381476,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 13211
    },
    {
      "epoch": 3.0905157291544847,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 13212
    },
    {
      "epoch": 3.090749619927494,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.629,
      "step": 13213
    },
    {
      "epoch": 3.090983510700503,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9504,
      "step": 13214
    },
    {
      "epoch": 3.0912174014735117,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6508,
      "step": 13215
    },
    {
      "epoch": 3.091451292246521,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5793,
      "step": 13216
    },
    {
      "epoch": 3.0916851830195298,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 13217
    },
    {
      "epoch": 3.091919073792539,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 13218
    },
    {
      "epoch": 3.092152964565548,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 13219
    },
    {
      "epoch": 3.0923868553385567,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 13220
    },
    {
      "epoch": 3.092620746111566,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8708,
      "step": 13221
    },
    {
      "epoch": 3.092854636884575,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6432,
      "step": 13222
    },
    {
      "epoch": 3.093088527657584,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 13223
    },
    {
      "epoch": 3.093322418430593,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 13224
    },
    {
      "epoch": 3.093556309203602,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 13225
    },
    {
      "epoch": 3.093790199976611,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.4458,
      "step": 13226
    },
    {
      "epoch": 3.09402409074962,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 13227
    },
    {
      "epoch": 3.0942579815226288,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 13228
    },
    {
      "epoch": 3.094491872295638,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 13229
    },
    {
      "epoch": 3.094725763068647,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6304,
      "step": 13230
    },
    {
      "epoch": 3.0949596538416557,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 13231
    },
    {
      "epoch": 3.095193544614665,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8608,
      "step": 13232
    },
    {
      "epoch": 3.095427435387674,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 13233
    },
    {
      "epoch": 3.095661326160683,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 13234
    },
    {
      "epoch": 3.095895216933692,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 13235
    },
    {
      "epoch": 3.096129107706701,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 13236
    },
    {
      "epoch": 3.09636299847971,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 13237
    },
    {
      "epoch": 3.096596889252719,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 13238
    },
    {
      "epoch": 3.096830780025728,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 13239
    },
    {
      "epoch": 3.097064670798737,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6609,
      "step": 13240
    },
    {
      "epoch": 3.097298561571746,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 13241
    },
    {
      "epoch": 3.097532452344755,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 13242
    },
    {
      "epoch": 3.097766343117764,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6543,
      "step": 13243
    },
    {
      "epoch": 3.098000233890773,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5234,
      "step": 13244
    },
    {
      "epoch": 3.098234124663782,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 13245
    },
    {
      "epoch": 3.098468015436791,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 13246
    },
    {
      "epoch": 3.0987019062098,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 13247
    },
    {
      "epoch": 3.098935796982809,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6155,
      "step": 13248
    },
    {
      "epoch": 3.099169687755818,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.739,
      "step": 13249
    },
    {
      "epoch": 3.099403578528827,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 13250
    },
    {
      "epoch": 3.099637469301836,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 13251
    },
    {
      "epoch": 3.099871360074845,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.705,
      "step": 13252
    },
    {
      "epoch": 3.100105250847854,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 13253
    },
    {
      "epoch": 3.100339141620863,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 13254
    },
    {
      "epoch": 3.1005730323938723,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 13255
    },
    {
      "epoch": 3.100806923166881,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7525,
      "step": 13256
    },
    {
      "epoch": 3.10104081393989,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 13257
    },
    {
      "epoch": 3.1012747047128992,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.453,
      "step": 13258
    },
    {
      "epoch": 3.101508595485908,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 13259
    },
    {
      "epoch": 3.101742486258917,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 13260
    },
    {
      "epoch": 3.101976377031926,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 13261
    },
    {
      "epoch": 3.102210267804935,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 13262
    },
    {
      "epoch": 3.1024441585779443,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 13263
    },
    {
      "epoch": 3.102678049350953,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 13264
    },
    {
      "epoch": 3.102911940123962,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9313,
      "step": 13265
    },
    {
      "epoch": 3.1031458308969713,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 13266
    },
    {
      "epoch": 3.10337972166998,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7931,
      "step": 13267
    },
    {
      "epoch": 3.103613612442989,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 13268
    },
    {
      "epoch": 3.1038475032159982,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 13269
    },
    {
      "epoch": 3.104081393989007,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 13270
    },
    {
      "epoch": 3.1043152847620163,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.381,
      "step": 13271
    },
    {
      "epoch": 3.104549175535025,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 13272
    },
    {
      "epoch": 3.104783066308034,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 13273
    },
    {
      "epoch": 3.1050169570810433,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 13274
    },
    {
      "epoch": 3.105250847854052,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7456,
      "step": 13275
    },
    {
      "epoch": 3.105484738627061,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 13276
    },
    {
      "epoch": 3.1057186294000703,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 13277
    },
    {
      "epoch": 3.105952520173079,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 13278
    },
    {
      "epoch": 3.1061864109460884,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 13279
    },
    {
      "epoch": 3.106420301719097,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 13280
    },
    {
      "epoch": 3.106654192492106,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 13281
    },
    {
      "epoch": 3.1068880832651153,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 13282
    },
    {
      "epoch": 3.107121974038124,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.591,
      "step": 13283
    },
    {
      "epoch": 3.107355864811133,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 13284
    },
    {
      "epoch": 3.1075897555841423,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6723,
      "step": 13285
    },
    {
      "epoch": 3.107823646357151,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 13286
    },
    {
      "epoch": 3.1080575371301604,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 13287
    },
    {
      "epoch": 3.1082914279031693,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6423,
      "step": 13288
    },
    {
      "epoch": 3.108525318676178,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 13289
    },
    {
      "epoch": 3.1087592094491874,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 13290
    },
    {
      "epoch": 3.108993100222196,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 13291
    },
    {
      "epoch": 3.109226990995205,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 13292
    },
    {
      "epoch": 3.1094608817682143,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 13293
    },
    {
      "epoch": 3.109694772541223,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 13294
    },
    {
      "epoch": 3.1099286633142325,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 13295
    },
    {
      "epoch": 3.1101625540872413,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 13296
    },
    {
      "epoch": 3.11039644486025,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5748,
      "step": 13297
    },
    {
      "epoch": 3.1106303356332594,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.5298,
      "step": 13298
    },
    {
      "epoch": 3.1108642264062683,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 13299
    },
    {
      "epoch": 3.111098117179277,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 13300
    },
    {
      "epoch": 3.111098117179277,
      "eval_runtime": 4.6102,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 13300
    },
    {
      "epoch": 3.1113320079522864,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 13301
    },
    {
      "epoch": 3.111565898725295,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 13302
    },
    {
      "epoch": 3.1117997894983045,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 13303
    },
    {
      "epoch": 3.1120336802713133,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 1.6068,
      "step": 13304
    },
    {
      "epoch": 3.112267571044322,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3114,
      "step": 13305
    },
    {
      "epoch": 3.1125014618173314,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 13306
    },
    {
      "epoch": 3.1127353525903403,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 13307
    },
    {
      "epoch": 3.112969243363349,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 13308
    },
    {
      "epoch": 3.1132031341363584,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 13309
    },
    {
      "epoch": 3.1134370249093672,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7156,
      "step": 13310
    },
    {
      "epoch": 3.1136709156823765,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 13311
    },
    {
      "epoch": 3.1139048064553854,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 13312
    },
    {
      "epoch": 3.114138697228394,
      "grad_norm": 9.9375,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 13313
    },
    {
      "epoch": 3.1143725880014035,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 13314
    },
    {
      "epoch": 3.1146064787744123,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.7502,
      "step": 13315
    },
    {
      "epoch": 3.114840369547421,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 13316
    },
    {
      "epoch": 3.1150742603204304,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 13317
    },
    {
      "epoch": 3.1153081510934393,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 13318
    },
    {
      "epoch": 3.1155420418664486,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 13319
    },
    {
      "epoch": 3.1157759326394574,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6648,
      "step": 13320
    },
    {
      "epoch": 3.1160098234124662,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 13321
    },
    {
      "epoch": 3.1162437141854755,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5479,
      "step": 13322
    },
    {
      "epoch": 3.1164776049584844,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 13323
    },
    {
      "epoch": 3.116711495731493,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 13324
    },
    {
      "epoch": 3.1169453865045025,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 13325
    },
    {
      "epoch": 3.1171792772775113,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 13326
    },
    {
      "epoch": 3.1174131680505206,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 13327
    },
    {
      "epoch": 3.1176470588235294,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 13328
    },
    {
      "epoch": 3.1178809495965383,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 13329
    },
    {
      "epoch": 3.1181148403695476,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 13330
    },
    {
      "epoch": 3.1183487311425564,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5855,
      "step": 13331
    },
    {
      "epoch": 3.1185826219155652,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 13332
    },
    {
      "epoch": 3.1188165126885745,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 13333
    },
    {
      "epoch": 3.1190504034615834,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 13334
    },
    {
      "epoch": 3.1192842942345926,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9017,
      "step": 13335
    },
    {
      "epoch": 3.1195181850076015,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7212,
      "step": 13336
    },
    {
      "epoch": 3.1197520757806103,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 13337
    },
    {
      "epoch": 3.1199859665536196,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5074,
      "step": 13338
    },
    {
      "epoch": 3.1202198573266284,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 13339
    },
    {
      "epoch": 3.1204537480996373,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 13340
    },
    {
      "epoch": 3.1206876388726466,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2264,
      "step": 13341
    },
    {
      "epoch": 3.1209215296456554,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 13342
    },
    {
      "epoch": 3.1211554204186647,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 13343
    },
    {
      "epoch": 3.1213893111916735,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 13344
    },
    {
      "epoch": 3.1216232019646823,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 13345
    },
    {
      "epoch": 3.1218570927376916,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.652,
      "step": 13346
    },
    {
      "epoch": 3.1220909835107005,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6817,
      "step": 13347
    },
    {
      "epoch": 3.1223248742837093,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9759,
      "step": 13348
    },
    {
      "epoch": 3.1225587650567186,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 13349
    },
    {
      "epoch": 3.1227926558297274,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 13350
    },
    {
      "epoch": 3.1230265466027367,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 13351
    },
    {
      "epoch": 3.1232604373757455,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 13352
    },
    {
      "epoch": 3.1234943281487544,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4894,
      "step": 13353
    },
    {
      "epoch": 3.1237282189217637,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7041,
      "step": 13354
    },
    {
      "epoch": 3.1239621096947725,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 13355
    },
    {
      "epoch": 3.1241960004677813,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 13356
    },
    {
      "epoch": 3.1244298912407906,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.557,
      "step": 13357
    },
    {
      "epoch": 3.1246637820137995,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.5828,
      "step": 13358
    },
    {
      "epoch": 3.1248976727868087,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 13359
    },
    {
      "epoch": 3.1251315635598176,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 13360
    },
    {
      "epoch": 3.1253654543328264,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 13361
    },
    {
      "epoch": 3.1255993451058357,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 13362
    },
    {
      "epoch": 3.1258332358788445,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6844,
      "step": 13363
    },
    {
      "epoch": 3.1260671266518534,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 13364
    },
    {
      "epoch": 3.1263010174248627,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8528,
      "step": 13365
    },
    {
      "epoch": 3.1265349081978715,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7781,
      "step": 13366
    },
    {
      "epoch": 3.126768798970881,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 13367
    },
    {
      "epoch": 3.1270026897438896,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 13368
    },
    {
      "epoch": 3.1272365805168985,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7655,
      "step": 13369
    },
    {
      "epoch": 3.1274704712899077,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.3992,
      "step": 13370
    },
    {
      "epoch": 3.1277043620629166,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 13371
    },
    {
      "epoch": 3.127938252835926,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5167,
      "step": 13372
    },
    {
      "epoch": 3.1281721436089347,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 13373
    },
    {
      "epoch": 3.1284060343819435,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 13374
    },
    {
      "epoch": 3.128639925154953,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 13375
    },
    {
      "epoch": 3.1288738159279617,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 13376
    },
    {
      "epoch": 3.1291077067009705,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6847,
      "step": 13377
    },
    {
      "epoch": 3.1293415974739798,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 13378
    },
    {
      "epoch": 3.1295754882469886,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 13379
    },
    {
      "epoch": 3.1298093790199975,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 13380
    },
    {
      "epoch": 3.1300432697930067,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 13381
    },
    {
      "epoch": 3.1302771605660156,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 13382
    },
    {
      "epoch": 3.130511051339025,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 13383
    },
    {
      "epoch": 3.1307449421120337,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7241,
      "step": 13384
    },
    {
      "epoch": 3.1309788328850425,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 13385
    },
    {
      "epoch": 3.131212723658052,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 13386
    },
    {
      "epoch": 3.1314466144310606,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 13387
    },
    {
      "epoch": 3.13168050520407,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6019,
      "step": 13388
    },
    {
      "epoch": 3.1319143959770788,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 13389
    },
    {
      "epoch": 3.1321482867500876,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 13390
    },
    {
      "epoch": 3.132382177523097,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 13391
    },
    {
      "epoch": 3.1326160682961057,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 13392
    },
    {
      "epoch": 3.1328499590691146,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 13393
    },
    {
      "epoch": 3.133083849842124,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 13394
    },
    {
      "epoch": 3.1333177406151327,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 13395
    },
    {
      "epoch": 3.1335516313881415,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 13396
    },
    {
      "epoch": 3.133785522161151,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 13397
    },
    {
      "epoch": 3.1340194129341596,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 13398
    },
    {
      "epoch": 3.134253303707169,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8417,
      "step": 13399
    },
    {
      "epoch": 3.1344871944801778,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6582,
      "step": 13400
    },
    {
      "epoch": 3.1344871944801778,
      "eval_runtime": 4.5996,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 13400
    },
    {
      "epoch": 3.1347210852531866,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4245,
      "step": 13401
    },
    {
      "epoch": 3.134954976026196,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5569,
      "step": 13402
    },
    {
      "epoch": 3.1351888667992047,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 13403
    },
    {
      "epoch": 3.135422757572214,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 13404
    },
    {
      "epoch": 3.135656648345223,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 13405
    },
    {
      "epoch": 3.1358905391182317,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 13406
    },
    {
      "epoch": 3.136124429891241,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 13407
    },
    {
      "epoch": 3.13635832066425,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 13408
    },
    {
      "epoch": 3.1365922114372586,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 13409
    },
    {
      "epoch": 3.136826102210268,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 13410
    },
    {
      "epoch": 3.1370599929832768,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 13411
    },
    {
      "epoch": 3.1372938837562856,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 13412
    },
    {
      "epoch": 3.137527774529295,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 13413
    },
    {
      "epoch": 3.1377616653023037,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 13414
    },
    {
      "epoch": 3.137995556075313,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 13415
    },
    {
      "epoch": 3.138229446848322,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 13416
    },
    {
      "epoch": 3.1384633376213307,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 13417
    },
    {
      "epoch": 3.13869722839434,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 13418
    },
    {
      "epoch": 3.138931119167349,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 13419
    },
    {
      "epoch": 3.139165009940358,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8007,
      "step": 13420
    },
    {
      "epoch": 3.139398900713367,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6271,
      "step": 13421
    },
    {
      "epoch": 3.1396327914863758,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 13422
    },
    {
      "epoch": 3.139866682259385,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6714,
      "step": 13423
    },
    {
      "epoch": 3.140100573032394,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 13424
    },
    {
      "epoch": 3.1403344638054027,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 13425
    },
    {
      "epoch": 3.140568354578412,
      "grad_norm": 8.375,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 13426
    },
    {
      "epoch": 3.140802245351421,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8701,
      "step": 13427
    },
    {
      "epoch": 3.1410361361244297,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5888,
      "step": 13428
    },
    {
      "epoch": 3.141270026897439,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 13429
    },
    {
      "epoch": 3.141503917670448,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6294,
      "step": 13430
    },
    {
      "epoch": 3.141737808443457,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5737,
      "step": 13431
    },
    {
      "epoch": 3.141971699216466,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 13432
    },
    {
      "epoch": 3.1422055899894747,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 13433
    },
    {
      "epoch": 3.142439480762484,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 13434
    },
    {
      "epoch": 3.142673371535493,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 13435
    },
    {
      "epoch": 3.142907262308502,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 13436
    },
    {
      "epoch": 3.143141153081511,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5576,
      "step": 13437
    },
    {
      "epoch": 3.14337504385452,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 13438
    },
    {
      "epoch": 3.143608934627529,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 13439
    },
    {
      "epoch": 3.143842825400538,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 13440
    },
    {
      "epoch": 3.144076716173547,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 13441
    },
    {
      "epoch": 3.144310606946556,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 13442
    },
    {
      "epoch": 3.144544497719565,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6578,
      "step": 13443
    },
    {
      "epoch": 3.1447783884925737,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 13444
    },
    {
      "epoch": 3.145012279265583,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 13445
    },
    {
      "epoch": 3.145246170038592,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 13446
    },
    {
      "epoch": 3.145480060811601,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 13447
    },
    {
      "epoch": 3.14571395158461,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 13448
    },
    {
      "epoch": 3.145947842357619,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7651,
      "step": 13449
    },
    {
      "epoch": 3.146181733130628,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 13450
    },
    {
      "epoch": 3.146415623903637,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 13451
    },
    {
      "epoch": 3.146649514676646,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 13452
    },
    {
      "epoch": 3.146883405449655,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 13453
    },
    {
      "epoch": 3.147117296222664,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 13454
    },
    {
      "epoch": 3.147351186995673,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8047,
      "step": 13455
    },
    {
      "epoch": 3.147585077768682,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 13456
    },
    {
      "epoch": 3.147818968541691,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 13457
    },
    {
      "epoch": 3.1480528593147,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7669,
      "step": 13458
    },
    {
      "epoch": 3.148286750087709,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 13459
    },
    {
      "epoch": 3.148520640860718,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 13460
    },
    {
      "epoch": 3.148754531633727,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 13461
    },
    {
      "epoch": 3.148988422406736,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 13462
    },
    {
      "epoch": 3.149222313179745,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 13463
    },
    {
      "epoch": 3.149456203952754,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5577,
      "step": 13464
    },
    {
      "epoch": 3.149690094725763,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6603,
      "step": 13465
    },
    {
      "epoch": 3.149923985498772,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 13466
    },
    {
      "epoch": 3.150157876271781,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 13467
    },
    {
      "epoch": 3.1503917670447903,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 13468
    },
    {
      "epoch": 3.150625657817799,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8645,
      "step": 13469
    },
    {
      "epoch": 3.150859548590808,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5436,
      "step": 13470
    },
    {
      "epoch": 3.1510934393638173,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 13471
    },
    {
      "epoch": 3.151327330136826,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 13472
    },
    {
      "epoch": 3.151561220909835,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 13473
    },
    {
      "epoch": 3.151795111682844,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 13474
    },
    {
      "epoch": 3.152029002455853,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 13475
    },
    {
      "epoch": 3.1522628932288623,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5885,
      "step": 13476
    },
    {
      "epoch": 3.152496784001871,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5025,
      "step": 13477
    },
    {
      "epoch": 3.15273067477488,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4435,
      "step": 13478
    },
    {
      "epoch": 3.1529645655478893,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 13479
    },
    {
      "epoch": 3.153198456320898,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 13480
    },
    {
      "epoch": 3.153432347093907,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 13481
    },
    {
      "epoch": 3.1536662378669162,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 13482
    },
    {
      "epoch": 3.153900128639925,
      "grad_norm": 10.375,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 13483
    },
    {
      "epoch": 3.1541340194129344,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 13484
    },
    {
      "epoch": 3.154367910185943,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 13485
    },
    {
      "epoch": 3.154601800958952,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7627,
      "step": 13486
    },
    {
      "epoch": 3.1548356917319613,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 13487
    },
    {
      "epoch": 3.15506958250497,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 13488
    },
    {
      "epoch": 3.155303473277979,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 13489
    },
    {
      "epoch": 3.1555373640509883,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 13490
    },
    {
      "epoch": 3.155771254823997,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 13491
    },
    {
      "epoch": 3.1560051455970064,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 13492
    },
    {
      "epoch": 3.1562390363700152,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5115,
      "step": 13493
    },
    {
      "epoch": 3.156472927143024,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 13494
    },
    {
      "epoch": 3.1567068179160334,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 13495
    },
    {
      "epoch": 3.156940708689042,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6906,
      "step": 13496
    },
    {
      "epoch": 3.157174599462051,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 13497
    },
    {
      "epoch": 3.1574084902350603,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 13498
    },
    {
      "epoch": 3.157642381008069,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 13499
    },
    {
      "epoch": 3.1578762717810784,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 13500
    },
    {
      "epoch": 3.1578762717810784,
      "eval_runtime": 4.6168,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 13500
    },
    {
      "epoch": 3.1581101625540873,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 13501
    },
    {
      "epoch": 3.158344053327096,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 13502
    },
    {
      "epoch": 3.1585779441001054,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 13503
    },
    {
      "epoch": 3.1588118348731142,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 13504
    },
    {
      "epoch": 3.159045725646123,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 13505
    },
    {
      "epoch": 3.1592796164191324,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 13506
    },
    {
      "epoch": 3.159513507192141,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 13507
    },
    {
      "epoch": 3.1597473979651505,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 13508
    },
    {
      "epoch": 3.1599812887381593,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6663,
      "step": 13509
    },
    {
      "epoch": 3.160215179511168,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 13510
    },
    {
      "epoch": 3.1604490702841774,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6825,
      "step": 13511
    },
    {
      "epoch": 3.1606829610571863,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 13512
    },
    {
      "epoch": 3.160916851830195,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7461,
      "step": 13513
    },
    {
      "epoch": 3.1611507426032044,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 13514
    },
    {
      "epoch": 3.1613846333762132,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.4226,
      "step": 13515
    },
    {
      "epoch": 3.1616185241492225,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 13516
    },
    {
      "epoch": 3.1618524149222313,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 13517
    },
    {
      "epoch": 3.16208630569524,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 13518
    },
    {
      "epoch": 3.1623201964682495,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 13519
    },
    {
      "epoch": 3.1625540872412583,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 13520
    },
    {
      "epoch": 3.162787978014267,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 13521
    },
    {
      "epoch": 3.1630218687872764,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 13522
    },
    {
      "epoch": 3.1632557595602853,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 1.96,
      "step": 13523
    },
    {
      "epoch": 3.1634896503332945,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 13524
    },
    {
      "epoch": 3.1637235411063034,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 13525
    },
    {
      "epoch": 3.163957431879312,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 13526
    },
    {
      "epoch": 3.1641913226523215,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6768,
      "step": 13527
    },
    {
      "epoch": 3.1644252134253303,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 13528
    },
    {
      "epoch": 3.164659104198339,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 13529
    },
    {
      "epoch": 3.1648929949713485,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 13530
    },
    {
      "epoch": 3.1651268857443573,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 13531
    },
    {
      "epoch": 3.1653607765173666,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 13532
    },
    {
      "epoch": 3.1655946672903754,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.645,
      "step": 13533
    },
    {
      "epoch": 3.1658285580633843,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 13534
    },
    {
      "epoch": 3.1660624488363935,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8417,
      "step": 13535
    },
    {
      "epoch": 3.1662963396094024,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6286,
      "step": 13536
    },
    {
      "epoch": 3.166530230382411,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7469,
      "step": 13537
    },
    {
      "epoch": 3.1667641211554205,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 13538
    },
    {
      "epoch": 3.1669980119284293,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 13539
    },
    {
      "epoch": 3.1672319027014386,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 13540
    },
    {
      "epoch": 3.1674657934744475,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6413,
      "step": 13541
    },
    {
      "epoch": 3.1676996842474563,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4389,
      "step": 13542
    },
    {
      "epoch": 3.1679335750204656,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 13543
    },
    {
      "epoch": 3.1681674657934744,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 13544
    },
    {
      "epoch": 3.1684013565664833,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 13545
    },
    {
      "epoch": 3.1686352473394925,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 13546
    },
    {
      "epoch": 3.1688691381125014,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 13547
    },
    {
      "epoch": 3.1691030288855107,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 13548
    },
    {
      "epoch": 3.1693369196585195,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 13549
    },
    {
      "epoch": 3.1695708104315283,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3931,
      "step": 13550
    },
    {
      "epoch": 3.1698047012045376,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5744,
      "step": 13551
    },
    {
      "epoch": 3.1700385919775464,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6743,
      "step": 13552
    },
    {
      "epoch": 3.1702724827505553,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.717,
      "step": 13553
    },
    {
      "epoch": 3.1705063735235646,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 13554
    },
    {
      "epoch": 3.1707402642965734,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 13555
    },
    {
      "epoch": 3.1709741550695827,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 13556
    },
    {
      "epoch": 3.1712080458425915,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 13557
    },
    {
      "epoch": 3.1714419366156004,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 13558
    },
    {
      "epoch": 3.1716758273886096,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6125,
      "step": 13559
    },
    {
      "epoch": 3.1719097181616185,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 13560
    },
    {
      "epoch": 3.1721436089346273,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 13561
    },
    {
      "epoch": 3.1723774997076366,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 13562
    },
    {
      "epoch": 3.1726113904806454,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 13563
    },
    {
      "epoch": 3.1728452812536547,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 13564
    },
    {
      "epoch": 3.1730791720266636,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8468,
      "step": 13565
    },
    {
      "epoch": 3.1733130627996724,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 13566
    },
    {
      "epoch": 3.1735469535726817,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 13567
    },
    {
      "epoch": 3.1737808443456905,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 13568
    },
    {
      "epoch": 3.1740147351186994,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 13569
    },
    {
      "epoch": 3.1742486258917086,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6741,
      "step": 13570
    },
    {
      "epoch": 3.1744825166647175,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 13571
    },
    {
      "epoch": 3.1747164074377268,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 13572
    },
    {
      "epoch": 3.1749502982107356,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 13573
    },
    {
      "epoch": 3.1751841889837444,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 13574
    },
    {
      "epoch": 3.1754180797567537,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.6678,
      "step": 13575
    },
    {
      "epoch": 3.1756519705297626,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 13576
    },
    {
      "epoch": 3.1758858613027714,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 13577
    },
    {
      "epoch": 3.1761197520757807,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 13578
    },
    {
      "epoch": 3.1763536428487895,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 13579
    },
    {
      "epoch": 3.176587533621799,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 13580
    },
    {
      "epoch": 3.1768214243948076,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 13581
    },
    {
      "epoch": 3.1770553151678165,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 13582
    },
    {
      "epoch": 3.1772892059408258,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 13583
    },
    {
      "epoch": 3.1775230967138346,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 13584
    },
    {
      "epoch": 3.177756987486844,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 13585
    },
    {
      "epoch": 3.1779908782598527,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 13586
    },
    {
      "epoch": 3.1782247690328616,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 13587
    },
    {
      "epoch": 3.178458659805871,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 13588
    },
    {
      "epoch": 3.1786925505788797,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7483,
      "step": 13589
    },
    {
      "epoch": 3.1789264413518885,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6663,
      "step": 13590
    },
    {
      "epoch": 3.179160332124898,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.634,
      "step": 13591
    },
    {
      "epoch": 3.1793942228979066,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 13592
    },
    {
      "epoch": 3.1796281136709155,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 13593
    },
    {
      "epoch": 3.1798620044439248,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 13594
    },
    {
      "epoch": 3.1800958952169336,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 13595
    },
    {
      "epoch": 3.180329785989943,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 13596
    },
    {
      "epoch": 3.1805636767629517,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 13597
    },
    {
      "epoch": 3.1807975675359605,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 13598
    },
    {
      "epoch": 3.18103145830897,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 13599
    },
    {
      "epoch": 3.1812653490819787,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 13600
    },
    {
      "epoch": 3.1812653490819787,
      "eval_runtime": 4.5972,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 13600
    },
    {
      "epoch": 3.181499239854988,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 13601
    },
    {
      "epoch": 3.181733130627997,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 13602
    },
    {
      "epoch": 3.1819670214010056,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 13603
    },
    {
      "epoch": 3.182200912174015,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5111,
      "step": 13604
    },
    {
      "epoch": 3.1824348029470237,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 13605
    },
    {
      "epoch": 3.1826686937200326,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 13606
    },
    {
      "epoch": 3.182902584493042,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 13607
    },
    {
      "epoch": 3.1831364752660507,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 13608
    },
    {
      "epoch": 3.1833703660390595,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 13609
    },
    {
      "epoch": 3.183604256812069,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 13610
    },
    {
      "epoch": 3.1838381475850777,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6881,
      "step": 13611
    },
    {
      "epoch": 3.184072038358087,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 13612
    },
    {
      "epoch": 3.184305929131096,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 13613
    },
    {
      "epoch": 3.1845398199041046,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.038,
      "step": 13614
    },
    {
      "epoch": 3.184773710677114,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 13615
    },
    {
      "epoch": 3.1850076014501227,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 13616
    },
    {
      "epoch": 3.185241492223132,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 13617
    },
    {
      "epoch": 3.185475382996141,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 13618
    },
    {
      "epoch": 3.1857092737691497,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 13619
    },
    {
      "epoch": 3.185943164542159,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 13620
    },
    {
      "epoch": 3.186177055315168,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 13621
    },
    {
      "epoch": 3.1864109460881767,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.729,
      "step": 13622
    },
    {
      "epoch": 3.186644836861186,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 13623
    },
    {
      "epoch": 3.1868787276341948,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 13624
    },
    {
      "epoch": 3.1871126184072036,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 13625
    },
    {
      "epoch": 3.187346509180213,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 13626
    },
    {
      "epoch": 3.1875803999532217,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 13627
    },
    {
      "epoch": 3.187814290726231,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 13628
    },
    {
      "epoch": 3.18804818149924,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 13629
    },
    {
      "epoch": 3.1882820722722487,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 13630
    },
    {
      "epoch": 3.188515963045258,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7699,
      "step": 13631
    },
    {
      "epoch": 3.188749853818267,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 13632
    },
    {
      "epoch": 3.188983744591276,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 13633
    },
    {
      "epoch": 3.189217635364285,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 13634
    },
    {
      "epoch": 3.1894515261372938,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 13635
    },
    {
      "epoch": 3.189685416910303,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 13636
    },
    {
      "epoch": 3.189919307683312,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 13637
    },
    {
      "epoch": 3.1901531984563207,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 13638
    },
    {
      "epoch": 3.19038708922933,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7114,
      "step": 13639
    },
    {
      "epoch": 3.190620980002339,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 13640
    },
    {
      "epoch": 3.1908548707753477,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.5604,
      "step": 13641
    },
    {
      "epoch": 3.191088761548357,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 13642
    },
    {
      "epoch": 3.191322652321366,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 13643
    },
    {
      "epoch": 3.191556543094375,
      "grad_norm": 9.1875,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 13644
    },
    {
      "epoch": 3.191790433867384,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 13645
    },
    {
      "epoch": 3.1920243246403928,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 13646
    },
    {
      "epoch": 3.192258215413402,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 13647
    },
    {
      "epoch": 3.192492106186411,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6887,
      "step": 13648
    },
    {
      "epoch": 3.19272599695942,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 13649
    },
    {
      "epoch": 3.192959887732429,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 13650
    },
    {
      "epoch": 3.193193778505438,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 13651
    },
    {
      "epoch": 3.193427669278447,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 13652
    },
    {
      "epoch": 3.193661560051456,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 13653
    },
    {
      "epoch": 3.193895450824465,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 13654
    },
    {
      "epoch": 3.194129341597474,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6867,
      "step": 13655
    },
    {
      "epoch": 3.194363232370483,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 13656
    },
    {
      "epoch": 3.1945971231434918,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 13657
    },
    {
      "epoch": 3.194831013916501,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 13658
    },
    {
      "epoch": 3.19506490468951,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 13659
    },
    {
      "epoch": 3.195298795462519,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4604,
      "step": 13660
    },
    {
      "epoch": 3.195532686235528,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 13661
    },
    {
      "epoch": 3.195766577008537,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 13662
    },
    {
      "epoch": 3.196000467781546,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 13663
    },
    {
      "epoch": 3.196234358554555,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7223,
      "step": 13664
    },
    {
      "epoch": 3.1964682493275642,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 13665
    },
    {
      "epoch": 3.196702140100573,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6373,
      "step": 13666
    },
    {
      "epoch": 3.196936030873582,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 13667
    },
    {
      "epoch": 3.197169921646591,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.793,
      "step": 13668
    },
    {
      "epoch": 3.1974038124196,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6625,
      "step": 13669
    },
    {
      "epoch": 3.197637703192609,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 13670
    },
    {
      "epoch": 3.197871593965618,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 13671
    },
    {
      "epoch": 3.198105484738627,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 13672
    },
    {
      "epoch": 3.1983393755116363,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 13673
    },
    {
      "epoch": 3.198573266284645,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 13674
    },
    {
      "epoch": 3.198807157057654,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 13675
    },
    {
      "epoch": 3.1990410478306632,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6137,
      "step": 13676
    },
    {
      "epoch": 3.199274938603672,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.584,
      "step": 13677
    },
    {
      "epoch": 3.199508829376681,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 13678
    },
    {
      "epoch": 3.19974272014969,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 13679
    },
    {
      "epoch": 3.199976610922699,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 13680
    },
    {
      "epoch": 3.2002105016957083,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7096,
      "step": 13681
    },
    {
      "epoch": 3.200444392468717,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7727,
      "step": 13682
    },
    {
      "epoch": 3.200678283241726,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 13683
    },
    {
      "epoch": 3.2009121740147353,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7069,
      "step": 13684
    },
    {
      "epoch": 3.201146064787744,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 13685
    },
    {
      "epoch": 3.201379955560753,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 13686
    },
    {
      "epoch": 3.2016138463337622,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 13687
    },
    {
      "epoch": 3.201847737106771,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 13688
    },
    {
      "epoch": 3.2020816278797803,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 13689
    },
    {
      "epoch": 3.202315518652789,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 13690
    },
    {
      "epoch": 3.202549409425798,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 13691
    },
    {
      "epoch": 3.2027833001988073,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5393,
      "step": 13692
    },
    {
      "epoch": 3.203017190971816,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 13693
    },
    {
      "epoch": 3.203251081744825,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 13694
    },
    {
      "epoch": 3.2034849725178343,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 13695
    },
    {
      "epoch": 3.203718863290843,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 13696
    },
    {
      "epoch": 3.2039527540638524,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 13697
    },
    {
      "epoch": 3.204186644836861,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 13698
    },
    {
      "epoch": 3.20442053560987,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5756,
      "step": 13699
    },
    {
      "epoch": 3.2046544263828793,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 13700
    },
    {
      "epoch": 3.2046544263828793,
      "eval_runtime": 4.6017,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 13700
    },
    {
      "epoch": 3.204888317155888,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 13701
    },
    {
      "epoch": 3.205122207928897,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 13702
    },
    {
      "epoch": 3.2053560987019063,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 13703
    },
    {
      "epoch": 3.205589989474915,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 13704
    },
    {
      "epoch": 3.2058238802479244,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 13705
    },
    {
      "epoch": 3.2060577710209333,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7423,
      "step": 13706
    },
    {
      "epoch": 3.206291661793942,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 13707
    },
    {
      "epoch": 3.2065255525669514,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 13708
    },
    {
      "epoch": 3.20675944333996,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 13709
    },
    {
      "epoch": 3.206993334112969,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 13710
    },
    {
      "epoch": 3.2072272248859783,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7184,
      "step": 13711
    },
    {
      "epoch": 3.207461115658987,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 13712
    },
    {
      "epoch": 3.2076950064319965,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 13713
    },
    {
      "epoch": 3.2079288972050053,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 13714
    },
    {
      "epoch": 3.208162787978014,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 13715
    },
    {
      "epoch": 3.2083966787510234,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 13716
    },
    {
      "epoch": 3.2086305695240322,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 13717
    },
    {
      "epoch": 3.208864460297041,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 13718
    },
    {
      "epoch": 3.2090983510700504,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 13719
    },
    {
      "epoch": 3.209332241843059,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 13720
    },
    {
      "epoch": 3.2095661326160685,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 13721
    },
    {
      "epoch": 3.2098000233890773,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 13722
    },
    {
      "epoch": 3.210033914162086,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 13723
    },
    {
      "epoch": 3.2102678049350954,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 13724
    },
    {
      "epoch": 3.2105016957081043,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 13725
    },
    {
      "epoch": 3.210735586481113,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 13726
    },
    {
      "epoch": 3.2109694772541224,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 13727
    },
    {
      "epoch": 3.2112033680271312,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 13728
    },
    {
      "epoch": 3.2114372588001405,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1715,
      "step": 13729
    },
    {
      "epoch": 3.2116711495731494,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 13730
    },
    {
      "epoch": 3.211905040346158,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 13731
    },
    {
      "epoch": 3.2121389311191675,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 13732
    },
    {
      "epoch": 3.2123728218921763,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 13733
    },
    {
      "epoch": 3.212606712665185,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 13734
    },
    {
      "epoch": 3.2128406034381944,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 13735
    },
    {
      "epoch": 3.2130744942112033,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 13736
    },
    {
      "epoch": 3.2133083849842126,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6077,
      "step": 13737
    },
    {
      "epoch": 3.2135422757572214,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 13738
    },
    {
      "epoch": 3.2137761665302302,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 13739
    },
    {
      "epoch": 3.2140100573032395,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7273,
      "step": 13740
    },
    {
      "epoch": 3.2142439480762484,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 13741
    },
    {
      "epoch": 3.214477838849257,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 13742
    },
    {
      "epoch": 3.2147117296222665,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 13743
    },
    {
      "epoch": 3.2149456203952753,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6919,
      "step": 13744
    },
    {
      "epoch": 3.2151795111682846,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4296,
      "step": 13745
    },
    {
      "epoch": 3.2154134019412934,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5411,
      "step": 13746
    },
    {
      "epoch": 3.2156472927143023,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 13747
    },
    {
      "epoch": 3.2158811834873116,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 13748
    },
    {
      "epoch": 3.2161150742603204,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 13749
    },
    {
      "epoch": 3.2163489650333292,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 13750
    },
    {
      "epoch": 3.2165828558063385,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 13751
    },
    {
      "epoch": 3.2168167465793474,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 13752
    },
    {
      "epoch": 3.2170506373523566,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 13753
    },
    {
      "epoch": 3.2172845281253655,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 13754
    },
    {
      "epoch": 3.2175184188983743,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 13755
    },
    {
      "epoch": 3.2177523096713836,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 13756
    },
    {
      "epoch": 3.2179862004443924,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 13757
    },
    {
      "epoch": 3.2182200912174013,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 13758
    },
    {
      "epoch": 3.2184539819904106,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 13759
    },
    {
      "epoch": 3.2186878727634194,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7228,
      "step": 13760
    },
    {
      "epoch": 3.2189217635364287,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7454,
      "step": 13761
    },
    {
      "epoch": 3.2191556543094375,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 13762
    },
    {
      "epoch": 3.2193895450824463,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 13763
    },
    {
      "epoch": 3.2196234358554556,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7651,
      "step": 13764
    },
    {
      "epoch": 3.2198573266284645,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 13765
    },
    {
      "epoch": 3.2200912174014733,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 13766
    },
    {
      "epoch": 3.2203251081744826,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 13767
    },
    {
      "epoch": 3.2205589989474914,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6023,
      "step": 13768
    },
    {
      "epoch": 3.2207928897205007,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 13769
    },
    {
      "epoch": 3.2210267804935095,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 13770
    },
    {
      "epoch": 3.2212606712665184,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 13771
    },
    {
      "epoch": 3.2214945620395277,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 13772
    },
    {
      "epoch": 3.2217284528125365,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 13773
    },
    {
      "epoch": 3.2219623435855453,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 13774
    },
    {
      "epoch": 3.2221962343585546,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 13775
    },
    {
      "epoch": 3.2224301251315635,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.798,
      "step": 13776
    },
    {
      "epoch": 3.2226640159045727,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 13777
    },
    {
      "epoch": 3.2228979066775816,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 13778
    },
    {
      "epoch": 3.2231317974505904,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6719,
      "step": 13779
    },
    {
      "epoch": 3.2233656882235997,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 13780
    },
    {
      "epoch": 3.2235995789966085,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6795,
      "step": 13781
    },
    {
      "epoch": 3.223833469769618,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6725,
      "step": 13782
    },
    {
      "epoch": 3.2240673605426267,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7011,
      "step": 13783
    },
    {
      "epoch": 3.2243012513156355,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7153,
      "step": 13784
    },
    {
      "epoch": 3.224535142088645,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 13785
    },
    {
      "epoch": 3.2247690328616536,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 13786
    },
    {
      "epoch": 3.2250029236346625,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 13787
    },
    {
      "epoch": 3.2252368144076717,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6604,
      "step": 13788
    },
    {
      "epoch": 3.2254707051806806,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 13789
    },
    {
      "epoch": 3.2257045959536894,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7531,
      "step": 13790
    },
    {
      "epoch": 3.2259384867266987,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7266,
      "step": 13791
    },
    {
      "epoch": 3.2261723774997075,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 13792
    },
    {
      "epoch": 3.226406268272717,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 13793
    },
    {
      "epoch": 3.2266401590457257,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.606,
      "step": 13794
    },
    {
      "epoch": 3.2268740498187345,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 13795
    },
    {
      "epoch": 3.2271079405917438,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 13796
    },
    {
      "epoch": 3.2273418313647526,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 13797
    },
    {
      "epoch": 3.227575722137762,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 13798
    },
    {
      "epoch": 3.2278096129107707,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 13799
    },
    {
      "epoch": 3.2280435036837796,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7908,
      "step": 13800
    },
    {
      "epoch": 3.2280435036837796,
      "eval_runtime": 4.5969,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 13800
    },
    {
      "epoch": 3.228277394456789,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8028,
      "step": 13801
    },
    {
      "epoch": 3.2285112852297977,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 13802
    },
    {
      "epoch": 3.2287451760028065,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 13803
    },
    {
      "epoch": 3.228979066775816,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 13804
    },
    {
      "epoch": 3.2292129575488246,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 13805
    },
    {
      "epoch": 3.2294468483218335,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 13806
    },
    {
      "epoch": 3.2296807390948428,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 13807
    },
    {
      "epoch": 3.2299146298678516,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 13808
    },
    {
      "epoch": 3.230148520640861,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 13809
    },
    {
      "epoch": 3.2303824114138697,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 13810
    },
    {
      "epoch": 3.2306163021868786,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.5837,
      "step": 13811
    },
    {
      "epoch": 3.230850192959888,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 13812
    },
    {
      "epoch": 3.2310840837328967,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2204,
      "step": 13813
    },
    {
      "epoch": 3.231317974505906,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 13814
    },
    {
      "epoch": 3.231551865278915,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5894,
      "step": 13815
    },
    {
      "epoch": 3.2317857560519236,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 13816
    },
    {
      "epoch": 3.232019646824933,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.6107,
      "step": 13817
    },
    {
      "epoch": 3.2322535375979418,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 13818
    },
    {
      "epoch": 3.2324874283709506,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 13819
    },
    {
      "epoch": 3.23272131914396,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 13820
    },
    {
      "epoch": 3.2329552099169687,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 13821
    },
    {
      "epoch": 3.2331891006899776,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 13822
    },
    {
      "epoch": 3.233422991462987,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 13823
    },
    {
      "epoch": 3.2336568822359957,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 13824
    },
    {
      "epoch": 3.233890773009005,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 13825
    },
    {
      "epoch": 3.234124663782014,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7939,
      "step": 13826
    },
    {
      "epoch": 3.2343585545550226,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 13827
    },
    {
      "epoch": 3.234592445328032,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 13828
    },
    {
      "epoch": 3.2348263361010408,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8093,
      "step": 13829
    },
    {
      "epoch": 3.23506022687405,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0437,
      "step": 13830
    },
    {
      "epoch": 3.235294117647059,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 13831
    },
    {
      "epoch": 3.2355280084200677,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 13832
    },
    {
      "epoch": 3.235761899193077,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6522,
      "step": 13833
    },
    {
      "epoch": 3.235995789966086,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9301,
      "step": 13834
    },
    {
      "epoch": 3.2362296807390947,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7727,
      "step": 13835
    },
    {
      "epoch": 3.236463571512104,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 13836
    },
    {
      "epoch": 3.236697462285113,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 13837
    },
    {
      "epoch": 3.2369313530581216,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6996,
      "step": 13838
    },
    {
      "epoch": 3.237165243831131,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4084,
      "step": 13839
    },
    {
      "epoch": 3.2373991346041397,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 13840
    },
    {
      "epoch": 3.237633025377149,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 13841
    },
    {
      "epoch": 3.237866916150158,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 13842
    },
    {
      "epoch": 3.2381008069231667,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 13843
    },
    {
      "epoch": 3.238334697696176,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 13844
    },
    {
      "epoch": 3.238568588469185,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.615,
      "step": 13845
    },
    {
      "epoch": 3.238802479242194,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 13846
    },
    {
      "epoch": 3.239036370015203,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 13847
    },
    {
      "epoch": 3.239270260788212,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 13848
    },
    {
      "epoch": 3.239504151561221,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 13849
    },
    {
      "epoch": 3.23973804233423,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 13850
    },
    {
      "epoch": 3.2399719331072387,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 13851
    },
    {
      "epoch": 3.240205823880248,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 13852
    },
    {
      "epoch": 3.240439714653257,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 13853
    },
    {
      "epoch": 3.2406736054262657,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 13854
    },
    {
      "epoch": 3.240907496199275,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 13855
    },
    {
      "epoch": 3.241141386972284,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7627,
      "step": 13856
    },
    {
      "epoch": 3.241375277745293,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 13857
    },
    {
      "epoch": 3.241609168518302,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 13858
    },
    {
      "epoch": 3.241843059291311,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 13859
    },
    {
      "epoch": 3.24207695006432,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 13860
    },
    {
      "epoch": 3.242310840837329,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 13861
    },
    {
      "epoch": 3.242544731610338,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 13862
    },
    {
      "epoch": 3.242778622383347,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 13863
    },
    {
      "epoch": 3.243012513156356,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 13864
    },
    {
      "epoch": 3.243246403929365,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 13865
    },
    {
      "epoch": 3.243480294702374,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 13866
    },
    {
      "epoch": 3.243714185475383,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 13867
    },
    {
      "epoch": 3.243948076248392,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 13868
    },
    {
      "epoch": 3.244181967021401,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6998,
      "step": 13869
    },
    {
      "epoch": 3.24441585779441,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 13870
    },
    {
      "epoch": 3.244649748567419,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 13871
    },
    {
      "epoch": 3.244883639340428,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 13872
    },
    {
      "epoch": 3.245117530113437,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 13873
    },
    {
      "epoch": 3.245351420886446,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 13874
    },
    {
      "epoch": 3.245585311659455,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 13875
    },
    {
      "epoch": 3.245819202432464,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 13876
    },
    {
      "epoch": 3.246053093205473,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 13877
    },
    {
      "epoch": 3.2462869839784823,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 13878
    },
    {
      "epoch": 3.246520874751491,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 13879
    },
    {
      "epoch": 3.2467547655245,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 13880
    },
    {
      "epoch": 3.246988656297509,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.4425,
      "step": 13881
    },
    {
      "epoch": 3.247222547070518,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 13882
    },
    {
      "epoch": 3.247456437843527,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1904,
      "step": 13883
    },
    {
      "epoch": 3.247690328616536,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.716,
      "step": 13884
    },
    {
      "epoch": 3.247924219389545,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 13885
    },
    {
      "epoch": 3.2481581101625543,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 13886
    },
    {
      "epoch": 3.248392000935563,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 13887
    },
    {
      "epoch": 3.248625891708572,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 13888
    },
    {
      "epoch": 3.2488597824815812,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6705,
      "step": 13889
    },
    {
      "epoch": 3.24909367325459,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 13890
    },
    {
      "epoch": 3.249327564027599,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0599,
      "step": 13891
    },
    {
      "epoch": 3.249561454800608,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 13892
    },
    {
      "epoch": 3.249795345573617,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 13893
    },
    {
      "epoch": 3.2500292363466263,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 13894
    },
    {
      "epoch": 3.250263127119635,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 13895
    },
    {
      "epoch": 3.250497017892644,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 13896
    },
    {
      "epoch": 3.2507309086656533,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6097,
      "step": 13897
    },
    {
      "epoch": 3.250964799438662,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 13898
    },
    {
      "epoch": 3.251198690211671,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 13899
    },
    {
      "epoch": 3.2514325809846802,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 13900
    },
    {
      "epoch": 3.2514325809846802,
      "eval_runtime": 4.6403,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 13900
    },
    {
      "epoch": 3.251666471757689,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 13901
    },
    {
      "epoch": 3.251900362530698,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8693,
      "step": 13902
    },
    {
      "epoch": 3.252134253303707,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 13903
    },
    {
      "epoch": 3.252368144076716,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 13904
    },
    {
      "epoch": 3.2526020348497253,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 13905
    },
    {
      "epoch": 3.252835925622734,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 13906
    },
    {
      "epoch": 3.253069816395743,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7013,
      "step": 13907
    },
    {
      "epoch": 3.2533037071687523,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 13908
    },
    {
      "epoch": 3.253537597941761,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 13909
    },
    {
      "epoch": 3.2537714887147704,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6556,
      "step": 13910
    },
    {
      "epoch": 3.2540053794877792,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 13911
    },
    {
      "epoch": 3.254239270260788,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 13912
    },
    {
      "epoch": 3.2544731610337974,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 13913
    },
    {
      "epoch": 3.254707051806806,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 13914
    },
    {
      "epoch": 3.254940942579815,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 13915
    },
    {
      "epoch": 3.2551748333528243,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 13916
    },
    {
      "epoch": 3.255408724125833,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 13917
    },
    {
      "epoch": 3.255642614898842,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 13918
    },
    {
      "epoch": 3.2558765056718513,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 13919
    },
    {
      "epoch": 3.25611039644486,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 13920
    },
    {
      "epoch": 3.2563442872178694,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 13921
    },
    {
      "epoch": 3.2565781779908782,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 13922
    },
    {
      "epoch": 3.256812068763887,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 13923
    },
    {
      "epoch": 3.2570459595368964,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 13924
    },
    {
      "epoch": 3.257279850309905,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 13925
    },
    {
      "epoch": 3.2575137410829145,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 13926
    },
    {
      "epoch": 3.2577476318559233,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 13927
    },
    {
      "epoch": 3.257981522628932,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 13928
    },
    {
      "epoch": 3.2582154134019414,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 13929
    },
    {
      "epoch": 3.2584493041749503,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 13930
    },
    {
      "epoch": 3.2586831949479595,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 13931
    },
    {
      "epoch": 3.2589170857209684,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 13932
    },
    {
      "epoch": 3.2591509764939772,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 13933
    },
    {
      "epoch": 3.2593848672669865,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 13934
    },
    {
      "epoch": 3.2596187580399953,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 13935
    },
    {
      "epoch": 3.259852648813004,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 13936
    },
    {
      "epoch": 3.2600865395860135,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 13937
    },
    {
      "epoch": 3.2603204303590223,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 13938
    },
    {
      "epoch": 3.260554321132031,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 13939
    },
    {
      "epoch": 3.2607882119050404,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 13940
    },
    {
      "epoch": 3.2610221026780493,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 13941
    },
    {
      "epoch": 3.2612559934510585,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6765,
      "step": 13942
    },
    {
      "epoch": 3.2614898842240674,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 13943
    },
    {
      "epoch": 3.261723774997076,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 13944
    },
    {
      "epoch": 3.2619576657700855,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 13945
    },
    {
      "epoch": 3.2621915565430943,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 13946
    },
    {
      "epoch": 3.2624254473161036,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 13947
    },
    {
      "epoch": 3.2626593380891125,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 13948
    },
    {
      "epoch": 3.2628932288621213,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 13949
    },
    {
      "epoch": 3.2631271196351306,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 13950
    },
    {
      "epoch": 3.2633610104081394,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 13951
    },
    {
      "epoch": 3.2635949011811483,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8136,
      "step": 13952
    },
    {
      "epoch": 3.2638287919541575,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 13953
    },
    {
      "epoch": 3.2640626827271664,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 13954
    },
    {
      "epoch": 3.264296573500175,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 13955
    },
    {
      "epoch": 3.2645304642731845,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 13956
    },
    {
      "epoch": 3.2647643550461933,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 13957
    },
    {
      "epoch": 3.2649982458192026,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 13958
    },
    {
      "epoch": 3.2652321365922115,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 13959
    },
    {
      "epoch": 3.2654660273652203,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 13960
    },
    {
      "epoch": 3.2656999181382296,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 13961
    },
    {
      "epoch": 3.2659338089112384,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7819,
      "step": 13962
    },
    {
      "epoch": 3.2661676996842477,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 13963
    },
    {
      "epoch": 3.2664015904572565,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 13964
    },
    {
      "epoch": 3.2666354812302654,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 13965
    },
    {
      "epoch": 3.2668693720032747,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 13966
    },
    {
      "epoch": 3.2671032627762835,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 13967
    },
    {
      "epoch": 3.2673371535492923,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 13968
    },
    {
      "epoch": 3.2675710443223016,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 13969
    },
    {
      "epoch": 3.2678049350953104,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 13970
    },
    {
      "epoch": 3.2680388258683193,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 13971
    },
    {
      "epoch": 3.2682727166413286,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 13972
    },
    {
      "epoch": 3.2685066074143374,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 13973
    },
    {
      "epoch": 3.2687404981873467,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8025,
      "step": 13974
    },
    {
      "epoch": 3.2689743889603555,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5425,
      "step": 13975
    },
    {
      "epoch": 3.2692082797333644,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 13976
    },
    {
      "epoch": 3.2694421705063736,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 13977
    },
    {
      "epoch": 3.2696760612793825,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 13978
    },
    {
      "epoch": 3.2699099520523918,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 13979
    },
    {
      "epoch": 3.2701438428254006,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 13980
    },
    {
      "epoch": 3.2703777335984094,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 13981
    },
    {
      "epoch": 3.2706116243714187,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 13982
    },
    {
      "epoch": 3.2708455151444276,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9831,
      "step": 13983
    },
    {
      "epoch": 3.2710794059174364,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 13984
    },
    {
      "epoch": 3.2713132966904457,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 13985
    },
    {
      "epoch": 3.2715471874634545,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 13986
    },
    {
      "epoch": 3.2717810782364634,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 13987
    },
    {
      "epoch": 3.2720149690094726,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 13988
    },
    {
      "epoch": 3.2722488597824815,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5686,
      "step": 13989
    },
    {
      "epoch": 3.2724827505554908,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 13990
    },
    {
      "epoch": 3.2727166413284996,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 13991
    },
    {
      "epoch": 3.2729505321015084,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.707,
      "step": 13992
    },
    {
      "epoch": 3.2731844228745177,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 13993
    },
    {
      "epoch": 3.2734183136475266,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 13994
    },
    {
      "epoch": 3.273652204420536,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.5296,
      "step": 13995
    },
    {
      "epoch": 3.2738860951935447,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 13996
    },
    {
      "epoch": 3.2741199859665535,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7157,
      "step": 13997
    },
    {
      "epoch": 3.274353876739563,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 13998
    },
    {
      "epoch": 3.2745877675125716,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7134,
      "step": 13999
    },
    {
      "epoch": 3.2748216582855805,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7397,
      "step": 14000
    },
    {
      "epoch": 3.2748216582855805,
      "eval_runtime": 4.6057,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 14000
    },
    {
      "epoch": 3.2750555490585898,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 14001
    },
    {
      "epoch": 3.2752894398315986,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8728,
      "step": 14002
    },
    {
      "epoch": 3.2755233306046074,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 14003
    },
    {
      "epoch": 3.2757572213776167,
      "grad_norm": 8.1875,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 14004
    },
    {
      "epoch": 3.2759911121506256,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6599,
      "step": 14005
    },
    {
      "epoch": 3.276225002923635,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 14006
    },
    {
      "epoch": 3.2764588936966437,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 14007
    },
    {
      "epoch": 3.2766927844696525,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 14008
    },
    {
      "epoch": 3.276926675242662,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 14009
    },
    {
      "epoch": 3.2771605660156706,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 14010
    },
    {
      "epoch": 3.27739445678868,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 14011
    },
    {
      "epoch": 3.2776283475616887,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7543,
      "step": 14012
    },
    {
      "epoch": 3.2778622383346976,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 14013
    },
    {
      "epoch": 3.278096129107707,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 14014
    },
    {
      "epoch": 3.2783300198807157,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 14015
    },
    {
      "epoch": 3.2785639106537245,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6695,
      "step": 14016
    },
    {
      "epoch": 3.278797801426734,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 14017
    },
    {
      "epoch": 3.2790316921997427,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4932,
      "step": 14018
    },
    {
      "epoch": 3.2792655829727515,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 14019
    },
    {
      "epoch": 3.279499473745761,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 14020
    },
    {
      "epoch": 3.2797333645187696,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 14021
    },
    {
      "epoch": 3.279967255291779,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 14022
    },
    {
      "epoch": 3.2802011460647877,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6366,
      "step": 14023
    },
    {
      "epoch": 3.2804350368377966,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 14024
    },
    {
      "epoch": 3.280668927610806,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.624,
      "step": 14025
    },
    {
      "epoch": 3.2809028183838147,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 14026
    },
    {
      "epoch": 3.281136709156824,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 14027
    },
    {
      "epoch": 3.281370599929833,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8196,
      "step": 14028
    },
    {
      "epoch": 3.2816044907028417,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 14029
    },
    {
      "epoch": 3.281838381475851,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 14030
    },
    {
      "epoch": 3.28207227224886,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.631,
      "step": 14031
    },
    {
      "epoch": 3.2823061630218686,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5334,
      "step": 14032
    },
    {
      "epoch": 3.282540053794878,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6763,
      "step": 14033
    },
    {
      "epoch": 3.2827739445678867,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 14034
    },
    {
      "epoch": 3.2830078353408956,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5987,
      "step": 14035
    },
    {
      "epoch": 3.283241726113905,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5987,
      "step": 14036
    },
    {
      "epoch": 3.2834756168869137,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 14037
    },
    {
      "epoch": 3.283709507659923,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 14038
    },
    {
      "epoch": 3.283943398432932,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 14039
    },
    {
      "epoch": 3.2841772892059407,
      "grad_norm": 8.4375,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 14040
    },
    {
      "epoch": 3.28441117997895,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 14041
    },
    {
      "epoch": 3.2846450707519588,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 14042
    },
    {
      "epoch": 3.284878961524968,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 14043
    },
    {
      "epoch": 3.285112852297977,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 14044
    },
    {
      "epoch": 3.2853467430709857,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 14045
    },
    {
      "epoch": 3.285580633843995,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 14046
    },
    {
      "epoch": 3.285814524617004,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 14047
    },
    {
      "epoch": 3.2860484153900127,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 14048
    },
    {
      "epoch": 3.286282306163022,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 14049
    },
    {
      "epoch": 3.286516196936031,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 14050
    },
    {
      "epoch": 3.2867500877090396,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 14051
    },
    {
      "epoch": 3.286983978482049,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 14052
    },
    {
      "epoch": 3.2872178692550578,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.594,
      "step": 14053
    },
    {
      "epoch": 3.287451760028067,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 14054
    },
    {
      "epoch": 3.287685650801076,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 14055
    },
    {
      "epoch": 3.2879195415740847,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 14056
    },
    {
      "epoch": 3.288153432347094,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.679,
      "step": 14057
    },
    {
      "epoch": 3.288387323120103,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 14058
    },
    {
      "epoch": 3.288621213893112,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 14059
    },
    {
      "epoch": 3.288855104666121,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 14060
    },
    {
      "epoch": 3.28908899543913,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 14061
    },
    {
      "epoch": 3.289322886212139,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 14062
    },
    {
      "epoch": 3.289556776985148,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8027,
      "step": 14063
    },
    {
      "epoch": 3.2897906677581568,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 14064
    },
    {
      "epoch": 3.290024558531166,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 14065
    },
    {
      "epoch": 3.290258449304175,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.652,
      "step": 14066
    },
    {
      "epoch": 3.2904923400771837,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 14067
    },
    {
      "epoch": 3.290726230850193,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 14068
    },
    {
      "epoch": 3.290960121623202,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 14069
    },
    {
      "epoch": 3.291194012396211,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 14070
    },
    {
      "epoch": 3.29142790316922,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 14071
    },
    {
      "epoch": 3.291661793942229,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.752,
      "step": 14072
    },
    {
      "epoch": 3.291895684715238,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 14073
    },
    {
      "epoch": 3.292129575488247,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 14074
    },
    {
      "epoch": 3.292363466261256,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 14075
    },
    {
      "epoch": 3.292597357034265,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 14076
    },
    {
      "epoch": 3.292831247807274,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 14077
    },
    {
      "epoch": 3.293065138580283,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 14078
    },
    {
      "epoch": 3.293299029353292,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 14079
    },
    {
      "epoch": 3.293532920126301,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 14080
    },
    {
      "epoch": 3.29376681089931,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 14081
    },
    {
      "epoch": 3.294000701672319,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 14082
    },
    {
      "epoch": 3.294234592445328,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 14083
    },
    {
      "epoch": 3.294468483218337,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.717,
      "step": 14084
    },
    {
      "epoch": 3.294702373991346,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 14085
    },
    {
      "epoch": 3.294936264764355,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5493,
      "step": 14086
    },
    {
      "epoch": 3.295170155537364,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 14087
    },
    {
      "epoch": 3.295404046310373,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 14088
    },
    {
      "epoch": 3.295637937083382,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 14089
    },
    {
      "epoch": 3.295871827856391,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6626,
      "step": 14090
    },
    {
      "epoch": 3.2961057186294003,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 14091
    },
    {
      "epoch": 3.296339609402409,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 14092
    },
    {
      "epoch": 3.296573500175418,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7187,
      "step": 14093
    },
    {
      "epoch": 3.2968073909484272,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 14094
    },
    {
      "epoch": 3.297041281721436,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6272,
      "step": 14095
    },
    {
      "epoch": 3.297275172494445,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 14096
    },
    {
      "epoch": 3.297509063267454,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 14097
    },
    {
      "epoch": 3.297742954040463,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7728,
      "step": 14098
    },
    {
      "epoch": 3.297976844813472,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 14099
    },
    {
      "epoch": 3.298210735586481,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 14100
    },
    {
      "epoch": 3.298210735586481,
      "eval_runtime": 4.601,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 14100
    },
    {
      "epoch": 3.29844462635949,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 14101
    },
    {
      "epoch": 3.2986785171324993,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 14102
    },
    {
      "epoch": 3.298912407905508,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 14103
    },
    {
      "epoch": 3.299146298678517,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6627,
      "step": 14104
    },
    {
      "epoch": 3.2993801894515262,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 14105
    },
    {
      "epoch": 3.299614080224535,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 14106
    },
    {
      "epoch": 3.2998479709975443,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6294,
      "step": 14107
    },
    {
      "epoch": 3.300081861770553,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6504,
      "step": 14108
    },
    {
      "epoch": 3.300315752543562,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 14109
    },
    {
      "epoch": 3.3005496433165713,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 14110
    },
    {
      "epoch": 3.30078353408958,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 14111
    },
    {
      "epoch": 3.301017424862589,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 14112
    },
    {
      "epoch": 3.3012513156355983,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 14113
    },
    {
      "epoch": 3.301485206408607,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6572,
      "step": 14114
    },
    {
      "epoch": 3.301719097181616,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 14115
    },
    {
      "epoch": 3.301952987954625,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 14116
    },
    {
      "epoch": 3.302186878727634,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.4884,
      "step": 14117
    },
    {
      "epoch": 3.3024207695006433,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 14118
    },
    {
      "epoch": 3.302654660273652,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 14119
    },
    {
      "epoch": 3.302888551046661,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 14120
    },
    {
      "epoch": 3.3031224418196703,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6778,
      "step": 14121
    },
    {
      "epoch": 3.303356332592679,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 14122
    },
    {
      "epoch": 3.3035902233656884,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 14123
    },
    {
      "epoch": 3.3038241141386973,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 14124
    },
    {
      "epoch": 3.304058004911706,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 14125
    },
    {
      "epoch": 3.3042918956847154,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 14126
    },
    {
      "epoch": 3.304525786457724,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 14127
    },
    {
      "epoch": 3.3047596772307335,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 14128
    },
    {
      "epoch": 3.3049935680037423,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 14129
    },
    {
      "epoch": 3.305227458776751,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 14130
    },
    {
      "epoch": 3.3054613495497605,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 14131
    },
    {
      "epoch": 3.3056952403227693,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 14132
    },
    {
      "epoch": 3.305929131095778,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0902,
      "step": 14133
    },
    {
      "epoch": 3.3061630218687874,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 14134
    },
    {
      "epoch": 3.3063969126417962,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 14135
    },
    {
      "epoch": 3.306630803414805,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 14136
    },
    {
      "epoch": 3.3068646941878144,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6429,
      "step": 14137
    },
    {
      "epoch": 3.307098584960823,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3659,
      "step": 14138
    },
    {
      "epoch": 3.3073324757338325,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 14139
    },
    {
      "epoch": 3.3075663665068413,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7082,
      "step": 14140
    },
    {
      "epoch": 3.30780025727985,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 14141
    },
    {
      "epoch": 3.3080341480528594,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 14142
    },
    {
      "epoch": 3.3082680388258683,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 14143
    },
    {
      "epoch": 3.3085019295988776,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 14144
    },
    {
      "epoch": 3.3087358203718864,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 14145
    },
    {
      "epoch": 3.3089697111448952,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 14146
    },
    {
      "epoch": 3.3092036019179045,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3927,
      "step": 14147
    },
    {
      "epoch": 3.3094374926909134,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 14148
    },
    {
      "epoch": 3.309671383463922,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 14149
    },
    {
      "epoch": 3.3099052742369315,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 14150
    },
    {
      "epoch": 3.3101391650099403,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.682,
      "step": 14151
    },
    {
      "epoch": 3.310373055782949,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 14152
    },
    {
      "epoch": 3.3106069465559584,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 14153
    },
    {
      "epoch": 3.3108408373289673,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4426,
      "step": 14154
    },
    {
      "epoch": 3.3110747281019766,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 14155
    },
    {
      "epoch": 3.3113086188749854,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 14156
    },
    {
      "epoch": 3.3115425096479942,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8988,
      "step": 14157
    },
    {
      "epoch": 3.3117764004210035,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6556,
      "step": 14158
    },
    {
      "epoch": 3.3120102911940124,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 14159
    },
    {
      "epoch": 3.3122441819670216,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 14160
    },
    {
      "epoch": 3.3124780727400305,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 14161
    },
    {
      "epoch": 3.3127119635130393,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 14162
    },
    {
      "epoch": 3.3129458542860486,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 14163
    },
    {
      "epoch": 3.3131797450590574,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4959,
      "step": 14164
    },
    {
      "epoch": 3.3134136358320663,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 14165
    },
    {
      "epoch": 3.3136475266050756,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 14166
    },
    {
      "epoch": 3.3138814173780844,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 14167
    },
    {
      "epoch": 3.3141153081510932,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5779,
      "step": 14168
    },
    {
      "epoch": 3.3143491989241025,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 14169
    },
    {
      "epoch": 3.3145830896971114,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.647,
      "step": 14170
    },
    {
      "epoch": 3.3148169804701206,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 14171
    },
    {
      "epoch": 3.3150508712431295,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 14172
    },
    {
      "epoch": 3.3152847620161383,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 14173
    },
    {
      "epoch": 3.3155186527891476,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 14174
    },
    {
      "epoch": 3.3157525435621564,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 14175
    },
    {
      "epoch": 3.3159864343351657,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 14176
    },
    {
      "epoch": 3.3162203251081745,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7866,
      "step": 14177
    },
    {
      "epoch": 3.3164542158811834,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 14178
    },
    {
      "epoch": 3.3166881066541927,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 14179
    },
    {
      "epoch": 3.3169219974272015,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 14180
    },
    {
      "epoch": 3.3171558882002103,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 14181
    },
    {
      "epoch": 3.3173897789732196,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7056,
      "step": 14182
    },
    {
      "epoch": 3.3176236697462285,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 14183
    },
    {
      "epoch": 3.3178575605192373,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 14184
    },
    {
      "epoch": 3.3180914512922466,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 14185
    },
    {
      "epoch": 3.3183253420652554,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 14186
    },
    {
      "epoch": 3.3185592328382647,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6059,
      "step": 14187
    },
    {
      "epoch": 3.3187931236112735,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9314,
      "step": 14188
    },
    {
      "epoch": 3.3190270143842824,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 14189
    },
    {
      "epoch": 3.3192609051572917,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8385,
      "step": 14190
    },
    {
      "epoch": 3.3194947959303005,
      "grad_norm": 8.5,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 14191
    },
    {
      "epoch": 3.31972868670331,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 14192
    },
    {
      "epoch": 3.3199625774763186,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 14193
    },
    {
      "epoch": 3.3201964682493275,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 14194
    },
    {
      "epoch": 3.3204303590223367,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.4903,
      "step": 14195
    },
    {
      "epoch": 3.3206642497953456,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 14196
    },
    {
      "epoch": 3.3208981405683544,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 14197
    },
    {
      "epoch": 3.3211320313413637,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 14198
    },
    {
      "epoch": 3.3213659221143725,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 14199
    },
    {
      "epoch": 3.3215998128873814,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 14200
    },
    {
      "epoch": 3.3215998128873814,
      "eval_runtime": 4.6037,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 14200
    },
    {
      "epoch": 3.3218337036603907,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 14201
    },
    {
      "epoch": 3.3220675944333995,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3462,
      "step": 14202
    },
    {
      "epoch": 3.3223014852064088,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 14203
    },
    {
      "epoch": 3.3225353759794176,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8688,
      "step": 14204
    },
    {
      "epoch": 3.3227692667524265,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 14205
    },
    {
      "epoch": 3.3230031575254357,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 14206
    },
    {
      "epoch": 3.3232370482984446,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7998,
      "step": 14207
    },
    {
      "epoch": 3.323470939071454,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 14208
    },
    {
      "epoch": 3.3237048298444627,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 14209
    },
    {
      "epoch": 3.3239387206174715,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 14210
    },
    {
      "epoch": 3.324172611390481,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 14211
    },
    {
      "epoch": 3.3244065021634897,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 14212
    },
    {
      "epoch": 3.3246403929364985,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 14213
    },
    {
      "epoch": 3.3248742837095078,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 14214
    },
    {
      "epoch": 3.3251081744825166,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 14215
    },
    {
      "epoch": 3.3253420652555254,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 14216
    },
    {
      "epoch": 3.3255759560285347,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7175,
      "step": 14217
    },
    {
      "epoch": 3.3258098468015436,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 14218
    },
    {
      "epoch": 3.326043737574553,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 14219
    },
    {
      "epoch": 3.3262776283475617,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 14220
    },
    {
      "epoch": 3.3265115191205705,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 14221
    },
    {
      "epoch": 3.32674540989358,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4048,
      "step": 14222
    },
    {
      "epoch": 3.3269793006665886,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 14223
    },
    {
      "epoch": 3.327213191439598,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 14224
    },
    {
      "epoch": 3.3274470822126068,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6642,
      "step": 14225
    },
    {
      "epoch": 3.3276809729856156,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 14226
    },
    {
      "epoch": 3.327914863758625,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 14227
    },
    {
      "epoch": 3.3281487545316337,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 14228
    },
    {
      "epoch": 3.3283826453046426,
      "grad_norm": 7.03125,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 14229
    },
    {
      "epoch": 3.328616536077652,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5583,
      "step": 14230
    },
    {
      "epoch": 3.3288504268506607,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 14231
    },
    {
      "epoch": 3.3290843176236695,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 14232
    },
    {
      "epoch": 3.329318208396679,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 14233
    },
    {
      "epoch": 3.3295520991696876,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 14234
    },
    {
      "epoch": 3.329785989942697,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 14235
    },
    {
      "epoch": 3.3300198807157058,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 14236
    },
    {
      "epoch": 3.3302537714887146,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2304,
      "step": 14237
    },
    {
      "epoch": 3.330487662261724,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 14238
    },
    {
      "epoch": 3.3307215530347327,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 14239
    },
    {
      "epoch": 3.330955443807742,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 14240
    },
    {
      "epoch": 3.331189334580751,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 14241
    },
    {
      "epoch": 3.3314232253537597,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 14242
    },
    {
      "epoch": 3.331657116126769,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 14243
    },
    {
      "epoch": 3.331891006899778,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 14244
    },
    {
      "epoch": 3.3321248976727866,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2816,
      "step": 14245
    },
    {
      "epoch": 3.332358788445796,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1872,
      "step": 14246
    },
    {
      "epoch": 3.3325926792188048,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 14247
    },
    {
      "epoch": 3.3328265699918136,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 14248
    },
    {
      "epoch": 3.333060460764823,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 14249
    },
    {
      "epoch": 3.3332943515378317,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 14250
    },
    {
      "epoch": 3.333528242310841,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6651,
      "step": 14251
    },
    {
      "epoch": 3.33376213308385,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5253,
      "step": 14252
    },
    {
      "epoch": 3.3339960238568587,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6242,
      "step": 14253
    },
    {
      "epoch": 3.334229914629868,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 14254
    },
    {
      "epoch": 3.334463805402877,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 14255
    },
    {
      "epoch": 3.334697696175886,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2638,
      "step": 14256
    },
    {
      "epoch": 3.334931586948895,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6585,
      "step": 14257
    },
    {
      "epoch": 3.3351654777219037,
      "grad_norm": 7.21875,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 14258
    },
    {
      "epoch": 3.335399368494913,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 14259
    },
    {
      "epoch": 3.335633259267922,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 14260
    },
    {
      "epoch": 3.3358671500409307,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 14261
    },
    {
      "epoch": 3.33610104081394,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 14262
    },
    {
      "epoch": 3.336334931586949,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 14263
    },
    {
      "epoch": 3.3365688223599577,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 14264
    },
    {
      "epoch": 3.336802713132967,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5667,
      "step": 14265
    },
    {
      "epoch": 3.337036603905976,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 14266
    },
    {
      "epoch": 3.337270494678985,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 14267
    },
    {
      "epoch": 3.337504385451994,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 14268
    },
    {
      "epoch": 3.3377382762250027,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 14269
    },
    {
      "epoch": 3.337972166998012,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 14270
    },
    {
      "epoch": 3.338206057771021,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7416,
      "step": 14271
    },
    {
      "epoch": 3.33843994854403,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3393,
      "step": 14272
    },
    {
      "epoch": 3.338673839317039,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 14273
    },
    {
      "epoch": 3.338907730090048,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 14274
    },
    {
      "epoch": 3.339141620863057,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 14275
    },
    {
      "epoch": 3.339375511636066,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 14276
    },
    {
      "epoch": 3.339609402409075,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 14277
    },
    {
      "epoch": 3.339843293182084,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 14278
    },
    {
      "epoch": 3.340077183955093,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 14279
    },
    {
      "epoch": 3.3403110747281017,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 14280
    },
    {
      "epoch": 3.340544965501111,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 14281
    },
    {
      "epoch": 3.34077885627412,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 14282
    },
    {
      "epoch": 3.341012747047129,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 14283
    },
    {
      "epoch": 3.341246637820138,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 14284
    },
    {
      "epoch": 3.341480528593147,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 14285
    },
    {
      "epoch": 3.341714419366156,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 14286
    },
    {
      "epoch": 3.341948310139165,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 14287
    },
    {
      "epoch": 3.342182200912174,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 14288
    },
    {
      "epoch": 3.342416091685183,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6187,
      "step": 14289
    },
    {
      "epoch": 3.342649982458192,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 14290
    },
    {
      "epoch": 3.342883873231201,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5996,
      "step": 14291
    },
    {
      "epoch": 3.34311776400421,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 14292
    },
    {
      "epoch": 3.343351654777219,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 14293
    },
    {
      "epoch": 3.343585545550228,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 14294
    },
    {
      "epoch": 3.343819436323237,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 14295
    },
    {
      "epoch": 3.344053327096246,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 14296
    },
    {
      "epoch": 3.344287217869255,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7059,
      "step": 14297
    },
    {
      "epoch": 3.344521108642264,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 14298
    },
    {
      "epoch": 3.344754999415273,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 14299
    },
    {
      "epoch": 3.344988890188282,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 14300
    },
    {
      "epoch": 3.344988890188282,
      "eval_runtime": 4.628,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 14300
    },
    {
      "epoch": 3.345222780961291,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 14301
    },
    {
      "epoch": 3.3454566717343,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 14302
    },
    {
      "epoch": 3.345690562507309,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8165,
      "step": 14303
    },
    {
      "epoch": 3.3459244532803183,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7232,
      "step": 14304
    },
    {
      "epoch": 3.346158344053327,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 14305
    },
    {
      "epoch": 3.346392234826336,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 14306
    },
    {
      "epoch": 3.3466261255993452,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 14307
    },
    {
      "epoch": 3.346860016372354,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.629,
      "step": 14308
    },
    {
      "epoch": 3.347093907145363,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 14309
    },
    {
      "epoch": 3.347327797918372,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0967,
      "step": 14310
    },
    {
      "epoch": 3.347561688691381,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 14311
    },
    {
      "epoch": 3.34779557946439,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 14312
    },
    {
      "epoch": 3.348029470237399,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 14313
    },
    {
      "epoch": 3.348263361010408,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 14314
    },
    {
      "epoch": 3.3484972517834173,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5206,
      "step": 14315
    },
    {
      "epoch": 3.348731142556426,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 14316
    },
    {
      "epoch": 3.348965033329435,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6151,
      "step": 14317
    },
    {
      "epoch": 3.3491989241024442,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 14318
    },
    {
      "epoch": 3.349432814875453,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 14319
    },
    {
      "epoch": 3.3496667056484624,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.4057,
      "step": 14320
    },
    {
      "epoch": 3.349900596421471,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 14321
    },
    {
      "epoch": 3.35013448719448,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6232,
      "step": 14322
    },
    {
      "epoch": 3.3503683779674893,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 14323
    },
    {
      "epoch": 3.350602268740498,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 14324
    },
    {
      "epoch": 3.3508361595135074,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 14325
    },
    {
      "epoch": 3.3510700502865163,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 14326
    },
    {
      "epoch": 3.351303941059525,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 14327
    },
    {
      "epoch": 3.3515378318325344,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 14328
    },
    {
      "epoch": 3.3517717226055432,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 14329
    },
    {
      "epoch": 3.352005613378552,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7339,
      "step": 14330
    },
    {
      "epoch": 3.3522395041515614,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 14331
    },
    {
      "epoch": 3.35247339492457,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 14332
    },
    {
      "epoch": 3.352707285697579,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 14333
    },
    {
      "epoch": 3.3529411764705883,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 14334
    },
    {
      "epoch": 3.353175067243597,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 14335
    },
    {
      "epoch": 3.3534089580166064,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 14336
    },
    {
      "epoch": 3.3536428487896153,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 14337
    },
    {
      "epoch": 3.353876739562624,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5245,
      "step": 14338
    },
    {
      "epoch": 3.3541106303356334,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 14339
    },
    {
      "epoch": 3.3543445211086422,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 14340
    },
    {
      "epoch": 3.3545784118816515,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5565,
      "step": 14341
    },
    {
      "epoch": 3.3548123026546603,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 14342
    },
    {
      "epoch": 3.355046193427669,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 14343
    },
    {
      "epoch": 3.3552800842006785,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 14344
    },
    {
      "epoch": 3.3555139749736873,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 14345
    },
    {
      "epoch": 3.355747865746696,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 14346
    },
    {
      "epoch": 3.3559817565197054,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6725,
      "step": 14347
    },
    {
      "epoch": 3.3562156472927143,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 14348
    },
    {
      "epoch": 3.356449538065723,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 14349
    },
    {
      "epoch": 3.3566834288387324,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 14350
    },
    {
      "epoch": 3.3569173196117412,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 14351
    },
    {
      "epoch": 3.3571512103847505,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 14352
    },
    {
      "epoch": 3.3573851011577593,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5978,
      "step": 14353
    },
    {
      "epoch": 3.357618991930768,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5342,
      "step": 14354
    },
    {
      "epoch": 3.3578528827037775,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 14355
    },
    {
      "epoch": 3.3580867734767863,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 14356
    },
    {
      "epoch": 3.3583206642497956,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 14357
    },
    {
      "epoch": 3.3585545550228044,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 14358
    },
    {
      "epoch": 3.3587884457958133,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 14359
    },
    {
      "epoch": 3.3590223365688225,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 14360
    },
    {
      "epoch": 3.3592562273418314,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 14361
    },
    {
      "epoch": 3.35949011811484,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7733,
      "step": 14362
    },
    {
      "epoch": 3.3597240088878495,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 14363
    },
    {
      "epoch": 3.3599578996608583,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 14364
    },
    {
      "epoch": 3.360191790433867,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 14365
    },
    {
      "epoch": 3.3604256812068765,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 14366
    },
    {
      "epoch": 3.3606595719798853,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 14367
    },
    {
      "epoch": 3.3608934627528946,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8801,
      "step": 14368
    },
    {
      "epoch": 3.3611273535259034,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 14369
    },
    {
      "epoch": 3.3613612442989123,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 14370
    },
    {
      "epoch": 3.3615951350719215,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8415,
      "step": 14371
    },
    {
      "epoch": 3.3618290258449304,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 14372
    },
    {
      "epoch": 3.3620629166179397,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5922,
      "step": 14373
    },
    {
      "epoch": 3.3622968073909485,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 14374
    },
    {
      "epoch": 3.3625306981639573,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7518,
      "step": 14375
    },
    {
      "epoch": 3.3627645889369666,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 14376
    },
    {
      "epoch": 3.3629984797099755,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 14377
    },
    {
      "epoch": 3.3632323704829843,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5249,
      "step": 14378
    },
    {
      "epoch": 3.3634662612559936,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 14379
    },
    {
      "epoch": 3.3637001520290024,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 14380
    },
    {
      "epoch": 3.3639340428020112,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 14381
    },
    {
      "epoch": 3.3641679335750205,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 14382
    },
    {
      "epoch": 3.3644018243480294,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 14383
    },
    {
      "epoch": 3.3646357151210387,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 14384
    },
    {
      "epoch": 3.3648696058940475,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 14385
    },
    {
      "epoch": 3.3651034966670563,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 14386
    },
    {
      "epoch": 3.3653373874400656,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 14387
    },
    {
      "epoch": 3.3655712782130744,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 14388
    },
    {
      "epoch": 3.3658051689860837,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 14389
    },
    {
      "epoch": 3.3660390597590926,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 14390
    },
    {
      "epoch": 3.3662729505321014,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 14391
    },
    {
      "epoch": 3.3665068413051107,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 14392
    },
    {
      "epoch": 3.3667407320781195,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 14393
    },
    {
      "epoch": 3.3669746228511284,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6084,
      "step": 14394
    },
    {
      "epoch": 3.3672085136241376,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0059,
      "step": 14395
    },
    {
      "epoch": 3.3674424043971465,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 14396
    },
    {
      "epoch": 3.3676762951701553,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.3144,
      "step": 14397
    },
    {
      "epoch": 3.3679101859431646,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 14398
    },
    {
      "epoch": 3.3681440767161734,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7936,
      "step": 14399
    },
    {
      "epoch": 3.3683779674891827,
      "grad_norm": 9.875,
      "learning_rate": 3e-05,
      "loss": 1.6966,
      "step": 14400
    },
    {
      "epoch": 3.3683779674891827,
      "eval_runtime": 4.6086,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 14400
    },
    {
      "epoch": 3.3686118582621916,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 14401
    },
    {
      "epoch": 3.3688457490352004,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 14402
    },
    {
      "epoch": 3.3690796398082097,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 14403
    },
    {
      "epoch": 3.3693135305812185,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6477,
      "step": 14404
    },
    {
      "epoch": 3.369547421354228,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 14405
    },
    {
      "epoch": 3.3697813121272366,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 14406
    },
    {
      "epoch": 3.3700152029002455,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6758,
      "step": 14407
    },
    {
      "epoch": 3.3702490936732548,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 14408
    },
    {
      "epoch": 3.3704829844462636,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7321,
      "step": 14409
    },
    {
      "epoch": 3.3707168752192724,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 14410
    },
    {
      "epoch": 3.3709507659922817,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7692,
      "step": 14411
    },
    {
      "epoch": 3.3711846567652906,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 14412
    },
    {
      "epoch": 3.3714185475382994,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 14413
    },
    {
      "epoch": 3.3716524383113087,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 14414
    },
    {
      "epoch": 3.3718863290843175,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 14415
    },
    {
      "epoch": 3.372120219857327,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 14416
    },
    {
      "epoch": 3.3723541106303356,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 14417
    },
    {
      "epoch": 3.3725880014033445,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 14418
    },
    {
      "epoch": 3.3728218921763538,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 14419
    },
    {
      "epoch": 3.3730557829493626,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7103,
      "step": 14420
    },
    {
      "epoch": 3.373289673722372,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 14421
    },
    {
      "epoch": 3.3735235644953807,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5032,
      "step": 14422
    },
    {
      "epoch": 3.3737574552683895,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 14423
    },
    {
      "epoch": 3.373991346041399,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 14424
    },
    {
      "epoch": 3.3742252368144077,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 14425
    },
    {
      "epoch": 3.3744591275874165,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3309,
      "step": 14426
    },
    {
      "epoch": 3.374693018360426,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 14427
    },
    {
      "epoch": 3.3749269091334346,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 14428
    },
    {
      "epoch": 3.3751607999064435,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7504,
      "step": 14429
    },
    {
      "epoch": 3.3753946906794527,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.307,
      "step": 14430
    },
    {
      "epoch": 3.3756285814524616,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.739,
      "step": 14431
    },
    {
      "epoch": 3.375862472225471,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 14432
    },
    {
      "epoch": 3.3760963629984797,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 14433
    },
    {
      "epoch": 3.3763302537714885,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 14434
    },
    {
      "epoch": 3.376564144544498,
      "grad_norm": 7.3125,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 14435
    },
    {
      "epoch": 3.3767980353175067,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 14436
    },
    {
      "epoch": 3.377031926090516,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 14437
    },
    {
      "epoch": 3.377265816863525,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 14438
    },
    {
      "epoch": 3.3774997076365336,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 14439
    },
    {
      "epoch": 3.377733598409543,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 14440
    },
    {
      "epoch": 3.3779674891825517,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 14441
    },
    {
      "epoch": 3.3782013799555606,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 14442
    },
    {
      "epoch": 3.37843527072857,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 14443
    },
    {
      "epoch": 3.3786691615015787,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 14444
    },
    {
      "epoch": 3.3789030522745875,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 14445
    },
    {
      "epoch": 3.379136943047597,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7663,
      "step": 14446
    },
    {
      "epoch": 3.3793708338206057,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 14447
    },
    {
      "epoch": 3.379604724593615,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 14448
    },
    {
      "epoch": 3.3798386153666238,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 14449
    },
    {
      "epoch": 3.3800725061396326,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 14450
    },
    {
      "epoch": 3.380306396912642,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 14451
    },
    {
      "epoch": 3.3805402876856507,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 14452
    },
    {
      "epoch": 3.38077417845866,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 14453
    },
    {
      "epoch": 3.381008069231669,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8617,
      "step": 14454
    },
    {
      "epoch": 3.3812419600046777,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6404,
      "step": 14455
    },
    {
      "epoch": 3.381475850777687,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 14456
    },
    {
      "epoch": 3.381709741550696,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6219,
      "step": 14457
    },
    {
      "epoch": 3.3819436323237047,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7947,
      "step": 14458
    },
    {
      "epoch": 3.382177523096714,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2187,
      "step": 14459
    },
    {
      "epoch": 3.3824114138697228,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 14460
    },
    {
      "epoch": 3.3826453046427316,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 14461
    },
    {
      "epoch": 3.382879195415741,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 14462
    },
    {
      "epoch": 3.3831130861887497,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7907,
      "step": 14463
    },
    {
      "epoch": 3.383346976961759,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 14464
    },
    {
      "epoch": 3.383580867734768,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 14465
    },
    {
      "epoch": 3.3838147585077767,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 14466
    },
    {
      "epoch": 3.384048649280786,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2241,
      "step": 14467
    },
    {
      "epoch": 3.384282540053795,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 14468
    },
    {
      "epoch": 3.384516430826804,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 14469
    },
    {
      "epoch": 3.384750321599813,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.758,
      "step": 14470
    },
    {
      "epoch": 3.3849842123728218,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 14471
    },
    {
      "epoch": 3.385218103145831,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 14472
    },
    {
      "epoch": 3.38545199391884,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 14473
    },
    {
      "epoch": 3.3856858846918487,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 14474
    },
    {
      "epoch": 3.385919775464858,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 14475
    },
    {
      "epoch": 3.386153666237867,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 14476
    },
    {
      "epoch": 3.3863875570108757,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 14477
    },
    {
      "epoch": 3.386621447783885,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 14478
    },
    {
      "epoch": 3.386855338556894,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 14479
    },
    {
      "epoch": 3.387089229329903,
      "grad_norm": 9.6875,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 14480
    },
    {
      "epoch": 3.387323120102912,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 14481
    },
    {
      "epoch": 3.3875570108759208,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 14482
    },
    {
      "epoch": 3.38779090164893,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 14483
    },
    {
      "epoch": 3.388024792421939,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 14484
    },
    {
      "epoch": 3.388258683194948,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 14485
    },
    {
      "epoch": 3.388492573967957,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 14486
    },
    {
      "epoch": 3.388726464740966,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 14487
    },
    {
      "epoch": 3.388960355513975,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 14488
    },
    {
      "epoch": 3.389194246286984,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 14489
    },
    {
      "epoch": 3.389428137059993,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 14490
    },
    {
      "epoch": 3.389662027833002,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 14491
    },
    {
      "epoch": 3.389895918606011,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6276,
      "step": 14492
    },
    {
      "epoch": 3.3901298093790198,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 14493
    },
    {
      "epoch": 3.390363700152029,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9724,
      "step": 14494
    },
    {
      "epoch": 3.390597590925038,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 14495
    },
    {
      "epoch": 3.390831481698047,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6784,
      "step": 14496
    },
    {
      "epoch": 3.391065372471056,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 14497
    },
    {
      "epoch": 3.391299263244065,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5499,
      "step": 14498
    },
    {
      "epoch": 3.391533154017074,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 14499
    },
    {
      "epoch": 3.391767044790083,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 14500
    },
    {
      "epoch": 3.391767044790083,
      "eval_runtime": 4.6199,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 14500
    },
    {
      "epoch": 3.3920009355630922,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 14501
    },
    {
      "epoch": 3.392234826336101,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 14502
    },
    {
      "epoch": 3.39246871710911,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 14503
    },
    {
      "epoch": 3.392702607882119,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 14504
    },
    {
      "epoch": 3.392936498655128,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 14505
    },
    {
      "epoch": 3.393170389428137,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6527,
      "step": 14506
    },
    {
      "epoch": 3.393404280201146,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 14507
    },
    {
      "epoch": 3.393638170974155,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 14508
    },
    {
      "epoch": 3.393872061747164,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 14509
    },
    {
      "epoch": 3.394105952520173,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 14510
    },
    {
      "epoch": 3.394339843293182,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 14511
    },
    {
      "epoch": 3.3945737340661912,
      "grad_norm": 17.75,
      "learning_rate": 3e-05,
      "loss": 2.3697,
      "step": 14512
    },
    {
      "epoch": 3.3948076248392,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 14513
    },
    {
      "epoch": 3.395041515612209,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 14514
    },
    {
      "epoch": 3.395275406385218,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5562,
      "step": 14515
    },
    {
      "epoch": 3.395509297158227,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 14516
    },
    {
      "epoch": 3.3957431879312363,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 14517
    },
    {
      "epoch": 3.395977078704245,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5821,
      "step": 14518
    },
    {
      "epoch": 3.396210969477254,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 14519
    },
    {
      "epoch": 3.3964448602502633,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 14520
    },
    {
      "epoch": 3.396678751023272,
      "grad_norm": 8.125,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 14521
    },
    {
      "epoch": 3.3969126417962814,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 14522
    },
    {
      "epoch": 3.3971465325692902,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 14523
    },
    {
      "epoch": 3.397380423342299,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6316,
      "step": 14524
    },
    {
      "epoch": 3.3976143141153083,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4755,
      "step": 14525
    },
    {
      "epoch": 3.397848204888317,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 14526
    },
    {
      "epoch": 3.398082095661326,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7449,
      "step": 14527
    },
    {
      "epoch": 3.3983159864343353,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 14528
    },
    {
      "epoch": 3.398549877207344,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 14529
    },
    {
      "epoch": 3.398783767980353,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8038,
      "step": 14530
    },
    {
      "epoch": 3.3990176587533623,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 14531
    },
    {
      "epoch": 3.399251549526371,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 14532
    },
    {
      "epoch": 3.3994854402993804,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 14533
    },
    {
      "epoch": 3.399719331072389,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 14534
    },
    {
      "epoch": 3.399953221845398,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 14535
    },
    {
      "epoch": 3.4001871126184073,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6269,
      "step": 14536
    },
    {
      "epoch": 3.400421003391416,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 14537
    },
    {
      "epoch": 3.4006548941644255,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 14538
    },
    {
      "epoch": 3.4008887849374343,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7377,
      "step": 14539
    },
    {
      "epoch": 3.401122675710443,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7333,
      "step": 14540
    },
    {
      "epoch": 3.4013565664834524,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 14541
    },
    {
      "epoch": 3.4015904572564613,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 14542
    },
    {
      "epoch": 3.40182434802947,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1965,
      "step": 14543
    },
    {
      "epoch": 3.4020582388024794,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 14544
    },
    {
      "epoch": 3.402292129575488,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 14545
    },
    {
      "epoch": 3.402526020348497,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 14546
    },
    {
      "epoch": 3.4027599111215063,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 14547
    },
    {
      "epoch": 3.402993801894515,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 14548
    },
    {
      "epoch": 3.4032276926675245,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 14549
    },
    {
      "epoch": 3.4034615834405333,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 14550
    },
    {
      "epoch": 3.403695474213542,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 14551
    },
    {
      "epoch": 3.4039293649865514,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.615,
      "step": 14552
    },
    {
      "epoch": 3.4041632557595602,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 14553
    },
    {
      "epoch": 3.4043971465325695,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7051,
      "step": 14554
    },
    {
      "epoch": 3.4046310373055784,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 14555
    },
    {
      "epoch": 3.404864928078587,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 14556
    },
    {
      "epoch": 3.4050988188515965,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6278,
      "step": 14557
    },
    {
      "epoch": 3.4053327096246053,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.651,
      "step": 14558
    },
    {
      "epoch": 3.405566600397614,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 14559
    },
    {
      "epoch": 3.4058004911706234,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1949,
      "step": 14560
    },
    {
      "epoch": 3.4060343819436323,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 14561
    },
    {
      "epoch": 3.406268272716641,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 14562
    },
    {
      "epoch": 3.4065021634896504,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 14563
    },
    {
      "epoch": 3.4067360542626592,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 14564
    },
    {
      "epoch": 3.4069699450356685,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 14565
    },
    {
      "epoch": 3.4072038358086774,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5794,
      "step": 14566
    },
    {
      "epoch": 3.407437726581686,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 14567
    },
    {
      "epoch": 3.4076716173546955,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 14568
    },
    {
      "epoch": 3.4079055081277043,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 14569
    },
    {
      "epoch": 3.4081393989007136,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 14570
    },
    {
      "epoch": 3.4083732896737224,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 14571
    },
    {
      "epoch": 3.4086071804467313,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 14572
    },
    {
      "epoch": 3.4088410712197406,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 14573
    },
    {
      "epoch": 3.4090749619927494,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 14574
    },
    {
      "epoch": 3.4093088527657582,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 14575
    },
    {
      "epoch": 3.4095427435387675,
      "grad_norm": 12.3125,
      "learning_rate": 3e-05,
      "loss": 2.3469,
      "step": 14576
    },
    {
      "epoch": 3.4097766343117764,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 14577
    },
    {
      "epoch": 3.410010525084785,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6879,
      "step": 14578
    },
    {
      "epoch": 3.4102444158577945,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 14579
    },
    {
      "epoch": 3.4104783066308033,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 14580
    },
    {
      "epoch": 3.4107121974038126,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 14581
    },
    {
      "epoch": 3.4109460881768214,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 14582
    },
    {
      "epoch": 3.4111799789498303,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 14583
    },
    {
      "epoch": 3.4114138697228396,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 14584
    },
    {
      "epoch": 3.4116477604958484,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 14585
    },
    {
      "epoch": 3.4118816512688577,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 14586
    },
    {
      "epoch": 3.4121155420418665,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 14587
    },
    {
      "epoch": 3.4123494328148753,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 14588
    },
    {
      "epoch": 3.4125833235878846,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 14589
    },
    {
      "epoch": 3.4128172143608935,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 14590
    },
    {
      "epoch": 3.4130511051339023,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8487,
      "step": 14591
    },
    {
      "epoch": 3.4132849959069116,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 14592
    },
    {
      "epoch": 3.4135188866799204,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5995,
      "step": 14593
    },
    {
      "epoch": 3.4137527774529293,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6748,
      "step": 14594
    },
    {
      "epoch": 3.4139866682259385,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 14595
    },
    {
      "epoch": 3.4142205589989474,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 14596
    },
    {
      "epoch": 3.4144544497719567,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 14597
    },
    {
      "epoch": 3.4146883405449655,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 14598
    },
    {
      "epoch": 3.4149222313179743,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 14599
    },
    {
      "epoch": 3.4151561220909836,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 14600
    },
    {
      "epoch": 3.4151561220909836,
      "eval_runtime": 4.6276,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 14600
    },
    {
      "epoch": 3.4153900128639925,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 14601
    },
    {
      "epoch": 3.4156239036370017,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7536,
      "step": 14602
    },
    {
      "epoch": 3.4158577944100106,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 14603
    },
    {
      "epoch": 3.4160916851830194,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 14604
    },
    {
      "epoch": 3.4163255759560287,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 14605
    },
    {
      "epoch": 3.4165594667290375,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5818,
      "step": 14606
    },
    {
      "epoch": 3.4167933575020464,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 14607
    },
    {
      "epoch": 3.4170272482750557,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 14608
    },
    {
      "epoch": 3.4172611390480645,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 14609
    },
    {
      "epoch": 3.4174950298210733,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 14610
    },
    {
      "epoch": 3.4177289205940826,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 14611
    },
    {
      "epoch": 3.4179628113670915,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7051,
      "step": 14612
    },
    {
      "epoch": 3.4181967021401007,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 14613
    },
    {
      "epoch": 3.4184305929131096,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7215,
      "step": 14614
    },
    {
      "epoch": 3.4186644836861184,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 14615
    },
    {
      "epoch": 3.4188983744591277,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 14616
    },
    {
      "epoch": 3.4191322652321365,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 14617
    },
    {
      "epoch": 3.419366156005146,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 14618
    },
    {
      "epoch": 3.4196000467781547,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 14619
    },
    {
      "epoch": 3.4198339375511635,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 14620
    },
    {
      "epoch": 3.4200678283241728,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 14621
    },
    {
      "epoch": 3.4203017190971816,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 14622
    },
    {
      "epoch": 3.4205356098701905,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7094,
      "step": 14623
    },
    {
      "epoch": 3.4207695006431997,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 14624
    },
    {
      "epoch": 3.4210033914162086,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 14625
    },
    {
      "epoch": 3.4212372821892174,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5253,
      "step": 14626
    },
    {
      "epoch": 3.4214711729622267,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 14627
    },
    {
      "epoch": 3.4217050637352355,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 14628
    },
    {
      "epoch": 3.421938954508245,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 14629
    },
    {
      "epoch": 3.4221728452812537,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5977,
      "step": 14630
    },
    {
      "epoch": 3.4224067360542625,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 14631
    },
    {
      "epoch": 3.4226406268272718,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5826,
      "step": 14632
    },
    {
      "epoch": 3.4228745176002806,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 14633
    },
    {
      "epoch": 3.42310840837329,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 14634
    },
    {
      "epoch": 3.4233422991462987,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 14635
    },
    {
      "epoch": 3.4235761899193076,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 14636
    },
    {
      "epoch": 3.423810080692317,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2451,
      "step": 14637
    },
    {
      "epoch": 3.4240439714653257,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 14638
    },
    {
      "epoch": 3.4242778622383345,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 14639
    },
    {
      "epoch": 3.424511753011344,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 14640
    },
    {
      "epoch": 3.4247456437843526,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 14641
    },
    {
      "epoch": 3.4249795345573615,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 14642
    },
    {
      "epoch": 3.4252134253303708,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 14643
    },
    {
      "epoch": 3.4254473161033796,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 14644
    },
    {
      "epoch": 3.425681206876389,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 14645
    },
    {
      "epoch": 3.4259150976493977,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 14646
    },
    {
      "epoch": 3.4261489884224066,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 14647
    },
    {
      "epoch": 3.426382879195416,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6864,
      "step": 14648
    },
    {
      "epoch": 3.4266167699684247,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 14649
    },
    {
      "epoch": 3.426850660741434,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7327,
      "step": 14650
    },
    {
      "epoch": 3.427084551514443,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7251,
      "step": 14651
    },
    {
      "epoch": 3.4273184422874516,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 14652
    },
    {
      "epoch": 3.427552333060461,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 14653
    },
    {
      "epoch": 3.4277862238334698,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 14654
    },
    {
      "epoch": 3.4280201146064786,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 14655
    },
    {
      "epoch": 3.428254005379488,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 14656
    },
    {
      "epoch": 3.4284878961524967,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 14657
    },
    {
      "epoch": 3.4287217869255056,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 14658
    },
    {
      "epoch": 3.428955677698515,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 14659
    },
    {
      "epoch": 3.4291895684715237,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 14660
    },
    {
      "epoch": 3.429423459244533,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 14661
    },
    {
      "epoch": 3.429657350017542,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 14662
    },
    {
      "epoch": 3.4298912407905506,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 14663
    },
    {
      "epoch": 3.43012513156356,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 14664
    },
    {
      "epoch": 3.4303590223365688,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 14665
    },
    {
      "epoch": 3.430592913109578,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 14666
    },
    {
      "epoch": 3.430826803882587,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 14667
    },
    {
      "epoch": 3.4310606946555957,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.382,
      "step": 14668
    },
    {
      "epoch": 3.431294585428605,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 14669
    },
    {
      "epoch": 3.431528476201614,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 14670
    },
    {
      "epoch": 3.4317623669746227,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 14671
    },
    {
      "epoch": 3.431996257747632,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 14672
    },
    {
      "epoch": 3.432230148520641,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 14673
    },
    {
      "epoch": 3.4324640392936496,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 14674
    },
    {
      "epoch": 3.432697930066659,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 14675
    },
    {
      "epoch": 3.4329318208396677,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 14676
    },
    {
      "epoch": 3.433165711612677,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 14677
    },
    {
      "epoch": 3.433399602385686,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 14678
    },
    {
      "epoch": 3.4336334931586947,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8131,
      "step": 14679
    },
    {
      "epoch": 3.433867383931704,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 14680
    },
    {
      "epoch": 3.434101274704713,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.7261,
      "step": 14681
    },
    {
      "epoch": 3.434335165477722,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 14682
    },
    {
      "epoch": 3.434569056250731,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 14683
    },
    {
      "epoch": 3.43480294702374,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 14684
    },
    {
      "epoch": 3.435036837796749,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 14685
    },
    {
      "epoch": 3.435270728569758,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 14686
    },
    {
      "epoch": 3.4355046193427667,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 14687
    },
    {
      "epoch": 3.435738510115776,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 14688
    },
    {
      "epoch": 3.435972400888785,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5554,
      "step": 14689
    },
    {
      "epoch": 3.4362062916617937,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6827,
      "step": 14690
    },
    {
      "epoch": 3.436440182434803,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.4542,
      "step": 14691
    },
    {
      "epoch": 3.436674073207812,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7664,
      "step": 14692
    },
    {
      "epoch": 3.436907963980821,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 14693
    },
    {
      "epoch": 3.43714185475383,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 14694
    },
    {
      "epoch": 3.4373757455268388,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 14695
    },
    {
      "epoch": 3.437609636299848,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6907,
      "step": 14696
    },
    {
      "epoch": 3.437843527072857,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 14697
    },
    {
      "epoch": 3.438077417845866,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6786,
      "step": 14698
    },
    {
      "epoch": 3.438311308618875,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 14699
    },
    {
      "epoch": 3.438545199391884,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 14700
    },
    {
      "epoch": 3.438545199391884,
      "eval_runtime": 4.5941,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 14700
    },
    {
      "epoch": 3.438779090164893,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2433,
      "step": 14701
    },
    {
      "epoch": 3.439012980937902,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 14702
    },
    {
      "epoch": 3.439246871710911,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 14703
    },
    {
      "epoch": 3.43948076248392,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6408,
      "step": 14704
    },
    {
      "epoch": 3.439714653256929,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 14705
    },
    {
      "epoch": 3.4399485440299378,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 14706
    },
    {
      "epoch": 3.440182434802947,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 14707
    },
    {
      "epoch": 3.440416325575956,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6777,
      "step": 14708
    },
    {
      "epoch": 3.440650216348965,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6482,
      "step": 14709
    },
    {
      "epoch": 3.440884107121974,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7954,
      "step": 14710
    },
    {
      "epoch": 3.441117997894983,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 14711
    },
    {
      "epoch": 3.441351888667992,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5641,
      "step": 14712
    },
    {
      "epoch": 3.441585779441001,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 14713
    },
    {
      "epoch": 3.4418196702140103,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 14714
    },
    {
      "epoch": 3.442053560987019,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 14715
    },
    {
      "epoch": 3.442287451760028,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.263,
      "step": 14716
    },
    {
      "epoch": 3.442521342533037,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 14717
    },
    {
      "epoch": 3.442755233306046,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 14718
    },
    {
      "epoch": 3.4429891240790553,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 14719
    },
    {
      "epoch": 3.443223014852064,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 14720
    },
    {
      "epoch": 3.443456905625073,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 14721
    },
    {
      "epoch": 3.4436907963980823,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 14722
    },
    {
      "epoch": 3.443924687171091,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6142,
      "step": 14723
    },
    {
      "epoch": 3.4441585779441,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 14724
    },
    {
      "epoch": 3.4443924687171092,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 14725
    },
    {
      "epoch": 3.444626359490118,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 14726
    },
    {
      "epoch": 3.444860250263127,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 14727
    },
    {
      "epoch": 3.445094141036136,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.621,
      "step": 14728
    },
    {
      "epoch": 3.445328031809145,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 14729
    },
    {
      "epoch": 3.4455619225821543,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.196,
      "step": 14730
    },
    {
      "epoch": 3.445795813355163,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 14731
    },
    {
      "epoch": 3.446029704128172,
      "grad_norm": 16.875,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 14732
    },
    {
      "epoch": 3.4462635949011813,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 14733
    },
    {
      "epoch": 3.44649748567419,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 14734
    },
    {
      "epoch": 3.4467313764471994,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5657,
      "step": 14735
    },
    {
      "epoch": 3.4469652672202082,
      "grad_norm": 8.9375,
      "learning_rate": 3e-05,
      "loss": 2.6285,
      "step": 14736
    },
    {
      "epoch": 3.447199157993217,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6907,
      "step": 14737
    },
    {
      "epoch": 3.4474330487662264,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5412,
      "step": 14738
    },
    {
      "epoch": 3.447666939539235,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 14739
    },
    {
      "epoch": 3.447900830312244,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 14740
    },
    {
      "epoch": 3.4481347210852533,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 14741
    },
    {
      "epoch": 3.448368611858262,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 14742
    },
    {
      "epoch": 3.448602502631271,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 14743
    },
    {
      "epoch": 3.4488363934042803,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 14744
    },
    {
      "epoch": 3.449070284177289,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 14745
    },
    {
      "epoch": 3.4493041749502984,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 14746
    },
    {
      "epoch": 3.4495380657233072,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 14747
    },
    {
      "epoch": 3.449771956496316,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 14748
    },
    {
      "epoch": 3.4500058472693254,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6089,
      "step": 14749
    },
    {
      "epoch": 3.450239738042334,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 14750
    },
    {
      "epoch": 3.4504736288153435,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 14751
    },
    {
      "epoch": 3.4507075195883523,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 14752
    },
    {
      "epoch": 3.450941410361361,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6312,
      "step": 14753
    },
    {
      "epoch": 3.4511753011343704,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 14754
    },
    {
      "epoch": 3.4514091919073793,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 14755
    },
    {
      "epoch": 3.451643082680388,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 14756
    },
    {
      "epoch": 3.4518769734533974,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7444,
      "step": 14757
    },
    {
      "epoch": 3.4521108642264062,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 14758
    },
    {
      "epoch": 3.452344754999415,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 14759
    },
    {
      "epoch": 3.4525786457724243,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 14760
    },
    {
      "epoch": 3.452812536545433,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8473,
      "step": 14761
    },
    {
      "epoch": 3.4530464273184425,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 14762
    },
    {
      "epoch": 3.4532803180914513,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 14763
    },
    {
      "epoch": 3.45351420886446,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7079,
      "step": 14764
    },
    {
      "epoch": 3.4537480996374694,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6827,
      "step": 14765
    },
    {
      "epoch": 3.4539819904104783,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 14766
    },
    {
      "epoch": 3.4542158811834875,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8434,
      "step": 14767
    },
    {
      "epoch": 3.4544497719564964,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5705,
      "step": 14768
    },
    {
      "epoch": 3.4546836627295052,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 14769
    },
    {
      "epoch": 3.4549175535025145,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 14770
    },
    {
      "epoch": 3.4551514442755233,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 14771
    },
    {
      "epoch": 3.455385335048532,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 14772
    },
    {
      "epoch": 3.4556192258215415,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 14773
    },
    {
      "epoch": 3.4558531165945503,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 14774
    },
    {
      "epoch": 3.456087007367559,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 14775
    },
    {
      "epoch": 3.4563208981405684,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 14776
    },
    {
      "epoch": 3.4565547889135773,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 14777
    },
    {
      "epoch": 3.4567886796865865,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 14778
    },
    {
      "epoch": 3.4570225704595954,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.3934,
      "step": 14779
    },
    {
      "epoch": 3.457256461232604,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 14780
    },
    {
      "epoch": 3.4574903520056135,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 14781
    },
    {
      "epoch": 3.4577242427786223,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 14782
    },
    {
      "epoch": 3.4579581335516316,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4035,
      "step": 14783
    },
    {
      "epoch": 3.4581920243246405,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 14784
    },
    {
      "epoch": 3.4584259150976493,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6463,
      "step": 14785
    },
    {
      "epoch": 3.4586598058706586,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 14786
    },
    {
      "epoch": 3.4588936966436674,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 14787
    },
    {
      "epoch": 3.4591275874166763,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6806,
      "step": 14788
    },
    {
      "epoch": 3.4593614781896855,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5976,
      "step": 14789
    },
    {
      "epoch": 3.4595953689626944,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 14790
    },
    {
      "epoch": 3.459829259735703,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5658,
      "step": 14791
    },
    {
      "epoch": 3.4600631505087125,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 14792
    },
    {
      "epoch": 3.4602970412817213,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 14793
    },
    {
      "epoch": 3.4605309320547306,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.542,
      "step": 14794
    },
    {
      "epoch": 3.4607648228277395,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 14795
    },
    {
      "epoch": 3.4609987136007483,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 14796
    },
    {
      "epoch": 3.4612326043737576,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 14797
    },
    {
      "epoch": 3.4614664951467664,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 14798
    },
    {
      "epoch": 3.4617003859197757,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 14799
    },
    {
      "epoch": 3.4619342766927845,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 14800
    },
    {
      "epoch": 3.4619342766927845,
      "eval_runtime": 4.6722,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 14800
    },
    {
      "epoch": 3.4621681674657934,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 14801
    },
    {
      "epoch": 3.4624020582388026,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 14802
    },
    {
      "epoch": 3.4626359490118115,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 14803
    },
    {
      "epoch": 3.4628698397848203,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7244,
      "step": 14804
    },
    {
      "epoch": 3.4631037305578296,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 14805
    },
    {
      "epoch": 3.4633376213308384,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 14806
    },
    {
      "epoch": 3.4635715121038473,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 14807
    },
    {
      "epoch": 3.4638054028768566,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 14808
    },
    {
      "epoch": 3.4640392936498654,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 14809
    },
    {
      "epoch": 3.4642731844228747,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 14810
    },
    {
      "epoch": 3.4645070751958835,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 14811
    },
    {
      "epoch": 3.4647409659688924,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1679,
      "step": 14812
    },
    {
      "epoch": 3.4649748567419016,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 14813
    },
    {
      "epoch": 3.4652087475149105,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 14814
    },
    {
      "epoch": 3.4654426382879198,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 14815
    },
    {
      "epoch": 3.4656765290609286,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 14816
    },
    {
      "epoch": 3.4659104198339374,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 14817
    },
    {
      "epoch": 3.4661443106069467,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7146,
      "step": 14818
    },
    {
      "epoch": 3.4663782013799556,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 14819
    },
    {
      "epoch": 3.4666120921529644,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 14820
    },
    {
      "epoch": 3.4668459829259737,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5902,
      "step": 14821
    },
    {
      "epoch": 3.4670798736989825,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7033,
      "step": 14822
    },
    {
      "epoch": 3.4673137644719914,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 14823
    },
    {
      "epoch": 3.4675476552450006,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 14824
    },
    {
      "epoch": 3.4677815460180095,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 14825
    },
    {
      "epoch": 3.4680154367910188,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 14826
    },
    {
      "epoch": 3.4682493275640276,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 14827
    },
    {
      "epoch": 3.4684832183370364,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.4938,
      "step": 14828
    },
    {
      "epoch": 3.4687171091100457,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 14829
    },
    {
      "epoch": 3.4689509998830546,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5513,
      "step": 14830
    },
    {
      "epoch": 3.469184890656064,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 14831
    },
    {
      "epoch": 3.4694187814290727,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 14832
    },
    {
      "epoch": 3.4696526722020815,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.575,
      "step": 14833
    },
    {
      "epoch": 3.469886562975091,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 14834
    },
    {
      "epoch": 3.4701204537480996,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 14835
    },
    {
      "epoch": 3.4703543445211085,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5032,
      "step": 14836
    },
    {
      "epoch": 3.4705882352941178,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 14837
    },
    {
      "epoch": 3.4708221260671266,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 14838
    },
    {
      "epoch": 3.4710560168401354,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8547,
      "step": 14839
    },
    {
      "epoch": 3.4712899076131447,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 14840
    },
    {
      "epoch": 3.4715237983861535,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6693,
      "step": 14841
    },
    {
      "epoch": 3.471757689159163,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 14842
    },
    {
      "epoch": 3.4719915799321717,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 14843
    },
    {
      "epoch": 3.4722254707051805,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 14844
    },
    {
      "epoch": 3.47245936147819,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 14845
    },
    {
      "epoch": 3.4726932522511986,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 14846
    },
    {
      "epoch": 3.472927143024208,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 14847
    },
    {
      "epoch": 3.4731610337972167,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 14848
    },
    {
      "epoch": 3.4733949245702256,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 14849
    },
    {
      "epoch": 3.473628815343235,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 14850
    },
    {
      "epoch": 3.4738627061162437,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 14851
    },
    {
      "epoch": 3.4740965968892525,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 14852
    },
    {
      "epoch": 3.474330487662262,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 14853
    },
    {
      "epoch": 3.4745643784352707,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 14854
    },
    {
      "epoch": 3.4747982692082795,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 14855
    },
    {
      "epoch": 3.475032159981289,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7448,
      "step": 14856
    },
    {
      "epoch": 3.4752660507542976,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 14857
    },
    {
      "epoch": 3.475499941527307,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 14858
    },
    {
      "epoch": 3.4757338323003157,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 14859
    },
    {
      "epoch": 3.4759677230733246,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5273,
      "step": 14860
    },
    {
      "epoch": 3.476201613846334,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 14861
    },
    {
      "epoch": 3.4764355046193427,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.608,
      "step": 14862
    },
    {
      "epoch": 3.476669395392352,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 14863
    },
    {
      "epoch": 3.476903286165361,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 14864
    },
    {
      "epoch": 3.4771371769383697,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 14865
    },
    {
      "epoch": 3.477371067711379,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 14866
    },
    {
      "epoch": 3.4776049584843878,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 14867
    },
    {
      "epoch": 3.4778388492573966,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 14868
    },
    {
      "epoch": 3.478072740030406,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 14869
    },
    {
      "epoch": 3.4783066308034147,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 14870
    },
    {
      "epoch": 3.4785405215764236,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 14871
    },
    {
      "epoch": 3.478774412349433,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 14872
    },
    {
      "epoch": 3.4790083031224417,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7771,
      "step": 14873
    },
    {
      "epoch": 3.479242193895451,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7858,
      "step": 14874
    },
    {
      "epoch": 3.47947608466846,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 14875
    },
    {
      "epoch": 3.4797099754414687,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 14876
    },
    {
      "epoch": 3.479943866214478,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 14877
    },
    {
      "epoch": 3.4801777569874868,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 14878
    },
    {
      "epoch": 3.480411647760496,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 14879
    },
    {
      "epoch": 3.480645538533505,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7039,
      "step": 14880
    },
    {
      "epoch": 3.4808794293065137,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7273,
      "step": 14881
    },
    {
      "epoch": 3.481113320079523,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 14882
    },
    {
      "epoch": 3.481347210852532,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 14883
    },
    {
      "epoch": 3.4815811016255407,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 14884
    },
    {
      "epoch": 3.48181499239855,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 14885
    },
    {
      "epoch": 3.482048883171559,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 14886
    },
    {
      "epoch": 3.4822827739445676,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 14887
    },
    {
      "epoch": 3.482516664717577,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5825,
      "step": 14888
    },
    {
      "epoch": 3.4827505554905858,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3176,
      "step": 14889
    },
    {
      "epoch": 3.482984446263595,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.677,
      "step": 14890
    },
    {
      "epoch": 3.483218337036604,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2355,
      "step": 14891
    },
    {
      "epoch": 3.4834522278096127,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7579,
      "step": 14892
    },
    {
      "epoch": 3.483686118582622,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 14893
    },
    {
      "epoch": 3.483920009355631,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5567,
      "step": 14894
    },
    {
      "epoch": 3.48415390012864,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 14895
    },
    {
      "epoch": 3.484387790901649,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 14896
    },
    {
      "epoch": 3.484621681674658,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 14897
    },
    {
      "epoch": 3.484855572447667,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 14898
    },
    {
      "epoch": 3.485089463220676,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 14899
    },
    {
      "epoch": 3.4853233539936848,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 14900
    },
    {
      "epoch": 3.4853233539936848,
      "eval_runtime": 4.589,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 14900
    },
    {
      "epoch": 3.485557244766694,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 14901
    },
    {
      "epoch": 3.485791135539703,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 14902
    },
    {
      "epoch": 3.4860250263127117,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 14903
    },
    {
      "epoch": 3.486258917085721,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 14904
    },
    {
      "epoch": 3.48649280785873,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 14905
    },
    {
      "epoch": 3.486726698631739,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3766,
      "step": 14906
    },
    {
      "epoch": 3.486960589404748,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 14907
    },
    {
      "epoch": 3.487194480177757,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 14908
    },
    {
      "epoch": 3.487428370950766,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 14909
    },
    {
      "epoch": 3.487662261723775,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 14910
    },
    {
      "epoch": 3.487896152496784,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 14911
    },
    {
      "epoch": 3.488130043269793,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 14912
    },
    {
      "epoch": 3.488363934042802,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 14913
    },
    {
      "epoch": 3.488597824815811,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 14914
    },
    {
      "epoch": 3.48883171558882,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 14915
    },
    {
      "epoch": 3.4890656063618293,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5844,
      "step": 14916
    },
    {
      "epoch": 3.489299497134838,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 14917
    },
    {
      "epoch": 3.489533387907847,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 14918
    },
    {
      "epoch": 3.4897672786808562,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8305,
      "step": 14919
    },
    {
      "epoch": 3.490001169453865,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6401,
      "step": 14920
    },
    {
      "epoch": 3.490235060226874,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 14921
    },
    {
      "epoch": 3.490468950999883,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 14922
    },
    {
      "epoch": 3.490702841772892,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 14923
    },
    {
      "epoch": 3.490936732545901,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 14924
    },
    {
      "epoch": 3.49117062331891,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 14925
    },
    {
      "epoch": 3.491404514091919,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 14926
    },
    {
      "epoch": 3.4916384048649283,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 14927
    },
    {
      "epoch": 3.491872295637937,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 14928
    },
    {
      "epoch": 3.492106186410946,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1728,
      "step": 14929
    },
    {
      "epoch": 3.4923400771839552,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5321,
      "step": 14930
    },
    {
      "epoch": 3.492573967956964,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 14931
    },
    {
      "epoch": 3.4928078587299733,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 14932
    },
    {
      "epoch": 3.493041749502982,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 14933
    },
    {
      "epoch": 3.493275640275991,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 14934
    },
    {
      "epoch": 3.4935095310490003,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 14935
    },
    {
      "epoch": 3.493743421822009,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 14936
    },
    {
      "epoch": 3.493977312595018,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.784,
      "step": 14937
    },
    {
      "epoch": 3.4942112033680273,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7757,
      "step": 14938
    },
    {
      "epoch": 3.494445094141036,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 14939
    },
    {
      "epoch": 3.494678984914045,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 14940
    },
    {
      "epoch": 3.494912875687054,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5642,
      "step": 14941
    },
    {
      "epoch": 3.495146766460063,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 14942
    },
    {
      "epoch": 3.4953806572330723,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 14943
    },
    {
      "epoch": 3.495614548006081,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 14944
    },
    {
      "epoch": 3.49584843877909,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3453,
      "step": 14945
    },
    {
      "epoch": 3.4960823295520993,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.7285,
      "step": 14946
    },
    {
      "epoch": 3.496316220325108,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 14947
    },
    {
      "epoch": 3.4965501110981174,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 14948
    },
    {
      "epoch": 3.4967840018711263,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 14949
    },
    {
      "epoch": 3.497017892644135,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1341,
      "step": 14950
    },
    {
      "epoch": 3.4972517834171444,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 14951
    },
    {
      "epoch": 3.497485674190153,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 14952
    },
    {
      "epoch": 3.497719564963162,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 14953
    },
    {
      "epoch": 3.4979534557361713,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 14954
    },
    {
      "epoch": 3.49818734650918,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 14955
    },
    {
      "epoch": 3.498421237282189,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 14956
    },
    {
      "epoch": 3.4986551280551983,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 14957
    },
    {
      "epoch": 3.498889018828207,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 14958
    },
    {
      "epoch": 3.4991229096012164,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 14959
    },
    {
      "epoch": 3.4993568003742253,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6787,
      "step": 14960
    },
    {
      "epoch": 3.499590691147234,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 14961
    },
    {
      "epoch": 3.4998245819202434,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 14962
    },
    {
      "epoch": 3.500058472693252,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.7509,
      "step": 14963
    },
    {
      "epoch": 3.5002923634662615,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 14964
    },
    {
      "epoch": 3.5005262542392703,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 14965
    },
    {
      "epoch": 3.500760145012279,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 14966
    },
    {
      "epoch": 3.500994035785288,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 14967
    },
    {
      "epoch": 3.5012279265582973,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6752,
      "step": 14968
    },
    {
      "epoch": 3.501461817331306,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 14969
    },
    {
      "epoch": 3.5016957081043154,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7749,
      "step": 14970
    },
    {
      "epoch": 3.5019295988773242,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 14971
    },
    {
      "epoch": 3.502163489650333,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 14972
    },
    {
      "epoch": 3.5023973804233424,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 14973
    },
    {
      "epoch": 3.502631271196351,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 14974
    },
    {
      "epoch": 3.5028651619693605,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 14975
    },
    {
      "epoch": 3.5030990527423693,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6163,
      "step": 14976
    },
    {
      "epoch": 3.503332943515378,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 14977
    },
    {
      "epoch": 3.5035668342883874,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2156,
      "step": 14978
    },
    {
      "epoch": 3.5038007250613963,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 14979
    },
    {
      "epoch": 3.5040346158344056,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6635,
      "step": 14980
    },
    {
      "epoch": 3.5042685066074144,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7052,
      "step": 14981
    },
    {
      "epoch": 3.5045023973804232,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 14982
    },
    {
      "epoch": 3.504736288153432,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 14983
    },
    {
      "epoch": 3.5049701789264414,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 14984
    },
    {
      "epoch": 3.50520406969945,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 14985
    },
    {
      "epoch": 3.5054379604724595,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 14986
    },
    {
      "epoch": 3.5056718512454683,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 14987
    },
    {
      "epoch": 3.505905742018477,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 14988
    },
    {
      "epoch": 3.5061396327914864,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 14989
    },
    {
      "epoch": 3.5063735235644953,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 14990
    },
    {
      "epoch": 3.5066074143375046,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 14991
    },
    {
      "epoch": 3.5068413051105134,
      "grad_norm": 8.125,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 14992
    },
    {
      "epoch": 3.5070751958835222,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.648,
      "step": 14993
    },
    {
      "epoch": 3.5073090866565315,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 14994
    },
    {
      "epoch": 3.5075429774295404,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 14995
    },
    {
      "epoch": 3.5077768682025496,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6783,
      "step": 14996
    },
    {
      "epoch": 3.5080107589755585,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 14997
    },
    {
      "epoch": 3.5082446497485673,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 14998
    },
    {
      "epoch": 3.508478540521576,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 14999
    },
    {
      "epoch": 3.5087124312945854,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 15000
    },
    {
      "epoch": 3.5087124312945854,
      "eval_runtime": 4.5924,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 15000
    },
    {
      "epoch": 3.5089463220675943,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 15001
    },
    {
      "epoch": 3.5091802128406036,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 15002
    },
    {
      "epoch": 3.5094141036136124,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 15003
    },
    {
      "epoch": 3.5096479943866212,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 15004
    },
    {
      "epoch": 3.5098818851596305,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 15005
    },
    {
      "epoch": 3.5101157759326393,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 15006
    },
    {
      "epoch": 3.5103496667056486,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 15007
    },
    {
      "epoch": 3.5105835574786575,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 15008
    },
    {
      "epoch": 3.5108174482516663,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 15009
    },
    {
      "epoch": 3.5110513390246756,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 15010
    },
    {
      "epoch": 3.5112852297976844,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 15011
    },
    {
      "epoch": 3.5115191205706937,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 15012
    },
    {
      "epoch": 3.5117530113437025,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 15013
    },
    {
      "epoch": 3.5119869021167114,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 15014
    },
    {
      "epoch": 3.5122207928897207,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 15015
    },
    {
      "epoch": 3.5124546836627295,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 15016
    },
    {
      "epoch": 3.512688574435739,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 15017
    },
    {
      "epoch": 3.5129224652087476,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 15018
    },
    {
      "epoch": 3.5131563559817565,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 15019
    },
    {
      "epoch": 3.5133902467547653,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 15020
    },
    {
      "epoch": 3.5136241375277746,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 15021
    },
    {
      "epoch": 3.5138580283007834,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 15022
    },
    {
      "epoch": 3.5140919190737927,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6493,
      "step": 15023
    },
    {
      "epoch": 3.5143258098468015,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 15024
    },
    {
      "epoch": 3.5145597006198104,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5261,
      "step": 15025
    },
    {
      "epoch": 3.5147935913928197,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.78,
      "step": 15026
    },
    {
      "epoch": 3.5150274821658285,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 15027
    },
    {
      "epoch": 3.515261372938838,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7415,
      "step": 15028
    },
    {
      "epoch": 3.5154952637118466,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.579,
      "step": 15029
    },
    {
      "epoch": 3.5157291544848555,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 15030
    },
    {
      "epoch": 3.5159630452578647,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 15031
    },
    {
      "epoch": 3.5161969360308736,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 15032
    },
    {
      "epoch": 3.516430826803883,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 15033
    },
    {
      "epoch": 3.5166647175768917,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8205,
      "step": 15034
    },
    {
      "epoch": 3.5168986083499005,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 15035
    },
    {
      "epoch": 3.5171324991229094,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 15036
    },
    {
      "epoch": 3.5173663898959187,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 15037
    },
    {
      "epoch": 3.5176002806689275,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7838,
      "step": 15038
    },
    {
      "epoch": 3.5178341714419368,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 15039
    },
    {
      "epoch": 3.5180680622149456,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6326,
      "step": 15040
    },
    {
      "epoch": 3.5183019529879545,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2623,
      "step": 15041
    },
    {
      "epoch": 3.5185358437609637,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 15042
    },
    {
      "epoch": 3.5187697345339726,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7332,
      "step": 15043
    },
    {
      "epoch": 3.519003625306982,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 15044
    },
    {
      "epoch": 3.5192375160799907,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 15045
    },
    {
      "epoch": 3.5194714068529995,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6205,
      "step": 15046
    },
    {
      "epoch": 3.519705297626009,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.681,
      "step": 15047
    },
    {
      "epoch": 3.5199391883990176,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 15048
    },
    {
      "epoch": 3.520173079172027,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 15049
    },
    {
      "epoch": 3.5204069699450358,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 15050
    },
    {
      "epoch": 3.5206408607180446,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 15051
    },
    {
      "epoch": 3.5208747514910534,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 15052
    },
    {
      "epoch": 3.5211086422640627,
      "grad_norm": 11.3125,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 15053
    },
    {
      "epoch": 3.5213425330370716,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 15054
    },
    {
      "epoch": 3.521576423810081,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 15055
    },
    {
      "epoch": 3.5218103145830897,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 15056
    },
    {
      "epoch": 3.5220442053560985,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 15057
    },
    {
      "epoch": 3.522278096129108,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 15058
    },
    {
      "epoch": 3.5225119869021166,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 15059
    },
    {
      "epoch": 3.522745877675126,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 15060
    },
    {
      "epoch": 3.5229797684481348,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1605,
      "step": 15061
    },
    {
      "epoch": 3.5232136592211436,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 15062
    },
    {
      "epoch": 3.523447549994153,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 15063
    },
    {
      "epoch": 3.5236814407671617,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 15064
    },
    {
      "epoch": 3.523915331540171,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 15065
    },
    {
      "epoch": 3.52414922231318,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6464,
      "step": 15066
    },
    {
      "epoch": 3.5243831130861887,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6844,
      "step": 15067
    },
    {
      "epoch": 3.5246170038591975,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 15068
    },
    {
      "epoch": 3.524850894632207,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 15069
    },
    {
      "epoch": 3.5250847854052156,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 15070
    },
    {
      "epoch": 3.525318676178225,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 15071
    },
    {
      "epoch": 3.5255525669512338,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 15072
    },
    {
      "epoch": 3.5257864577242426,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7651,
      "step": 15073
    },
    {
      "epoch": 3.526020348497252,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 15074
    },
    {
      "epoch": 3.5262542392702607,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 15075
    },
    {
      "epoch": 3.52648813004327,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 15076
    },
    {
      "epoch": 3.526722020816279,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 15077
    },
    {
      "epoch": 3.5269559115892877,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 15078
    },
    {
      "epoch": 3.527189802362297,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 15079
    },
    {
      "epoch": 3.527423693135306,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 15080
    },
    {
      "epoch": 3.527657583908315,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 15081
    },
    {
      "epoch": 3.527891474681324,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 15082
    },
    {
      "epoch": 3.5281253654543328,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 15083
    },
    {
      "epoch": 3.5283592562273416,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8153,
      "step": 15084
    },
    {
      "epoch": 3.528593147000351,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 15085
    },
    {
      "epoch": 3.5288270377733597,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.5891,
      "step": 15086
    },
    {
      "epoch": 3.529060928546369,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 15087
    },
    {
      "epoch": 3.529294819319378,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 15088
    },
    {
      "epoch": 3.5295287100923867,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7439,
      "step": 15089
    },
    {
      "epoch": 3.529762600865396,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 15090
    },
    {
      "epoch": 3.529996491638405,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 15091
    },
    {
      "epoch": 3.530230382411414,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 15092
    },
    {
      "epoch": 3.530464273184423,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 15093
    },
    {
      "epoch": 3.5306981639574317,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 15094
    },
    {
      "epoch": 3.530932054730441,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 15095
    },
    {
      "epoch": 3.53116594550345,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8359,
      "step": 15096
    },
    {
      "epoch": 3.531399836276459,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 15097
    },
    {
      "epoch": 3.531633727049468,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 15098
    },
    {
      "epoch": 3.531867617822477,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 15099
    },
    {
      "epoch": 3.5321015085954857,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 15100
    },
    {
      "epoch": 3.5321015085954857,
      "eval_runtime": 4.6343,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 15100
    },
    {
      "epoch": 3.532335399368495,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 15101
    },
    {
      "epoch": 3.532569290141504,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 15102
    },
    {
      "epoch": 3.532803180914513,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.761,
      "step": 15103
    },
    {
      "epoch": 3.533037071687522,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 15104
    },
    {
      "epoch": 3.5332709624605307,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 15105
    },
    {
      "epoch": 3.53350485323354,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.7628,
      "step": 15106
    },
    {
      "epoch": 3.533738744006549,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 15107
    },
    {
      "epoch": 3.533972634779558,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 15108
    },
    {
      "epoch": 3.534206525552567,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 15109
    },
    {
      "epoch": 3.534440416325576,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 15110
    },
    {
      "epoch": 3.534674307098585,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8196,
      "step": 15111
    },
    {
      "epoch": 3.534908197871594,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.772,
      "step": 15112
    },
    {
      "epoch": 3.535142088644603,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 15113
    },
    {
      "epoch": 3.535375979417612,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 15114
    },
    {
      "epoch": 3.535609870190621,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6794,
      "step": 15115
    },
    {
      "epoch": 3.5358437609636297,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6341,
      "step": 15116
    },
    {
      "epoch": 3.536077651736639,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 15117
    },
    {
      "epoch": 3.536311542509648,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 15118
    },
    {
      "epoch": 3.536545433282657,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 15119
    },
    {
      "epoch": 3.536779324055666,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 15120
    },
    {
      "epoch": 3.537013214828675,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 15121
    },
    {
      "epoch": 3.537247105601684,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5337,
      "step": 15122
    },
    {
      "epoch": 3.537480996374693,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9478,
      "step": 15123
    },
    {
      "epoch": 3.537714887147702,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 15124
    },
    {
      "epoch": 3.537948777920711,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.579,
      "step": 15125
    },
    {
      "epoch": 3.53818266869372,
      "grad_norm": 9.625,
      "learning_rate": 3e-05,
      "loss": 2.2567,
      "step": 15126
    },
    {
      "epoch": 3.538416559466729,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 15127
    },
    {
      "epoch": 3.538650450239738,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 15128
    },
    {
      "epoch": 3.5388843410127473,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 15129
    },
    {
      "epoch": 3.539118231785756,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 15130
    },
    {
      "epoch": 3.539352122558765,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 15131
    },
    {
      "epoch": 3.539586013331774,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.8001,
      "step": 15132
    },
    {
      "epoch": 3.539819904104783,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.609,
      "step": 15133
    },
    {
      "epoch": 3.540053794877792,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 15134
    },
    {
      "epoch": 3.540287685650801,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 15135
    },
    {
      "epoch": 3.54052157642381,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 15136
    },
    {
      "epoch": 3.540755467196819,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6985,
      "step": 15137
    },
    {
      "epoch": 3.540989357969828,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 15138
    },
    {
      "epoch": 3.541223248742837,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5771,
      "step": 15139
    },
    {
      "epoch": 3.5414571395158463,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 15140
    },
    {
      "epoch": 3.541691030288855,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 15141
    },
    {
      "epoch": 3.541924921061864,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 15142
    },
    {
      "epoch": 3.5421588118348732,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 15143
    },
    {
      "epoch": 3.542392702607882,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 15144
    },
    {
      "epoch": 3.5426265933808914,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 15145
    },
    {
      "epoch": 3.5428604841539,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 15146
    },
    {
      "epoch": 3.543094374926909,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 15147
    },
    {
      "epoch": 3.543328265699918,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 15148
    },
    {
      "epoch": 3.543562156472927,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 15149
    },
    {
      "epoch": 3.543796047245936,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5394,
      "step": 15150
    },
    {
      "epoch": 3.5440299380189453,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 15151
    },
    {
      "epoch": 3.544263828791954,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5842,
      "step": 15152
    },
    {
      "epoch": 3.544497719564963,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 15153
    },
    {
      "epoch": 3.5447316103379722,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 15154
    },
    {
      "epoch": 3.544965501110981,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9314,
      "step": 15155
    },
    {
      "epoch": 3.5451993918839904,
      "grad_norm": 7.78125,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 15156
    },
    {
      "epoch": 3.545433282656999,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0503,
      "step": 15157
    },
    {
      "epoch": 3.545667173430008,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 15158
    },
    {
      "epoch": 3.5459010642030173,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 15159
    },
    {
      "epoch": 3.546134954976026,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 15160
    },
    {
      "epoch": 3.5463688457490354,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8945,
      "step": 15161
    },
    {
      "epoch": 3.5466027365220443,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 15162
    },
    {
      "epoch": 3.546836627295053,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 15163
    },
    {
      "epoch": 3.547070518068062,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 15164
    },
    {
      "epoch": 3.5473044088410712,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 15165
    },
    {
      "epoch": 3.54753829961408,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 15166
    },
    {
      "epoch": 3.5477721903870894,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 15167
    },
    {
      "epoch": 3.548006081160098,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 15168
    },
    {
      "epoch": 3.548239971933107,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 15169
    },
    {
      "epoch": 3.5484738627061163,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 15170
    },
    {
      "epoch": 3.548707753479125,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 15171
    },
    {
      "epoch": 3.5489416442521344,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6369,
      "step": 15172
    },
    {
      "epoch": 3.5491755350251433,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 15173
    },
    {
      "epoch": 3.549409425798152,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6163,
      "step": 15174
    },
    {
      "epoch": 3.5496433165711614,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 15175
    },
    {
      "epoch": 3.5498772073441702,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 15176
    },
    {
      "epoch": 3.5501110981171795,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1376,
      "step": 15177
    },
    {
      "epoch": 3.5503449888901883,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6182,
      "step": 15178
    },
    {
      "epoch": 3.550578879663197,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 15179
    },
    {
      "epoch": 3.550812770436206,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 15180
    },
    {
      "epoch": 3.5510466612092153,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 15181
    },
    {
      "epoch": 3.551280551982224,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 15182
    },
    {
      "epoch": 3.5515144427552334,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6208,
      "step": 15183
    },
    {
      "epoch": 3.5517483335282423,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9824,
      "step": 15184
    },
    {
      "epoch": 3.551982224301251,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 15185
    },
    {
      "epoch": 3.5522161150742604,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7969,
      "step": 15186
    },
    {
      "epoch": 3.552450005847269,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 15187
    },
    {
      "epoch": 3.5526838966202785,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.5033,
      "step": 15188
    },
    {
      "epoch": 3.5529177873932873,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 15189
    },
    {
      "epoch": 3.553151678166296,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 15190
    },
    {
      "epoch": 3.5533855689393055,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 15191
    },
    {
      "epoch": 3.5536194597123143,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 15192
    },
    {
      "epoch": 3.5538533504853236,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6655,
      "step": 15193
    },
    {
      "epoch": 3.5540872412583324,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 15194
    },
    {
      "epoch": 3.5543211320313413,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 15195
    },
    {
      "epoch": 3.55455502280435,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 15196
    },
    {
      "epoch": 3.5547889135773594,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 15197
    },
    {
      "epoch": 3.555022804350368,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 15198
    },
    {
      "epoch": 3.5552566951233775,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.688,
      "step": 15199
    },
    {
      "epoch": 3.5554905858963863,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 15200
    },
    {
      "epoch": 3.5554905858963863,
      "eval_runtime": 4.602,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 15200
    },
    {
      "epoch": 3.555724476669395,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 15201
    },
    {
      "epoch": 3.5559583674424045,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 15202
    },
    {
      "epoch": 3.5561922582154133,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5444,
      "step": 15203
    },
    {
      "epoch": 3.5564261489884226,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 15204
    },
    {
      "epoch": 3.5566600397614314,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 15205
    },
    {
      "epoch": 3.5568939305344403,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 15206
    },
    {
      "epoch": 3.5571278213074495,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 15207
    },
    {
      "epoch": 3.5573617120804584,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8135,
      "step": 15208
    },
    {
      "epoch": 3.5575956028534677,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 15209
    },
    {
      "epoch": 3.5578294936264765,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 15210
    },
    {
      "epoch": 3.5580633843994853,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 15211
    },
    {
      "epoch": 3.5582972751724946,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 15212
    },
    {
      "epoch": 3.5585311659455034,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 15213
    },
    {
      "epoch": 3.5587650567185127,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 15214
    },
    {
      "epoch": 3.5589989474915216,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 15215
    },
    {
      "epoch": 3.5592328382645304,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 15216
    },
    {
      "epoch": 3.5594667290375392,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 15217
    },
    {
      "epoch": 3.5597006198105485,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6611,
      "step": 15218
    },
    {
      "epoch": 3.5599345105835574,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 15219
    },
    {
      "epoch": 3.5601684013565666,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 15220
    },
    {
      "epoch": 3.5604022921295755,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7616,
      "step": 15221
    },
    {
      "epoch": 3.5606361829025843,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 15222
    },
    {
      "epoch": 3.5608700736755936,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 15223
    },
    {
      "epoch": 3.5611039644486024,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.7595,
      "step": 15224
    },
    {
      "epoch": 3.5613378552216117,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 15225
    },
    {
      "epoch": 3.5615717459946206,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 15226
    },
    {
      "epoch": 3.5618056367676294,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 15227
    },
    {
      "epoch": 3.5620395275406387,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 15228
    },
    {
      "epoch": 3.5622734183136475,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.261,
      "step": 15229
    },
    {
      "epoch": 3.562507309086657,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 15230
    },
    {
      "epoch": 3.5627411998596656,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 15231
    },
    {
      "epoch": 3.5629750906326745,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 15232
    },
    {
      "epoch": 3.5632089814056833,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 15233
    },
    {
      "epoch": 3.5634428721786926,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7207,
      "step": 15234
    },
    {
      "epoch": 3.5636767629517014,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 15235
    },
    {
      "epoch": 3.5639106537247107,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 15236
    },
    {
      "epoch": 3.5641445444977196,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6885,
      "step": 15237
    },
    {
      "epoch": 3.5643784352707284,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2789,
      "step": 15238
    },
    {
      "epoch": 3.5646123260437377,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 15239
    },
    {
      "epoch": 3.5648462168167465,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 15240
    },
    {
      "epoch": 3.565080107589756,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 15241
    },
    {
      "epoch": 3.5653139983627646,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 15242
    },
    {
      "epoch": 3.5655478891357735,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4326,
      "step": 15243
    },
    {
      "epoch": 3.5657817799087828,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6082,
      "step": 15244
    },
    {
      "epoch": 3.5660156706817916,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4932,
      "step": 15245
    },
    {
      "epoch": 3.566249561454801,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 15246
    },
    {
      "epoch": 3.5664834522278097,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 15247
    },
    {
      "epoch": 3.5667173430008186,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 15248
    },
    {
      "epoch": 3.5669512337738274,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 15249
    },
    {
      "epoch": 3.5671851245468367,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 15250
    },
    {
      "epoch": 3.5674190153198455,
      "grad_norm": 9.125,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 15251
    },
    {
      "epoch": 3.567652906092855,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 15252
    },
    {
      "epoch": 3.5678867968658636,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5426,
      "step": 15253
    },
    {
      "epoch": 3.5681206876388725,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7426,
      "step": 15254
    },
    {
      "epoch": 3.5683545784118818,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 15255
    },
    {
      "epoch": 3.5685884691848906,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7866,
      "step": 15256
    },
    {
      "epoch": 3.5688223599579,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 15257
    },
    {
      "epoch": 3.5690562507309087,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6622,
      "step": 15258
    },
    {
      "epoch": 3.5692901415039175,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 15259
    },
    {
      "epoch": 3.569524032276927,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 15260
    },
    {
      "epoch": 3.5697579230499357,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.7481,
      "step": 15261
    },
    {
      "epoch": 3.569991813822945,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 15262
    },
    {
      "epoch": 3.570225704595954,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 15263
    },
    {
      "epoch": 3.5704595953689626,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 15264
    },
    {
      "epoch": 3.5706934861419715,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7934,
      "step": 15265
    },
    {
      "epoch": 3.5709273769149807,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 15266
    },
    {
      "epoch": 3.5711612676879896,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 15267
    },
    {
      "epoch": 3.571395158460999,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 15268
    },
    {
      "epoch": 3.5716290492340077,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 15269
    },
    {
      "epoch": 3.5718629400070165,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 15270
    },
    {
      "epoch": 3.572096830780026,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7999,
      "step": 15271
    },
    {
      "epoch": 3.5723307215530347,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 15272
    },
    {
      "epoch": 3.572564612326044,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7526,
      "step": 15273
    },
    {
      "epoch": 3.572798503099053,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7295,
      "step": 15274
    },
    {
      "epoch": 3.5730323938720616,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 15275
    },
    {
      "epoch": 3.573266284645071,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.715,
      "step": 15276
    },
    {
      "epoch": 3.5735001754180797,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 15277
    },
    {
      "epoch": 3.573734066191089,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 15278
    },
    {
      "epoch": 3.573967956964098,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 15279
    },
    {
      "epoch": 3.5742018477371067,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 15280
    },
    {
      "epoch": 3.5744357385101155,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 15281
    },
    {
      "epoch": 3.574669629283125,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 15282
    },
    {
      "epoch": 3.5749035200561337,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 15283
    },
    {
      "epoch": 3.575137410829143,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 15284
    },
    {
      "epoch": 3.5753713016021518,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 15285
    },
    {
      "epoch": 3.5756051923751606,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 15286
    },
    {
      "epoch": 3.57583908314817,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 15287
    },
    {
      "epoch": 3.5760729739211787,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6316,
      "step": 15288
    },
    {
      "epoch": 3.576306864694188,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 15289
    },
    {
      "epoch": 3.576540755467197,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5138,
      "step": 15290
    },
    {
      "epoch": 3.5767746462402057,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7406,
      "step": 15291
    },
    {
      "epoch": 3.577008537013215,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 15292
    },
    {
      "epoch": 3.577242427786224,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 15293
    },
    {
      "epoch": 3.577476318559233,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 15294
    },
    {
      "epoch": 3.577710209332242,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 15295
    },
    {
      "epoch": 3.5779441001052508,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 15296
    },
    {
      "epoch": 3.5781779908782596,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 15297
    },
    {
      "epoch": 3.578411881651269,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5172,
      "step": 15298
    },
    {
      "epoch": 3.5786457724242777,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 15299
    },
    {
      "epoch": 3.578879663197287,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 15300
    },
    {
      "epoch": 3.578879663197287,
      "eval_runtime": 4.6214,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 15300
    },
    {
      "epoch": 3.579113553970296,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 15301
    },
    {
      "epoch": 3.5793474447433047,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 15302
    },
    {
      "epoch": 3.579581335516314,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 15303
    },
    {
      "epoch": 3.579815226289323,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.6667,
      "step": 15304
    },
    {
      "epoch": 3.580049117062332,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 15305
    },
    {
      "epoch": 3.580283007835341,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 15306
    },
    {
      "epoch": 3.5805168986083498,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 15307
    },
    {
      "epoch": 3.580750789381359,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 15308
    },
    {
      "epoch": 3.580984680154368,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 15309
    },
    {
      "epoch": 3.581218570927377,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 15310
    },
    {
      "epoch": 3.581452461700386,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8195,
      "step": 15311
    },
    {
      "epoch": 3.581686352473395,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 15312
    },
    {
      "epoch": 3.5819202432464037,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 15313
    },
    {
      "epoch": 3.582154134019413,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7229,
      "step": 15314
    },
    {
      "epoch": 3.582388024792422,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 15315
    },
    {
      "epoch": 3.582621915565431,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5375,
      "step": 15316
    },
    {
      "epoch": 3.58285580633844,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 15317
    },
    {
      "epoch": 3.5830896971114488,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7907,
      "step": 15318
    },
    {
      "epoch": 3.583323587884458,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 15319
    },
    {
      "epoch": 3.583557478657467,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 15320
    },
    {
      "epoch": 3.583791369430476,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6428,
      "step": 15321
    },
    {
      "epoch": 3.584025260203485,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.517,
      "step": 15322
    },
    {
      "epoch": 3.584259150976494,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 15323
    },
    {
      "epoch": 3.584493041749503,
      "grad_norm": 13.5,
      "learning_rate": 3e-05,
      "loss": 2.4398,
      "step": 15324
    },
    {
      "epoch": 3.584726932522512,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 15325
    },
    {
      "epoch": 3.5849608232955212,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 15326
    },
    {
      "epoch": 3.58519471406853,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 15327
    },
    {
      "epoch": 3.585428604841539,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 15328
    },
    {
      "epoch": 3.5856624956145478,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6597,
      "step": 15329
    },
    {
      "epoch": 3.585896386387557,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8252,
      "step": 15330
    },
    {
      "epoch": 3.586130277160566,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 15331
    },
    {
      "epoch": 3.586364167933575,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7128,
      "step": 15332
    },
    {
      "epoch": 3.586598058706584,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 15333
    },
    {
      "epoch": 3.586831949479593,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 15334
    },
    {
      "epoch": 3.587065840252602,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 15335
    },
    {
      "epoch": 3.587299731025611,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 15336
    },
    {
      "epoch": 3.5875336217986202,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 15337
    },
    {
      "epoch": 3.587767512571629,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 15338
    },
    {
      "epoch": 3.588001403344638,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 15339
    },
    {
      "epoch": 3.588235294117647,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7488,
      "step": 15340
    },
    {
      "epoch": 3.588469184890656,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 15341
    },
    {
      "epoch": 3.5887030756636653,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 15342
    },
    {
      "epoch": 3.588936966436674,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 15343
    },
    {
      "epoch": 3.589170857209683,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8976,
      "step": 15344
    },
    {
      "epoch": 3.589404747982692,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 15345
    },
    {
      "epoch": 3.589638638755701,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 15346
    },
    {
      "epoch": 3.58987252952871,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.754,
      "step": 15347
    },
    {
      "epoch": 3.5901064203017192,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6688,
      "step": 15348
    },
    {
      "epoch": 3.590340311074728,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 15349
    },
    {
      "epoch": 3.590574201847737,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 15350
    },
    {
      "epoch": 3.590808092620746,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 15351
    },
    {
      "epoch": 3.591041983393755,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 15352
    },
    {
      "epoch": 3.5912758741667643,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8187,
      "step": 15353
    },
    {
      "epoch": 3.591509764939773,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 15354
    },
    {
      "epoch": 3.591743655712782,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6761,
      "step": 15355
    },
    {
      "epoch": 3.5919775464857913,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 15356
    },
    {
      "epoch": 3.5922114372588,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 15357
    },
    {
      "epoch": 3.5924453280318094,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 15358
    },
    {
      "epoch": 3.592679218804818,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 15359
    },
    {
      "epoch": 3.592913109577827,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 15360
    },
    {
      "epoch": 3.593147000350836,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 15361
    },
    {
      "epoch": 3.593380891123845,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 15362
    },
    {
      "epoch": 3.593614781896854,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 15363
    },
    {
      "epoch": 3.5938486726698633,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 15364
    },
    {
      "epoch": 3.594082563442872,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 15365
    },
    {
      "epoch": 3.594316454215881,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 15366
    },
    {
      "epoch": 3.5945503449888903,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 15367
    },
    {
      "epoch": 3.594784235761899,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 15368
    },
    {
      "epoch": 3.5950181265349084,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 15369
    },
    {
      "epoch": 3.595252017307917,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 15370
    },
    {
      "epoch": 3.595485908080926,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7987,
      "step": 15371
    },
    {
      "epoch": 3.5957197988539353,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4114,
      "step": 15372
    },
    {
      "epoch": 3.595953689626944,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 15373
    },
    {
      "epoch": 3.5961875803999535,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 15374
    },
    {
      "epoch": 3.5964214711729623,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7343,
      "step": 15375
    },
    {
      "epoch": 3.596655361945971,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 15376
    },
    {
      "epoch": 3.59688925271898,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 15377
    },
    {
      "epoch": 3.5971231434919893,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 15378
    },
    {
      "epoch": 3.597357034264998,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 15379
    },
    {
      "epoch": 3.5975909250380074,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 15380
    },
    {
      "epoch": 3.597824815811016,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 15381
    },
    {
      "epoch": 3.598058706584025,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 15382
    },
    {
      "epoch": 3.5982925973570343,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6607,
      "step": 15383
    },
    {
      "epoch": 3.598526488130043,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 15384
    },
    {
      "epoch": 3.5987603789030524,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 15385
    },
    {
      "epoch": 3.5989942696760613,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 15386
    },
    {
      "epoch": 3.59922816044907,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 15387
    },
    {
      "epoch": 3.5994620512220794,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 15388
    },
    {
      "epoch": 3.5996959419950882,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7,
      "step": 15389
    },
    {
      "epoch": 3.5999298327680975,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 15390
    },
    {
      "epoch": 3.6001637235411064,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 15391
    },
    {
      "epoch": 3.600397614314115,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 15392
    },
    {
      "epoch": 3.600631505087124,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6629,
      "step": 15393
    },
    {
      "epoch": 3.6008653958601333,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 15394
    },
    {
      "epoch": 3.601099286633142,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 15395
    },
    {
      "epoch": 3.6013331774061514,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 15396
    },
    {
      "epoch": 3.6015670681791603,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 15397
    },
    {
      "epoch": 3.601800958952169,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.6153,
      "step": 15398
    },
    {
      "epoch": 3.6020348497251784,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 15399
    },
    {
      "epoch": 3.6022687404981872,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 15400
    },
    {
      "epoch": 3.6022687404981872,
      "eval_runtime": 4.5927,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 15400
    },
    {
      "epoch": 3.6025026312711965,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9538,
      "step": 15401
    },
    {
      "epoch": 3.6027365220442054,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 15402
    },
    {
      "epoch": 3.602970412817214,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 15403
    },
    {
      "epoch": 3.6032043035902235,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 15404
    },
    {
      "epoch": 3.6034381943632323,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 15405
    },
    {
      "epoch": 3.6036720851362416,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 15406
    },
    {
      "epoch": 3.6039059759092504,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 15407
    },
    {
      "epoch": 3.6041398666822593,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 15408
    },
    {
      "epoch": 3.604373757455268,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 15409
    },
    {
      "epoch": 3.6046076482282774,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9244,
      "step": 15410
    },
    {
      "epoch": 3.6048415390012867,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 15411
    },
    {
      "epoch": 3.6050754297742955,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 15412
    },
    {
      "epoch": 3.6053093205473044,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 15413
    },
    {
      "epoch": 3.605543211320313,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 15414
    },
    {
      "epoch": 3.6057771020933225,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 15415
    },
    {
      "epoch": 3.6060109928663313,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6677,
      "step": 15416
    },
    {
      "epoch": 3.6062448836393406,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 15417
    },
    {
      "epoch": 3.6064787744123494,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 15418
    },
    {
      "epoch": 3.6067126651853583,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 15419
    },
    {
      "epoch": 3.6069465559583676,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 15420
    },
    {
      "epoch": 3.6071804467313764,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.6624,
      "step": 15421
    },
    {
      "epoch": 3.6074143375043857,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 15422
    },
    {
      "epoch": 3.6076482282773945,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 15423
    },
    {
      "epoch": 3.6078821190504033,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 15424
    },
    {
      "epoch": 3.6081160098234126,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7074,
      "step": 15425
    },
    {
      "epoch": 3.6083499005964215,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6418,
      "step": 15426
    },
    {
      "epoch": 3.6085837913694307,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 15427
    },
    {
      "epoch": 3.6088176821424396,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 15428
    },
    {
      "epoch": 3.6090515729154484,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 15429
    },
    {
      "epoch": 3.6092854636884573,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 15430
    },
    {
      "epoch": 3.6095193544614665,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7369,
      "step": 15431
    },
    {
      "epoch": 3.6097532452344754,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 15432
    },
    {
      "epoch": 3.6099871360074847,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 15433
    },
    {
      "epoch": 3.6102210267804935,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 15434
    },
    {
      "epoch": 3.6104549175535023,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 15435
    },
    {
      "epoch": 3.6106888083265116,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8096,
      "step": 15436
    },
    {
      "epoch": 3.6109226990995205,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 15437
    },
    {
      "epoch": 3.6111565898725297,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 15438
    },
    {
      "epoch": 3.6113904806455386,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 15439
    },
    {
      "epoch": 3.6116243714185474,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 15440
    },
    {
      "epoch": 3.6118582621915567,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 15441
    },
    {
      "epoch": 3.6120921529645655,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 15442
    },
    {
      "epoch": 3.612326043737575,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 15443
    },
    {
      "epoch": 3.6125599345105837,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 15444
    },
    {
      "epoch": 3.6127938252835925,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8363,
      "step": 15445
    },
    {
      "epoch": 3.6130277160566013,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 15446
    },
    {
      "epoch": 3.6132616068296106,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 15447
    },
    {
      "epoch": 3.6134954976026195,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5906,
      "step": 15448
    },
    {
      "epoch": 3.6137293883756287,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7406,
      "step": 15449
    },
    {
      "epoch": 3.6139632791486376,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7846,
      "step": 15450
    },
    {
      "epoch": 3.6141971699216464,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8276,
      "step": 15451
    },
    {
      "epoch": 3.6144310606946557,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.827,
      "step": 15452
    },
    {
      "epoch": 3.6146649514676645,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7818,
      "step": 15453
    },
    {
      "epoch": 3.614898842240674,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4455,
      "step": 15454
    },
    {
      "epoch": 3.6151327330136827,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3058,
      "step": 15455
    },
    {
      "epoch": 3.6153666237866915,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 15456
    },
    {
      "epoch": 3.6156005145597008,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 15457
    },
    {
      "epoch": 3.6158344053327096,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 15458
    },
    {
      "epoch": 3.616068296105719,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 15459
    },
    {
      "epoch": 3.6163021868787277,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3467,
      "step": 15460
    },
    {
      "epoch": 3.6165360776517366,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 15461
    },
    {
      "epoch": 3.6167699684247454,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7478,
      "step": 15462
    },
    {
      "epoch": 3.6170038591977547,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 15463
    },
    {
      "epoch": 3.6172377499707635,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 15464
    },
    {
      "epoch": 3.617471640743773,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 15465
    },
    {
      "epoch": 3.6177055315167816,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 15466
    },
    {
      "epoch": 3.6179394222897905,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 15467
    },
    {
      "epoch": 3.6181733130627998,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 15468
    },
    {
      "epoch": 3.6184072038358086,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5973,
      "step": 15469
    },
    {
      "epoch": 3.618641094608818,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 15470
    },
    {
      "epoch": 3.6188749853818267,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 15471
    },
    {
      "epoch": 3.6191088761548356,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 15472
    },
    {
      "epoch": 3.619342766927845,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 15473
    },
    {
      "epoch": 3.6195766577008537,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 15474
    },
    {
      "epoch": 3.619810548473863,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 15475
    },
    {
      "epoch": 3.620044439246872,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 15476
    },
    {
      "epoch": 3.6202783300198806,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 15477
    },
    {
      "epoch": 3.6205122207928895,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.5692,
      "step": 15478
    },
    {
      "epoch": 3.6207461115658988,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 15479
    },
    {
      "epoch": 3.6209800023389076,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 15480
    },
    {
      "epoch": 3.621213893111917,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6972,
      "step": 15481
    },
    {
      "epoch": 3.6214477838849257,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.73,
      "step": 15482
    },
    {
      "epoch": 3.6216816746579346,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 15483
    },
    {
      "epoch": 3.621915565430944,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 15484
    },
    {
      "epoch": 3.6221494562039527,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7479,
      "step": 15485
    },
    {
      "epoch": 3.622383346976962,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 15486
    },
    {
      "epoch": 3.622617237749971,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 15487
    },
    {
      "epoch": 3.6228511285229796,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 15488
    },
    {
      "epoch": 3.623085019295989,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 15489
    },
    {
      "epoch": 3.6233189100689978,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 15490
    },
    {
      "epoch": 3.623552800842007,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 15491
    },
    {
      "epoch": 3.623786691615016,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5789,
      "step": 15492
    },
    {
      "epoch": 3.6240205823880247,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 15493
    },
    {
      "epoch": 3.6242544731610336,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 15494
    },
    {
      "epoch": 3.624488363934043,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 15495
    },
    {
      "epoch": 3.6247222547070517,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 15496
    },
    {
      "epoch": 3.624956145480061,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 15497
    },
    {
      "epoch": 3.62519003625307,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.4011,
      "step": 15498
    },
    {
      "epoch": 3.6254239270260786,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 15499
    },
    {
      "epoch": 3.625657817799088,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6512,
      "step": 15500
    },
    {
      "epoch": 3.625657817799088,
      "eval_runtime": 4.5872,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 15500
    },
    {
      "epoch": 3.6258917085720967,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3135,
      "step": 15501
    },
    {
      "epoch": 3.626125599345106,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 15502
    },
    {
      "epoch": 3.626359490118115,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 15503
    },
    {
      "epoch": 3.6265933808911237,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 15504
    },
    {
      "epoch": 3.626827271664133,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 15505
    },
    {
      "epoch": 3.627061162437142,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 15506
    },
    {
      "epoch": 3.627295053210151,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.229,
      "step": 15507
    },
    {
      "epoch": 3.62752894398316,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 15508
    },
    {
      "epoch": 3.627762834756169,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7082,
      "step": 15509
    },
    {
      "epoch": 3.6279967255291776,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 15510
    },
    {
      "epoch": 3.628230616302187,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 15511
    },
    {
      "epoch": 3.6284645070751957,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6688,
      "step": 15512
    },
    {
      "epoch": 3.628698397848205,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9303,
      "step": 15513
    },
    {
      "epoch": 3.628932288621214,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 15514
    },
    {
      "epoch": 3.6291661793942227,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 15515
    },
    {
      "epoch": 3.629400070167232,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 15516
    },
    {
      "epoch": 3.629633960940241,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6879,
      "step": 15517
    },
    {
      "epoch": 3.62986785171325,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7424,
      "step": 15518
    },
    {
      "epoch": 3.630101742486259,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6501,
      "step": 15519
    },
    {
      "epoch": 3.630335633259268,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 15520
    },
    {
      "epoch": 3.630569524032277,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 15521
    },
    {
      "epoch": 3.630803414805286,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 15522
    },
    {
      "epoch": 3.631037305578295,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6406,
      "step": 15523
    },
    {
      "epoch": 3.631271196351304,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 15524
    },
    {
      "epoch": 3.631505087124313,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7994,
      "step": 15525
    },
    {
      "epoch": 3.6317389778973217,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 15526
    },
    {
      "epoch": 3.631972868670331,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8824,
      "step": 15527
    },
    {
      "epoch": 3.63220675944334,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 15528
    },
    {
      "epoch": 3.632440650216349,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 15529
    },
    {
      "epoch": 3.632674540989358,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 15530
    },
    {
      "epoch": 3.6329084317623668,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 15531
    },
    {
      "epoch": 3.633142322535376,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 15532
    },
    {
      "epoch": 3.633376213308385,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 15533
    },
    {
      "epoch": 3.633610104081394,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 15534
    },
    {
      "epoch": 3.633843994854403,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 15535
    },
    {
      "epoch": 3.634077885627412,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 15536
    },
    {
      "epoch": 3.634311776400421,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.66,
      "step": 15537
    },
    {
      "epoch": 3.63454566717343,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 15538
    },
    {
      "epoch": 3.6347795579464393,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 15539
    },
    {
      "epoch": 3.635013448719448,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6854,
      "step": 15540
    },
    {
      "epoch": 3.635247339492457,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 15541
    },
    {
      "epoch": 3.6354812302654658,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 15542
    },
    {
      "epoch": 3.635715121038475,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.4724,
      "step": 15543
    },
    {
      "epoch": 3.635949011811484,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 15544
    },
    {
      "epoch": 3.636182902584493,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 15545
    },
    {
      "epoch": 3.636416793357502,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 15546
    },
    {
      "epoch": 3.636650684130511,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 15547
    },
    {
      "epoch": 3.63688457490352,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 15548
    },
    {
      "epoch": 3.637118465676529,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 15549
    },
    {
      "epoch": 3.6373523564495382,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 15550
    },
    {
      "epoch": 3.637586247222547,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 15551
    },
    {
      "epoch": 3.637820137995556,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 15552
    },
    {
      "epoch": 3.638054028768565,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 15553
    },
    {
      "epoch": 3.638287919541574,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 15554
    },
    {
      "epoch": 3.6385218103145833,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 15555
    },
    {
      "epoch": 3.638755701087592,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 15556
    },
    {
      "epoch": 3.638989591860601,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 15557
    },
    {
      "epoch": 3.63922348263361,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6425,
      "step": 15558
    },
    {
      "epoch": 3.639457373406619,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8026,
      "step": 15559
    },
    {
      "epoch": 3.639691264179628,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3311,
      "step": 15560
    },
    {
      "epoch": 3.6399251549526372,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 15561
    },
    {
      "epoch": 3.640159045725646,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 15562
    },
    {
      "epoch": 3.640392936498655,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.518,
      "step": 15563
    },
    {
      "epoch": 3.640626827271664,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6872,
      "step": 15564
    },
    {
      "epoch": 3.640860718044673,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 15565
    },
    {
      "epoch": 3.6410946088176823,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 15566
    },
    {
      "epoch": 3.641328499590691,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7561,
      "step": 15567
    },
    {
      "epoch": 3.6415623903637,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 15568
    },
    {
      "epoch": 3.6417962811367093,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 15569
    },
    {
      "epoch": 3.642030171909718,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 15570
    },
    {
      "epoch": 3.6422640626827274,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 15571
    },
    {
      "epoch": 3.6424979534557362,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 15572
    },
    {
      "epoch": 3.642731844228745,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 15573
    },
    {
      "epoch": 3.642965735001754,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7789,
      "step": 15574
    },
    {
      "epoch": 3.643199625774763,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 15575
    },
    {
      "epoch": 3.643433516547772,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 15576
    },
    {
      "epoch": 3.6436674073207813,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 15577
    },
    {
      "epoch": 3.64390129809379,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 15578
    },
    {
      "epoch": 3.644135188866799,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 15579
    },
    {
      "epoch": 3.6443690796398083,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 15580
    },
    {
      "epoch": 3.644602970412817,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 15581
    },
    {
      "epoch": 3.6448368611858264,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 15582
    },
    {
      "epoch": 3.6450707519588352,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7202,
      "step": 15583
    },
    {
      "epoch": 3.645304642731844,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 15584
    },
    {
      "epoch": 3.6455385335048534,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 15585
    },
    {
      "epoch": 3.645772424277862,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 15586
    },
    {
      "epoch": 3.6460063150508715,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 15587
    },
    {
      "epoch": 3.6462402058238803,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 15588
    },
    {
      "epoch": 3.646474096596889,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 15589
    },
    {
      "epoch": 3.646707987369898,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3345,
      "step": 15590
    },
    {
      "epoch": 3.6469418781429073,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 15591
    },
    {
      "epoch": 3.647175768915916,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6066,
      "step": 15592
    },
    {
      "epoch": 3.6474096596889254,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 15593
    },
    {
      "epoch": 3.6476435504619342,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.656,
      "step": 15594
    },
    {
      "epoch": 3.647877441234943,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6216,
      "step": 15595
    },
    {
      "epoch": 3.6481113320079523,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 15596
    },
    {
      "epoch": 3.648345222780961,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 15597
    },
    {
      "epoch": 3.6485791135539705,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 15598
    },
    {
      "epoch": 3.6488130043269793,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 15599
    },
    {
      "epoch": 3.649046895099988,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 15600
    },
    {
      "epoch": 3.649046895099988,
      "eval_runtime": 4.6281,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 15600
    },
    {
      "epoch": 3.6492807858729974,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 15601
    },
    {
      "epoch": 3.6495146766460063,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 15602
    },
    {
      "epoch": 3.6497485674190155,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 15603
    },
    {
      "epoch": 3.6499824581920244,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 15604
    },
    {
      "epoch": 3.650216348965033,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 15605
    },
    {
      "epoch": 3.650450239738042,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 15606
    },
    {
      "epoch": 3.6506841305110513,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 15607
    },
    {
      "epoch": 3.65091802128406,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 15608
    },
    {
      "epoch": 3.6511519120570695,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 15609
    },
    {
      "epoch": 3.6513858028300783,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 15610
    },
    {
      "epoch": 3.651619693603087,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 15611
    },
    {
      "epoch": 3.6518535843760964,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 15612
    },
    {
      "epoch": 3.6520874751491053,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6256,
      "step": 15613
    },
    {
      "epoch": 3.6523213659221145,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 15614
    },
    {
      "epoch": 3.6525552566951234,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 15615
    },
    {
      "epoch": 3.652789147468132,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 15616
    },
    {
      "epoch": 3.6530230382411415,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9262,
      "step": 15617
    },
    {
      "epoch": 3.6532569290141503,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6949,
      "step": 15618
    },
    {
      "epoch": 3.6534908197871596,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4641,
      "step": 15619
    },
    {
      "epoch": 3.6537247105601685,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 15620
    },
    {
      "epoch": 3.6539586013331773,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 15621
    },
    {
      "epoch": 3.6541924921061866,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 15622
    },
    {
      "epoch": 3.6544263828791954,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7745,
      "step": 15623
    },
    {
      "epoch": 3.6546602736522047,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 15624
    },
    {
      "epoch": 3.6548941644252135,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 15625
    },
    {
      "epoch": 3.6551280551982224,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 15626
    },
    {
      "epoch": 3.655361945971231,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 15627
    },
    {
      "epoch": 3.6555958367442405,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 15628
    },
    {
      "epoch": 3.6558297275172493,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.59,
      "step": 15629
    },
    {
      "epoch": 3.6560636182902586,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 15630
    },
    {
      "epoch": 3.6562975090632674,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 15631
    },
    {
      "epoch": 3.6565313998362763,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 15632
    },
    {
      "epoch": 3.6567652906092856,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 15633
    },
    {
      "epoch": 3.6569991813822944,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 15634
    },
    {
      "epoch": 3.6572330721553037,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 15635
    },
    {
      "epoch": 3.6574669629283125,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 15636
    },
    {
      "epoch": 3.6577008537013214,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 15637
    },
    {
      "epoch": 3.6579347444743306,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 15638
    },
    {
      "epoch": 3.6581686352473395,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5366,
      "step": 15639
    },
    {
      "epoch": 3.6584025260203488,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7147,
      "step": 15640
    },
    {
      "epoch": 3.6586364167933576,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 15641
    },
    {
      "epoch": 3.6588703075663664,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 15642
    },
    {
      "epoch": 3.6591041983393753,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 15643
    },
    {
      "epoch": 3.6593380891123846,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 15644
    },
    {
      "epoch": 3.6595719798853934,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 15645
    },
    {
      "epoch": 3.6598058706584027,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 15646
    },
    {
      "epoch": 3.6600397614314115,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 15647
    },
    {
      "epoch": 3.6602736522044204,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 15648
    },
    {
      "epoch": 3.6605075429774296,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 15649
    },
    {
      "epoch": 3.6607414337504385,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 15650
    },
    {
      "epoch": 3.6609753245234478,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7018,
      "step": 15651
    },
    {
      "epoch": 3.6612092152964566,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 15652
    },
    {
      "epoch": 3.6614431060694654,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 15653
    },
    {
      "epoch": 3.6616769968424747,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.3466,
      "step": 15654
    },
    {
      "epoch": 3.6619108876154836,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 15655
    },
    {
      "epoch": 3.662144778388493,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7554,
      "step": 15656
    },
    {
      "epoch": 3.6623786691615017,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 15657
    },
    {
      "epoch": 3.6626125599345105,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3153,
      "step": 15658
    },
    {
      "epoch": 3.6628464507075194,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 15659
    },
    {
      "epoch": 3.6630803414805286,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 15660
    },
    {
      "epoch": 3.6633142322535375,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8308,
      "step": 15661
    },
    {
      "epoch": 3.6635481230265468,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5073,
      "step": 15662
    },
    {
      "epoch": 3.6637820137995556,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 1.6643,
      "step": 15663
    },
    {
      "epoch": 3.6640159045725644,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 15664
    },
    {
      "epoch": 3.6642497953455737,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 15665
    },
    {
      "epoch": 3.6644836861185826,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 15666
    },
    {
      "epoch": 3.664717576891592,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8607,
      "step": 15667
    },
    {
      "epoch": 3.6649514676646007,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 15668
    },
    {
      "epoch": 3.6651853584376095,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 15669
    },
    {
      "epoch": 3.665419249210619,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 15670
    },
    {
      "epoch": 3.6656531399836276,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.492,
      "step": 15671
    },
    {
      "epoch": 3.665887030756637,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6792,
      "step": 15672
    },
    {
      "epoch": 3.6661209215296457,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 15673
    },
    {
      "epoch": 3.6663548123026546,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 15674
    },
    {
      "epoch": 3.6665887030756634,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 15675
    },
    {
      "epoch": 3.6668225938486727,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 15676
    },
    {
      "epoch": 3.6670564846216815,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 15677
    },
    {
      "epoch": 3.667290375394691,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 15678
    },
    {
      "epoch": 3.6675242661676997,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 15679
    },
    {
      "epoch": 3.6677581569407085,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 15680
    },
    {
      "epoch": 3.667992047713718,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 15681
    },
    {
      "epoch": 3.6682259384867266,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6262,
      "step": 15682
    },
    {
      "epoch": 3.668459829259736,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 15683
    },
    {
      "epoch": 3.6686937200327447,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6793,
      "step": 15684
    },
    {
      "epoch": 3.6689276108057536,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 15685
    },
    {
      "epoch": 3.669161501578763,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 15686
    },
    {
      "epoch": 3.6693953923517717,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 15687
    },
    {
      "epoch": 3.669629283124781,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 15688
    },
    {
      "epoch": 3.66986317389779,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 15689
    },
    {
      "epoch": 3.6700970646707987,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7146,
      "step": 15690
    },
    {
      "epoch": 3.6703309554438075,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 15691
    },
    {
      "epoch": 3.670564846216817,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 15692
    },
    {
      "epoch": 3.6707987369898256,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5628,
      "step": 15693
    },
    {
      "epoch": 3.671032627762835,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 15694
    },
    {
      "epoch": 3.6712665185358437,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 15695
    },
    {
      "epoch": 3.6715004093088526,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7454,
      "step": 15696
    },
    {
      "epoch": 3.671734300081862,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8503,
      "step": 15697
    },
    {
      "epoch": 3.6719681908548707,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 15698
    },
    {
      "epoch": 3.67220208162788,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 15699
    },
    {
      "epoch": 3.672435972400889,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 15700
    },
    {
      "epoch": 3.672435972400889,
      "eval_runtime": 4.6269,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 15700
    },
    {
      "epoch": 3.6726698631738977,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 15701
    },
    {
      "epoch": 3.672903753946907,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 15702
    },
    {
      "epoch": 3.6731376447199158,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 15703
    },
    {
      "epoch": 3.673371535492925,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 15704
    },
    {
      "epoch": 3.673605426265934,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 15705
    },
    {
      "epoch": 3.6738393170389427,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6716,
      "step": 15706
    },
    {
      "epoch": 3.6740732078119516,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9094,
      "step": 15707
    },
    {
      "epoch": 3.674307098584961,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 15708
    },
    {
      "epoch": 3.6745409893579697,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.3682,
      "step": 15709
    },
    {
      "epoch": 3.674774880130979,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 15710
    },
    {
      "epoch": 3.675008770903988,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 15711
    },
    {
      "epoch": 3.6752426616769966,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 15712
    },
    {
      "epoch": 3.675476552450006,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 15713
    },
    {
      "epoch": 3.6757104432230148,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 15714
    },
    {
      "epoch": 3.675944333996024,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 15715
    },
    {
      "epoch": 3.676178224769033,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 15716
    },
    {
      "epoch": 3.6764121155420417,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5091,
      "step": 15717
    },
    {
      "epoch": 3.676646006315051,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 15718
    },
    {
      "epoch": 3.67687989708806,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 15719
    },
    {
      "epoch": 3.677113787861069,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 15720
    },
    {
      "epoch": 3.677347678634078,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 15721
    },
    {
      "epoch": 3.677581569407087,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 15722
    },
    {
      "epoch": 3.6778154601800956,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 15723
    },
    {
      "epoch": 3.678049350953105,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 15724
    },
    {
      "epoch": 3.6782832417261138,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 15725
    },
    {
      "epoch": 3.678517132499123,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9388,
      "step": 15726
    },
    {
      "epoch": 3.678751023272132,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 15727
    },
    {
      "epoch": 3.6789849140451407,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8354,
      "step": 15728
    },
    {
      "epoch": 3.67921880481815,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 15729
    },
    {
      "epoch": 3.679452695591159,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 15730
    },
    {
      "epoch": 3.679686586364168,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 15731
    },
    {
      "epoch": 3.679920477137177,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 15732
    },
    {
      "epoch": 3.680154367910186,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 15733
    },
    {
      "epoch": 3.680388258683195,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7295,
      "step": 15734
    },
    {
      "epoch": 3.680622149456204,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7488,
      "step": 15735
    },
    {
      "epoch": 3.680856040229213,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 15736
    },
    {
      "epoch": 3.681089931002222,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 15737
    },
    {
      "epoch": 3.681323821775231,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 15738
    },
    {
      "epoch": 3.6815577125482397,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 15739
    },
    {
      "epoch": 3.681791603321249,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 15740
    },
    {
      "epoch": 3.682025494094258,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 15741
    },
    {
      "epoch": 3.682259384867267,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 15742
    },
    {
      "epoch": 3.682493275640276,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6889,
      "step": 15743
    },
    {
      "epoch": 3.682727166413285,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4979,
      "step": 15744
    },
    {
      "epoch": 3.682961057186294,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 15745
    },
    {
      "epoch": 3.683194947959303,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 15746
    },
    {
      "epoch": 3.683428838732312,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 15747
    },
    {
      "epoch": 3.683662729505321,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7084,
      "step": 15748
    },
    {
      "epoch": 3.68389662027833,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 15749
    },
    {
      "epoch": 3.684130511051339,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 15750
    },
    {
      "epoch": 3.684364401824348,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6388,
      "step": 15751
    },
    {
      "epoch": 3.6845982925973573,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 15752
    },
    {
      "epoch": 3.684832183370366,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 15753
    },
    {
      "epoch": 3.685066074143375,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 15754
    },
    {
      "epoch": 3.685299964916384,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5999,
      "step": 15755
    },
    {
      "epoch": 3.685533855689393,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 15756
    },
    {
      "epoch": 3.685767746462402,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 15757
    },
    {
      "epoch": 3.686001637235411,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 15758
    },
    {
      "epoch": 3.68623552800842,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 15759
    },
    {
      "epoch": 3.686469418781429,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.705,
      "step": 15760
    },
    {
      "epoch": 3.686703309554438,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7255,
      "step": 15761
    },
    {
      "epoch": 3.686937200327447,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 15762
    },
    {
      "epoch": 3.6871710911004563,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 15763
    },
    {
      "epoch": 3.687404981873465,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 15764
    },
    {
      "epoch": 3.687638872646474,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 15765
    },
    {
      "epoch": 3.6878727634194832,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 15766
    },
    {
      "epoch": 3.688106654192492,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.676,
      "step": 15767
    },
    {
      "epoch": 3.6883405449655013,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6731,
      "step": 15768
    },
    {
      "epoch": 3.68857443573851,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 15769
    },
    {
      "epoch": 3.688808326511519,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 15770
    },
    {
      "epoch": 3.689042217284528,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 15771
    },
    {
      "epoch": 3.689276108057537,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 15772
    },
    {
      "epoch": 3.689509998830546,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7415,
      "step": 15773
    },
    {
      "epoch": 3.6897438896035553,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 15774
    },
    {
      "epoch": 3.689977780376564,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 15775
    },
    {
      "epoch": 3.690211671149573,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5576,
      "step": 15776
    },
    {
      "epoch": 3.690445561922582,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 15777
    },
    {
      "epoch": 3.690679452695591,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6375,
      "step": 15778
    },
    {
      "epoch": 3.6909133434686003,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 15779
    },
    {
      "epoch": 3.691147234241609,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 15780
    },
    {
      "epoch": 3.691381125014618,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7391,
      "step": 15781
    },
    {
      "epoch": 3.6916150157876273,
      "grad_norm": 11.8125,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 15782
    },
    {
      "epoch": 3.691848906560636,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 15783
    },
    {
      "epoch": 3.6920827973336454,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 15784
    },
    {
      "epoch": 3.6923166881066543,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 15785
    },
    {
      "epoch": 3.692550578879663,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 15786
    },
    {
      "epoch": 3.692784469652672,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 15787
    },
    {
      "epoch": 3.693018360425681,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 15788
    },
    {
      "epoch": 3.69325225119869,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 15789
    },
    {
      "epoch": 3.6934861419716993,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 15790
    },
    {
      "epoch": 3.693720032744708,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 15791
    },
    {
      "epoch": 3.693953923517717,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5634,
      "step": 15792
    },
    {
      "epoch": 3.6941878142907263,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6317,
      "step": 15793
    },
    {
      "epoch": 3.694421705063735,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 15794
    },
    {
      "epoch": 3.6946555958367444,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 15795
    },
    {
      "epoch": 3.6948894866097532,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 15796
    },
    {
      "epoch": 3.695123377382762,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 15797
    },
    {
      "epoch": 3.6953572681557714,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 2.2963,
      "step": 15798
    },
    {
      "epoch": 3.69559115892878,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 15799
    },
    {
      "epoch": 3.6958250497017895,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 15800
    },
    {
      "epoch": 3.6958250497017895,
      "eval_runtime": 4.6112,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 15800
    },
    {
      "epoch": 3.6960589404747983,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.4485,
      "step": 15801
    },
    {
      "epoch": 3.696292831247807,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2516,
      "step": 15802
    },
    {
      "epoch": 3.696526722020816,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 15803
    },
    {
      "epoch": 3.6967606127938253,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 15804
    },
    {
      "epoch": 3.696994503566834,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 15805
    },
    {
      "epoch": 3.6972283943398434,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 15806
    },
    {
      "epoch": 3.6974622851128522,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7052,
      "step": 15807
    },
    {
      "epoch": 3.697696175885861,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6804,
      "step": 15808
    },
    {
      "epoch": 3.6979300666588704,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2409,
      "step": 15809
    },
    {
      "epoch": 3.698163957431879,
      "grad_norm": 2.515625,
      "learning_rate": 3e-05,
      "loss": 1.5635,
      "step": 15810
    },
    {
      "epoch": 3.6983978482048885,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 15811
    },
    {
      "epoch": 3.6986317389778973,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 15812
    },
    {
      "epoch": 3.698865629750906,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 15813
    },
    {
      "epoch": 3.6990995205239154,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5595,
      "step": 15814
    },
    {
      "epoch": 3.6993334112969243,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 15815
    },
    {
      "epoch": 3.6995673020699336,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 15816
    },
    {
      "epoch": 3.6998011928429424,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7932,
      "step": 15817
    },
    {
      "epoch": 3.7000350836159512,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 15818
    },
    {
      "epoch": 3.7002689743889605,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 15819
    },
    {
      "epoch": 3.7005028651619694,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 15820
    },
    {
      "epoch": 3.7007367559349786,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 15821
    },
    {
      "epoch": 3.7009706467079875,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 15822
    },
    {
      "epoch": 3.7012045374809963,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 15823
    },
    {
      "epoch": 3.701438428254005,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 15824
    },
    {
      "epoch": 3.7016723190270144,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 15825
    },
    {
      "epoch": 3.7019062098000233,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 15826
    },
    {
      "epoch": 3.7021401005730326,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 15827
    },
    {
      "epoch": 3.7023739913460414,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 15828
    },
    {
      "epoch": 3.7026078821190502,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4953,
      "step": 15829
    },
    {
      "epoch": 3.7028417728920595,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 15830
    },
    {
      "epoch": 3.7030756636650684,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 15831
    },
    {
      "epoch": 3.7033095544380776,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 15832
    },
    {
      "epoch": 3.7035434452110865,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 15833
    },
    {
      "epoch": 3.7037773359840953,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 15834
    },
    {
      "epoch": 3.7040112267571046,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 15835
    },
    {
      "epoch": 3.7042451175301134,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5452,
      "step": 15836
    },
    {
      "epoch": 3.7044790083031227,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 15837
    },
    {
      "epoch": 3.7047128990761315,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 15838
    },
    {
      "epoch": 3.7049467898491404,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 15839
    },
    {
      "epoch": 3.7051806806221492,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 15840
    },
    {
      "epoch": 3.7054145713951585,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 15841
    },
    {
      "epoch": 3.7056484621681673,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 15842
    },
    {
      "epoch": 3.7058823529411766,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 15843
    },
    {
      "epoch": 3.7061162437141855,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 15844
    },
    {
      "epoch": 3.7063501344871943,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.3421,
      "step": 15845
    },
    {
      "epoch": 3.7065840252602036,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 15846
    },
    {
      "epoch": 3.7068179160332124,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 15847
    },
    {
      "epoch": 3.7070518068062217,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 15848
    },
    {
      "epoch": 3.7072856975792305,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 15849
    },
    {
      "epoch": 3.7075195883522394,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 15850
    },
    {
      "epoch": 3.7077534791252487,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 15851
    },
    {
      "epoch": 3.7079873698982575,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 15852
    },
    {
      "epoch": 3.708221260671267,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5375,
      "step": 15853
    },
    {
      "epoch": 3.7084551514442756,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.596,
      "step": 15854
    },
    {
      "epoch": 3.7086890422172845,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 15855
    },
    {
      "epoch": 3.7089229329902933,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5518,
      "step": 15856
    },
    {
      "epoch": 3.7091568237633026,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5954,
      "step": 15857
    },
    {
      "epoch": 3.7093907145363114,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 15858
    },
    {
      "epoch": 3.7096246053093207,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 15859
    },
    {
      "epoch": 3.7098584960823295,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7023,
      "step": 15860
    },
    {
      "epoch": 3.7100923868553384,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 15861
    },
    {
      "epoch": 3.7103262776283477,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 15862
    },
    {
      "epoch": 3.7105601684013565,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7156,
      "step": 15863
    },
    {
      "epoch": 3.7107940591743658,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5863,
      "step": 15864
    },
    {
      "epoch": 3.7110279499473746,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 15865
    },
    {
      "epoch": 3.7112618407203835,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 15866
    },
    {
      "epoch": 3.7114957314933927,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7141,
      "step": 15867
    },
    {
      "epoch": 3.7117296222664016,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 15868
    },
    {
      "epoch": 3.711963513039411,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7068,
      "step": 15869
    },
    {
      "epoch": 3.7121974038124197,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 15870
    },
    {
      "epoch": 3.7124312945854285,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 15871
    },
    {
      "epoch": 3.7126651853584374,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 15872
    },
    {
      "epoch": 3.7128990761314467,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 15873
    },
    {
      "epoch": 3.7131329669044555,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 15874
    },
    {
      "epoch": 3.7133668576774648,
      "grad_norm": 13.9375,
      "learning_rate": 3e-05,
      "loss": 2.2513,
      "step": 15875
    },
    {
      "epoch": 3.7136007484504736,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 15876
    },
    {
      "epoch": 3.7138346392234824,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 15877
    },
    {
      "epoch": 3.7140685299964917,
      "grad_norm": 9.875,
      "learning_rate": 3e-05,
      "loss": 2.345,
      "step": 15878
    },
    {
      "epoch": 3.7143024207695006,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7215,
      "step": 15879
    },
    {
      "epoch": 3.71453631154251,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 15880
    },
    {
      "epoch": 3.7147702023155187,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 15881
    },
    {
      "epoch": 3.7150040930885275,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1725,
      "step": 15882
    },
    {
      "epoch": 3.715237983861537,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 15883
    },
    {
      "epoch": 3.7154718746345456,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7173,
      "step": 15884
    },
    {
      "epoch": 3.715705765407555,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 15885
    },
    {
      "epoch": 3.7159396561805638,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.752,
      "step": 15886
    },
    {
      "epoch": 3.7161735469535726,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 15887
    },
    {
      "epoch": 3.7164074377265814,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5785,
      "step": 15888
    },
    {
      "epoch": 3.7166413284995907,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 15889
    },
    {
      "epoch": 3.7168752192725996,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 15890
    },
    {
      "epoch": 3.717109110045609,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 15891
    },
    {
      "epoch": 3.7173430008186177,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 15892
    },
    {
      "epoch": 3.7175768915916265,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4382,
      "step": 15893
    },
    {
      "epoch": 3.717810782364636,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3655,
      "step": 15894
    },
    {
      "epoch": 3.7180446731376446,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 15895
    },
    {
      "epoch": 3.718278563910654,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 15896
    },
    {
      "epoch": 3.7185124546836628,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 15897
    },
    {
      "epoch": 3.7187463454566716,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 15898
    },
    {
      "epoch": 3.718980236229681,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 15899
    },
    {
      "epoch": 3.7192141270026897,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 15900
    },
    {
      "epoch": 3.7192141270026897,
      "eval_runtime": 4.6113,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 15900
    },
    {
      "epoch": 3.719448017775699,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.743,
      "step": 15901
    },
    {
      "epoch": 3.719681908548708,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6343,
      "step": 15902
    },
    {
      "epoch": 3.7199157993217167,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 15903
    },
    {
      "epoch": 3.7201496900947255,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 15904
    },
    {
      "epoch": 3.720383580867735,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6234,
      "step": 15905
    },
    {
      "epoch": 3.7206174716407436,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 15906
    },
    {
      "epoch": 3.720851362413753,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 15907
    },
    {
      "epoch": 3.7210852531867618,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 15908
    },
    {
      "epoch": 3.7213191439597706,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 15909
    },
    {
      "epoch": 3.72155303473278,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5726,
      "step": 15910
    },
    {
      "epoch": 3.7217869255057887,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 15911
    },
    {
      "epoch": 3.722020816278798,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 15912
    },
    {
      "epoch": 3.722254707051807,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 15913
    },
    {
      "epoch": 3.7224885978248157,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 15914
    },
    {
      "epoch": 3.722722488597825,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 15915
    },
    {
      "epoch": 3.722956379370834,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 15916
    },
    {
      "epoch": 3.723190270143843,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7439,
      "step": 15917
    },
    {
      "epoch": 3.723424160916852,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 15918
    },
    {
      "epoch": 3.7236580516898607,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 15919
    },
    {
      "epoch": 3.7238919424628696,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2452,
      "step": 15920
    },
    {
      "epoch": 3.724125833235879,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 15921
    },
    {
      "epoch": 3.7243597240088877,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 15922
    },
    {
      "epoch": 3.724593614781897,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 15923
    },
    {
      "epoch": 3.724827505554906,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 15924
    },
    {
      "epoch": 3.7250613963279147,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 15925
    },
    {
      "epoch": 3.725295287100924,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6636,
      "step": 15926
    },
    {
      "epoch": 3.725529177873933,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.663,
      "step": 15927
    },
    {
      "epoch": 3.725763068646942,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7076,
      "step": 15928
    },
    {
      "epoch": 3.725996959419951,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 15929
    },
    {
      "epoch": 3.7262308501929597,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0492,
      "step": 15930
    },
    {
      "epoch": 3.726464740965969,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 15931
    },
    {
      "epoch": 3.726698631738978,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.155,
      "step": 15932
    },
    {
      "epoch": 3.726932522511987,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.3879,
      "step": 15933
    },
    {
      "epoch": 3.727166413284996,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 15934
    },
    {
      "epoch": 3.727400304058005,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 15935
    },
    {
      "epoch": 3.7276341948310137,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 15936
    },
    {
      "epoch": 3.727868085604023,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 15937
    },
    {
      "epoch": 3.728101976377032,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 15938
    },
    {
      "epoch": 3.728335867150041,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 15939
    },
    {
      "epoch": 3.72856975792305,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7908,
      "step": 15940
    },
    {
      "epoch": 3.7288036486960587,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 15941
    },
    {
      "epoch": 3.729037539469068,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 15942
    },
    {
      "epoch": 3.729271430242077,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 15943
    },
    {
      "epoch": 3.729505321015086,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.838,
      "step": 15944
    },
    {
      "epoch": 3.729739211788095,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 15945
    },
    {
      "epoch": 3.729973102561104,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6361,
      "step": 15946
    },
    {
      "epoch": 3.730206993334113,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 15947
    },
    {
      "epoch": 3.730440884107122,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 15948
    },
    {
      "epoch": 3.730674774880131,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6195,
      "step": 15949
    },
    {
      "epoch": 3.73090866565314,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6462,
      "step": 15950
    },
    {
      "epoch": 3.731142556426149,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 15951
    },
    {
      "epoch": 3.7313764471991577,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 15952
    },
    {
      "epoch": 3.731610337972167,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.187,
      "step": 15953
    },
    {
      "epoch": 3.731844228745176,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 15954
    },
    {
      "epoch": 3.732078119518185,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 15955
    },
    {
      "epoch": 3.732312010291194,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 15956
    },
    {
      "epoch": 3.732545901064203,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7615,
      "step": 15957
    },
    {
      "epoch": 3.732779791837212,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 15958
    },
    {
      "epoch": 3.733013682610221,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 15959
    },
    {
      "epoch": 3.73324757338323,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 15960
    },
    {
      "epoch": 3.733481464156239,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 15961
    },
    {
      "epoch": 3.733715354929248,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 15962
    },
    {
      "epoch": 3.733949245702257,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.3784,
      "step": 15963
    },
    {
      "epoch": 3.734183136475266,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 15964
    },
    {
      "epoch": 3.7344170272482753,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 15965
    },
    {
      "epoch": 3.734650918021284,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 15966
    },
    {
      "epoch": 3.734884808794293,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 15967
    },
    {
      "epoch": 3.735118699567302,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 15968
    },
    {
      "epoch": 3.735352590340311,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 15969
    },
    {
      "epoch": 3.73558648111332,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 15970
    },
    {
      "epoch": 3.735820371886329,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 15971
    },
    {
      "epoch": 3.736054262659338,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2421,
      "step": 15972
    },
    {
      "epoch": 3.736288153432347,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 15973
    },
    {
      "epoch": 3.736522044205356,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 15974
    },
    {
      "epoch": 3.736755934978365,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 15975
    },
    {
      "epoch": 3.7369898257513743,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5637,
      "step": 15976
    },
    {
      "epoch": 3.737223716524383,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 15977
    },
    {
      "epoch": 3.737457607297392,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8087,
      "step": 15978
    },
    {
      "epoch": 3.7376914980704012,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 15979
    },
    {
      "epoch": 3.73792538884341,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 15980
    },
    {
      "epoch": 3.7381592796164194,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4591,
      "step": 15981
    },
    {
      "epoch": 3.738393170389428,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 15982
    },
    {
      "epoch": 3.738627061162437,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 15983
    },
    {
      "epoch": 3.738860951935446,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 15984
    },
    {
      "epoch": 3.739094842708455,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 15985
    },
    {
      "epoch": 3.739328733481464,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 15986
    },
    {
      "epoch": 3.7395626242544733,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 15987
    },
    {
      "epoch": 3.739796515027482,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 15988
    },
    {
      "epoch": 3.740030405800491,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 15989
    },
    {
      "epoch": 3.7402642965735002,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 15990
    },
    {
      "epoch": 3.740498187346509,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 15991
    },
    {
      "epoch": 3.7407320781195184,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 15992
    },
    {
      "epoch": 3.740965968892527,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 15993
    },
    {
      "epoch": 3.741199859665536,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 15994
    },
    {
      "epoch": 3.7414337504385453,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 15995
    },
    {
      "epoch": 3.741667641211554,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.73,
      "step": 15996
    },
    {
      "epoch": 3.7419015319845634,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 15997
    },
    {
      "epoch": 3.7421354227575723,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 15998
    },
    {
      "epoch": 3.742369313530581,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 15999
    },
    {
      "epoch": 3.74260320430359,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8553,
      "step": 16000
    },
    {
      "epoch": 3.74260320430359,
      "eval_runtime": 4.7198,
      "eval_samples_per_second": 0.212,
      "eval_steps_per_second": 0.212,
      "step": 16000
    },
    {
      "epoch": 3.7428370950765992,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 16001
    },
    {
      "epoch": 3.743070985849608,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 16002
    },
    {
      "epoch": 3.7433048766226173,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7401,
      "step": 16003
    },
    {
      "epoch": 3.743538767395626,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 16004
    },
    {
      "epoch": 3.743772658168635,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 16005
    },
    {
      "epoch": 3.7440065489416443,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 16006
    },
    {
      "epoch": 3.744240439714653,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.626,
      "step": 16007
    },
    {
      "epoch": 3.7444743304876624,
      "grad_norm": 8.375,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 16008
    },
    {
      "epoch": 3.7447082212606713,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 16009
    },
    {
      "epoch": 3.74494211203368,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 16010
    },
    {
      "epoch": 3.7451760028066894,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6851,
      "step": 16011
    },
    {
      "epoch": 3.7454098935796982,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 16012
    },
    {
      "epoch": 3.7456437843527075,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 16013
    },
    {
      "epoch": 3.7458776751257163,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 16014
    },
    {
      "epoch": 3.746111565898725,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 16015
    },
    {
      "epoch": 3.7463454566717345,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 16016
    },
    {
      "epoch": 3.7465793474447433,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6834,
      "step": 16017
    },
    {
      "epoch": 3.7468132382177526,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 16018
    },
    {
      "epoch": 3.7470471289907614,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 16019
    },
    {
      "epoch": 3.7472810197637703,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 16020
    },
    {
      "epoch": 3.747514910536779,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 16021
    },
    {
      "epoch": 3.7477488013097884,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 16022
    },
    {
      "epoch": 3.747982692082797,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 16023
    },
    {
      "epoch": 3.7482165828558065,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 16024
    },
    {
      "epoch": 3.7484504736288153,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 16025
    },
    {
      "epoch": 3.748684364401824,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 16026
    },
    {
      "epoch": 3.7489182551748335,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 16027
    },
    {
      "epoch": 3.7491521459478423,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6353,
      "step": 16028
    },
    {
      "epoch": 3.7493860367208516,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 16029
    },
    {
      "epoch": 3.7496199274938604,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 16030
    },
    {
      "epoch": 3.7498538182668693,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6412,
      "step": 16031
    },
    {
      "epoch": 3.7500877090398785,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 16032
    },
    {
      "epoch": 3.7503215998128874,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5241,
      "step": 16033
    },
    {
      "epoch": 3.7505554905858967,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 16034
    },
    {
      "epoch": 3.7507893813589055,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 16035
    },
    {
      "epoch": 3.7510232721319143,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 16036
    },
    {
      "epoch": 3.751257162904923,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 16037
    },
    {
      "epoch": 3.7514910536779325,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5775,
      "step": 16038
    },
    {
      "epoch": 3.7517249444509413,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 16039
    },
    {
      "epoch": 3.7519588352239506,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 16040
    },
    {
      "epoch": 3.7521927259969594,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 16041
    },
    {
      "epoch": 3.7524266167699682,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 16042
    },
    {
      "epoch": 3.7526605075429775,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 16043
    },
    {
      "epoch": 3.7528943983159864,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 16044
    },
    {
      "epoch": 3.7531282890889957,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 16045
    },
    {
      "epoch": 3.7533621798620045,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 16046
    },
    {
      "epoch": 3.7535960706350133,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 16047
    },
    {
      "epoch": 3.7538299614080226,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 16048
    },
    {
      "epoch": 3.7540638521810314,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1881,
      "step": 16049
    },
    {
      "epoch": 3.7542977429540407,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 16050
    },
    {
      "epoch": 3.7545316337270496,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 16051
    },
    {
      "epoch": 3.7547655245000584,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 16052
    },
    {
      "epoch": 3.7549994152730672,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7663,
      "step": 16053
    },
    {
      "epoch": 3.7552333060460765,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 16054
    },
    {
      "epoch": 3.7554671968190854,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 16055
    },
    {
      "epoch": 3.7557010875920946,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 16056
    },
    {
      "epoch": 3.7559349783651035,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.687,
      "step": 16057
    },
    {
      "epoch": 3.7561688691381123,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 16058
    },
    {
      "epoch": 3.7564027599111216,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 16059
    },
    {
      "epoch": 3.7566366506841304,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8284,
      "step": 16060
    },
    {
      "epoch": 3.7568705414571397,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 16061
    },
    {
      "epoch": 3.7571044322301486,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 16062
    },
    {
      "epoch": 3.7573383230031574,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1859,
      "step": 16063
    },
    {
      "epoch": 3.7575722137761667,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 16064
    },
    {
      "epoch": 3.7578061045491755,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 16065
    },
    {
      "epoch": 3.758039995322185,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 16066
    },
    {
      "epoch": 3.7582738860951936,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 16067
    },
    {
      "epoch": 3.7585077768682025,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 16068
    },
    {
      "epoch": 3.7587416676412113,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 16069
    },
    {
      "epoch": 3.7589755584142206,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 16070
    },
    {
      "epoch": 3.7592094491872294,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7971,
      "step": 16071
    },
    {
      "epoch": 3.7594433399602387,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 16072
    },
    {
      "epoch": 3.7596772307332476,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 16073
    },
    {
      "epoch": 3.7599111215062564,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 16074
    },
    {
      "epoch": 3.7601450122792657,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 16075
    },
    {
      "epoch": 3.7603789030522745,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 16076
    },
    {
      "epoch": 3.760612793825284,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 16077
    },
    {
      "epoch": 3.7608466845982926,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 16078
    },
    {
      "epoch": 3.7610805753713015,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 16079
    },
    {
      "epoch": 3.7613144661443108,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 16080
    },
    {
      "epoch": 3.7615483569173196,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 16081
    },
    {
      "epoch": 3.761782247690329,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 16082
    },
    {
      "epoch": 3.7620161384633377,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7455,
      "step": 16083
    },
    {
      "epoch": 3.7622500292363465,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5622,
      "step": 16084
    },
    {
      "epoch": 3.7624839200093554,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 16085
    },
    {
      "epoch": 3.7627178107823647,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 16086
    },
    {
      "epoch": 3.7629517015553735,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5529,
      "step": 16087
    },
    {
      "epoch": 3.763185592328383,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 16088
    },
    {
      "epoch": 3.7634194831013916,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 16089
    },
    {
      "epoch": 3.7636533738744005,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 16090
    },
    {
      "epoch": 3.7638872646474097,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6354,
      "step": 16091
    },
    {
      "epoch": 3.7641211554204186,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 16092
    },
    {
      "epoch": 3.764355046193428,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 16093
    },
    {
      "epoch": 3.7645889369664367,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 16094
    },
    {
      "epoch": 3.7648228277394455,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7413,
      "step": 16095
    },
    {
      "epoch": 3.765056718512455,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4535,
      "step": 16096
    },
    {
      "epoch": 3.7652906092854637,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 16097
    },
    {
      "epoch": 3.765524500058473,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7289,
      "step": 16098
    },
    {
      "epoch": 3.765758390831482,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 16099
    },
    {
      "epoch": 3.7659922816044906,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 16100
    },
    {
      "epoch": 3.7659922816044906,
      "eval_runtime": 4.6827,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 16100
    },
    {
      "epoch": 3.7662261723774995,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.037,
      "step": 16101
    },
    {
      "epoch": 3.7664600631505087,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 16102
    },
    {
      "epoch": 3.7666939539235176,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 16103
    },
    {
      "epoch": 3.766927844696527,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8524,
      "step": 16104
    },
    {
      "epoch": 3.7671617354695357,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2009,
      "step": 16105
    },
    {
      "epoch": 3.7673956262425445,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6918,
      "step": 16106
    },
    {
      "epoch": 3.767629517015554,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 16107
    },
    {
      "epoch": 3.7678634077885627,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 16108
    },
    {
      "epoch": 3.768097298561572,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 16109
    },
    {
      "epoch": 3.7683311893345808,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 16110
    },
    {
      "epoch": 3.7685650801075896,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 16111
    },
    {
      "epoch": 3.768798970880599,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 16112
    },
    {
      "epoch": 3.7690328616536077,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 16113
    },
    {
      "epoch": 3.769266752426617,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 16114
    },
    {
      "epoch": 3.769500643199626,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 16115
    },
    {
      "epoch": 3.7697345339726347,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 16116
    },
    {
      "epoch": 3.7699684247456435,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 16117
    },
    {
      "epoch": 3.770202315518653,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 16118
    },
    {
      "epoch": 3.7704362062916617,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 16119
    },
    {
      "epoch": 3.770670097064671,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 16120
    },
    {
      "epoch": 3.7709039878376798,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 16121
    },
    {
      "epoch": 3.7711378786106886,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 16122
    },
    {
      "epoch": 3.771371769383698,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7651,
      "step": 16123
    },
    {
      "epoch": 3.7716056601567067,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 16124
    },
    {
      "epoch": 3.771839550929716,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 16125
    },
    {
      "epoch": 3.772073441702725,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 16126
    },
    {
      "epoch": 3.7723073324757337,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5972,
      "step": 16127
    },
    {
      "epoch": 3.772541223248743,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 16128
    },
    {
      "epoch": 3.772775114021752,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 16129
    },
    {
      "epoch": 3.773009004794761,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 16130
    },
    {
      "epoch": 3.77324289556777,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 16131
    },
    {
      "epoch": 3.7734767863407788,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 16132
    },
    {
      "epoch": 3.7737106771137876,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 16133
    },
    {
      "epoch": 3.773944567886797,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 16134
    },
    {
      "epoch": 3.7741784586598057,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 16135
    },
    {
      "epoch": 3.774412349432815,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 16136
    },
    {
      "epoch": 3.774646240205824,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 16137
    },
    {
      "epoch": 3.7748801309788327,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 16138
    },
    {
      "epoch": 3.775114021751842,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 16139
    },
    {
      "epoch": 3.775347912524851,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 16140
    },
    {
      "epoch": 3.77558180329786,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 16141
    },
    {
      "epoch": 3.775815694070869,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 16142
    },
    {
      "epoch": 3.7760495848438778,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 16143
    },
    {
      "epoch": 3.776283475616887,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 16144
    },
    {
      "epoch": 3.776517366389896,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 16145
    },
    {
      "epoch": 3.776751257162905,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 16146
    },
    {
      "epoch": 3.776985147935914,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 16147
    },
    {
      "epoch": 3.777219038708923,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 16148
    },
    {
      "epoch": 3.7774529294819317,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5796,
      "step": 16149
    },
    {
      "epoch": 3.777686820254941,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 16150
    },
    {
      "epoch": 3.77792071102795,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 16151
    },
    {
      "epoch": 3.778154601800959,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 16152
    },
    {
      "epoch": 3.778388492573968,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 16153
    },
    {
      "epoch": 3.7786223833469768,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 16154
    },
    {
      "epoch": 3.778856274119986,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6697,
      "step": 16155
    },
    {
      "epoch": 3.779090164892995,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 16156
    },
    {
      "epoch": 3.779324055666004,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5475,
      "step": 16157
    },
    {
      "epoch": 3.779557946439013,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 16158
    },
    {
      "epoch": 3.779791837212022,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7401,
      "step": 16159
    },
    {
      "epoch": 3.780025727985031,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 16160
    },
    {
      "epoch": 3.78025961875804,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6794,
      "step": 16161
    },
    {
      "epoch": 3.7804935095310492,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 16162
    },
    {
      "epoch": 3.780727400304058,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6747,
      "step": 16163
    },
    {
      "epoch": 3.780961291077067,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0059,
      "step": 16164
    },
    {
      "epoch": 3.7811951818500757,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8476,
      "step": 16165
    },
    {
      "epoch": 3.781429072623085,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 16166
    },
    {
      "epoch": 3.781662963396094,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 16167
    },
    {
      "epoch": 3.781896854169103,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 16168
    },
    {
      "epoch": 3.782130744942112,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 16169
    },
    {
      "epoch": 3.782364635715121,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 16170
    },
    {
      "epoch": 3.78259852648813,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 16171
    },
    {
      "epoch": 3.782832417261139,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 16172
    },
    {
      "epoch": 3.7830663080341482,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 16173
    },
    {
      "epoch": 3.783300198807157,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7624,
      "step": 16174
    },
    {
      "epoch": 3.783534089580166,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 16175
    },
    {
      "epoch": 3.783767980353175,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 16176
    },
    {
      "epoch": 3.784001871126184,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 16177
    },
    {
      "epoch": 3.7842357618991933,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 16178
    },
    {
      "epoch": 3.784469652672202,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 16179
    },
    {
      "epoch": 3.784703543445211,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 16180
    },
    {
      "epoch": 3.78493743421822,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 16181
    },
    {
      "epoch": 3.785171324991229,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2931,
      "step": 16182
    },
    {
      "epoch": 3.785405215764238,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7819,
      "step": 16183
    },
    {
      "epoch": 3.7856391065372472,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 16184
    },
    {
      "epoch": 3.785872997310256,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 16185
    },
    {
      "epoch": 3.786106888083265,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 16186
    },
    {
      "epoch": 3.786340778856274,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 16187
    },
    {
      "epoch": 3.786574669629283,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.586,
      "step": 16188
    },
    {
      "epoch": 3.7868085604022923,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 16189
    },
    {
      "epoch": 3.787042451175301,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 16190
    },
    {
      "epoch": 3.78727634194831,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 16191
    },
    {
      "epoch": 3.7875102327213193,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 16192
    },
    {
      "epoch": 3.787744123494328,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6157,
      "step": 16193
    },
    {
      "epoch": 3.7879780142673374,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5263,
      "step": 16194
    },
    {
      "epoch": 3.788211905040346,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5637,
      "step": 16195
    },
    {
      "epoch": 3.788445795813355,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 16196
    },
    {
      "epoch": 3.788679686586364,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 16197
    },
    {
      "epoch": 3.788913577359373,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 16198
    },
    {
      "epoch": 3.789147468132382,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 16199
    },
    {
      "epoch": 3.7893813589053913,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6016,
      "step": 16200
    },
    {
      "epoch": 3.7893813589053913,
      "eval_runtime": 4.7568,
      "eval_samples_per_second": 0.21,
      "eval_steps_per_second": 0.21,
      "step": 16200
    },
    {
      "epoch": 3.7896152496784,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 16201
    },
    {
      "epoch": 3.789849140451409,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7432,
      "step": 16202
    },
    {
      "epoch": 3.7900830312244183,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 16203
    },
    {
      "epoch": 3.790316921997427,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 16204
    },
    {
      "epoch": 3.7905508127704364,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7604,
      "step": 16205
    },
    {
      "epoch": 3.790784703543445,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 16206
    },
    {
      "epoch": 3.791018594316454,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 16207
    },
    {
      "epoch": 3.7912524850894633,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 16208
    },
    {
      "epoch": 3.791486375862472,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6454,
      "step": 16209
    },
    {
      "epoch": 3.7917202666354815,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6116,
      "step": 16210
    },
    {
      "epoch": 3.7919541574084903,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3989,
      "step": 16211
    },
    {
      "epoch": 3.792188048181499,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.832,
      "step": 16212
    },
    {
      "epoch": 3.7924219389545084,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 16213
    },
    {
      "epoch": 3.7926558297275172,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 16214
    },
    {
      "epoch": 3.7928897205005265,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.29,
      "step": 16215
    },
    {
      "epoch": 3.7931236112735354,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 16216
    },
    {
      "epoch": 3.793357502046544,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.5074,
      "step": 16217
    },
    {
      "epoch": 3.793591392819553,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5338,
      "step": 16218
    },
    {
      "epoch": 3.7938252835925623,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 16219
    },
    {
      "epoch": 3.794059174365571,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.114,
      "step": 16220
    },
    {
      "epoch": 3.7942930651385804,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 16221
    },
    {
      "epoch": 3.7945269559115893,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 16222
    },
    {
      "epoch": 3.794760846684598,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 16223
    },
    {
      "epoch": 3.7949947374576074,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 16224
    },
    {
      "epoch": 3.7952286282306162,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9676,
      "step": 16225
    },
    {
      "epoch": 3.7954625190036255,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 16226
    },
    {
      "epoch": 3.7956964097766344,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 16227
    },
    {
      "epoch": 3.795930300549643,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 16228
    },
    {
      "epoch": 3.7961641913226525,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 16229
    },
    {
      "epoch": 3.7963980820956613,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 16230
    },
    {
      "epoch": 3.7966319728686706,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 16231
    },
    {
      "epoch": 3.7968658636416794,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.8417,
      "step": 16232
    },
    {
      "epoch": 3.7970997544146883,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6473,
      "step": 16233
    },
    {
      "epoch": 3.797333645187697,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 16234
    },
    {
      "epoch": 3.7975675359607064,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 16235
    },
    {
      "epoch": 3.7978014267337152,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2115,
      "step": 16236
    },
    {
      "epoch": 3.7980353175067245,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5665,
      "step": 16237
    },
    {
      "epoch": 3.7982692082797334,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6345,
      "step": 16238
    },
    {
      "epoch": 3.798503099052742,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.6987,
      "step": 16239
    },
    {
      "epoch": 3.7987369898257515,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 16240
    },
    {
      "epoch": 3.7989708805987603,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 16241
    },
    {
      "epoch": 3.7992047713717696,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7442,
      "step": 16242
    },
    {
      "epoch": 3.7994386621447784,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 16243
    },
    {
      "epoch": 3.7996725529177873,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6295,
      "step": 16244
    },
    {
      "epoch": 3.7999064436907966,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 16245
    },
    {
      "epoch": 3.8001403344638054,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 16246
    },
    {
      "epoch": 3.8003742252368147,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 16247
    },
    {
      "epoch": 3.8006081160098235,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 16248
    },
    {
      "epoch": 3.8008420067828323,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 16249
    },
    {
      "epoch": 3.801075897555841,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 16250
    },
    {
      "epoch": 3.8013097883288505,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 16251
    },
    {
      "epoch": 3.8015436791018593,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 16252
    },
    {
      "epoch": 3.8017775698748686,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 16253
    },
    {
      "epoch": 3.8020114606478774,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 16254
    },
    {
      "epoch": 3.8022453514208863,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 16255
    },
    {
      "epoch": 3.8024792421938955,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 16256
    },
    {
      "epoch": 3.8027131329669044,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5862,
      "step": 16257
    },
    {
      "epoch": 3.8029470237399137,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 16258
    },
    {
      "epoch": 3.8031809145129225,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 16259
    },
    {
      "epoch": 3.8034148052859313,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5158,
      "step": 16260
    },
    {
      "epoch": 3.8036486960589406,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 16261
    },
    {
      "epoch": 3.8038825868319495,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 16262
    },
    {
      "epoch": 3.8041164776049587,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 16263
    },
    {
      "epoch": 3.8043503683779676,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7147,
      "step": 16264
    },
    {
      "epoch": 3.8045842591509764,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 16265
    },
    {
      "epoch": 3.8048181499239853,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 16266
    },
    {
      "epoch": 3.8050520406969945,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 16267
    },
    {
      "epoch": 3.8052859314700034,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 16268
    },
    {
      "epoch": 3.8055198222430127,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 16269
    },
    {
      "epoch": 3.8057537130160215,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 16270
    },
    {
      "epoch": 3.8059876037890303,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8748,
      "step": 16271
    },
    {
      "epoch": 3.8062214945620396,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 16272
    },
    {
      "epoch": 3.8064553853350485,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 16273
    },
    {
      "epoch": 3.8066892761080577,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 16274
    },
    {
      "epoch": 3.8069231668810666,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6481,
      "step": 16275
    },
    {
      "epoch": 3.8071570576540754,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 16276
    },
    {
      "epoch": 3.8073909484270847,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 16277
    },
    {
      "epoch": 3.8076248392000935,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 16278
    },
    {
      "epoch": 3.807858729973103,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 16279
    },
    {
      "epoch": 3.8080926207461117,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 16280
    },
    {
      "epoch": 3.8083265115191205,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 16281
    },
    {
      "epoch": 3.8085604022921293,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 16282
    },
    {
      "epoch": 3.8087942930651386,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5537,
      "step": 16283
    },
    {
      "epoch": 3.8090281838381475,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 16284
    },
    {
      "epoch": 3.8092620746111567,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 16285
    },
    {
      "epoch": 3.8094959653841656,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 16286
    },
    {
      "epoch": 3.8097298561571744,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7255,
      "step": 16287
    },
    {
      "epoch": 3.8099637469301837,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7576,
      "step": 16288
    },
    {
      "epoch": 3.8101976377031925,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 16289
    },
    {
      "epoch": 3.810431528476202,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 16290
    },
    {
      "epoch": 3.8106654192492107,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 16291
    },
    {
      "epoch": 3.8108993100222195,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 16292
    },
    {
      "epoch": 3.8111332007952288,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 16293
    },
    {
      "epoch": 3.8113670915682376,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 16294
    },
    {
      "epoch": 3.811600982341247,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 16295
    },
    {
      "epoch": 3.8118348731142557,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 16296
    },
    {
      "epoch": 3.8120687638872646,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.562,
      "step": 16297
    },
    {
      "epoch": 3.8123026546602734,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 16298
    },
    {
      "epoch": 3.8125365454332827,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 16299
    },
    {
      "epoch": 3.8127704362062915,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 16300
    },
    {
      "epoch": 3.8127704362062915,
      "eval_runtime": 4.6256,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 16300
    },
    {
      "epoch": 3.813004326979301,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 16301
    },
    {
      "epoch": 3.8132382177523096,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.951,
      "step": 16302
    },
    {
      "epoch": 3.8134721085253185,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 16303
    },
    {
      "epoch": 3.8137059992983278,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 16304
    },
    {
      "epoch": 3.8139398900713366,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6971,
      "step": 16305
    },
    {
      "epoch": 3.814173780844346,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 16306
    },
    {
      "epoch": 3.8144076716173547,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 16307
    },
    {
      "epoch": 3.8146415623903636,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 16308
    },
    {
      "epoch": 3.814875453163373,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 16309
    },
    {
      "epoch": 3.8151093439363817,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 16310
    },
    {
      "epoch": 3.815343234709391,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 16311
    },
    {
      "epoch": 3.8155771254824,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6033,
      "step": 16312
    },
    {
      "epoch": 3.8158110162554086,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7238,
      "step": 16313
    },
    {
      "epoch": 3.8160449070284175,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 16314
    },
    {
      "epoch": 3.8162787978014268,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5222,
      "step": 16315
    },
    {
      "epoch": 3.8165126885744356,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 16316
    },
    {
      "epoch": 3.816746579347445,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 16317
    },
    {
      "epoch": 3.8169804701204537,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 16318
    },
    {
      "epoch": 3.8172143608934626,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 16319
    },
    {
      "epoch": 3.817448251666472,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 16320
    },
    {
      "epoch": 3.8176821424394807,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 16321
    },
    {
      "epoch": 3.81791603321249,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 16322
    },
    {
      "epoch": 3.818149923985499,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 16323
    },
    {
      "epoch": 3.8183838147585076,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 16324
    },
    {
      "epoch": 3.818617705531517,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 16325
    },
    {
      "epoch": 3.8188515963045258,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 16326
    },
    {
      "epoch": 3.819085487077535,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 16327
    },
    {
      "epoch": 3.819319377850544,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 16328
    },
    {
      "epoch": 3.8195532686235527,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 16329
    },
    {
      "epoch": 3.8197871593965615,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9724,
      "step": 16330
    },
    {
      "epoch": 3.820021050169571,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5925,
      "step": 16331
    },
    {
      "epoch": 3.8202549409425797,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 16332
    },
    {
      "epoch": 3.820488831715589,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 16333
    },
    {
      "epoch": 3.820722722488598,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 16334
    },
    {
      "epoch": 3.8209566132616066,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 16335
    },
    {
      "epoch": 3.821190504034616,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 16336
    },
    {
      "epoch": 3.8214243948076247,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7741,
      "step": 16337
    },
    {
      "epoch": 3.821658285580634,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 16338
    },
    {
      "epoch": 3.821892176353643,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 16339
    },
    {
      "epoch": 3.8221260671266517,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 16340
    },
    {
      "epoch": 3.822359957899661,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 16341
    },
    {
      "epoch": 3.82259384867267,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5195,
      "step": 16342
    },
    {
      "epoch": 3.822827739445679,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 16343
    },
    {
      "epoch": 3.823061630218688,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 16344
    },
    {
      "epoch": 3.823295520991697,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8151,
      "step": 16345
    },
    {
      "epoch": 3.8235294117647056,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 16346
    },
    {
      "epoch": 3.823763302537715,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 16347
    },
    {
      "epoch": 3.8239971933107237,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 16348
    },
    {
      "epoch": 3.824231084083733,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9912,
      "step": 16349
    },
    {
      "epoch": 3.824464974856742,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7101,
      "step": 16350
    },
    {
      "epoch": 3.8246988656297507,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 16351
    },
    {
      "epoch": 3.82493275640276,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2082,
      "step": 16352
    },
    {
      "epoch": 3.825166647175769,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 16353
    },
    {
      "epoch": 3.825400537948778,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 16354
    },
    {
      "epoch": 3.825634428721787,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 16355
    },
    {
      "epoch": 3.8258683194947958,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 16356
    },
    {
      "epoch": 3.826102210267805,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 16357
    },
    {
      "epoch": 3.826336101040814,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7999,
      "step": 16358
    },
    {
      "epoch": 3.826569991813823,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 16359
    },
    {
      "epoch": 3.826803882586832,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 16360
    },
    {
      "epoch": 3.827037773359841,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 16361
    },
    {
      "epoch": 3.8272716641328497,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 16362
    },
    {
      "epoch": 3.827505554905859,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 16363
    },
    {
      "epoch": 3.827739445678868,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 16364
    },
    {
      "epoch": 3.827973336451877,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8058,
      "step": 16365
    },
    {
      "epoch": 3.828207227224886,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 16366
    },
    {
      "epoch": 3.8284411179978948,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 16367
    },
    {
      "epoch": 3.828675008770904,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 16368
    },
    {
      "epoch": 3.828908899543913,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 16369
    },
    {
      "epoch": 3.829142790316922,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8096,
      "step": 16370
    },
    {
      "epoch": 3.829376681089931,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.885,
      "step": 16371
    },
    {
      "epoch": 3.82961057186294,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 16372
    },
    {
      "epoch": 3.829844462635949,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 16373
    },
    {
      "epoch": 3.830078353408958,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2806,
      "step": 16374
    },
    {
      "epoch": 3.8303122441819673,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9513,
      "step": 16375
    },
    {
      "epoch": 3.830546134954976,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 16376
    },
    {
      "epoch": 3.830780025727985,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 16377
    },
    {
      "epoch": 3.8310139165009938,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 16378
    },
    {
      "epoch": 3.831247807274003,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 16379
    },
    {
      "epoch": 3.831481698047012,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 16380
    },
    {
      "epoch": 3.831715588820021,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 16381
    },
    {
      "epoch": 3.83194947959303,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7745,
      "step": 16382
    },
    {
      "epoch": 3.832183370366039,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 16383
    },
    {
      "epoch": 3.832417261139048,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 16384
    },
    {
      "epoch": 3.832651151912057,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6524,
      "step": 16385
    },
    {
      "epoch": 3.8328850426850662,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 16386
    },
    {
      "epoch": 3.833118933458075,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 16387
    },
    {
      "epoch": 3.833352824231084,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 16388
    },
    {
      "epoch": 3.833586715004093,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 16389
    },
    {
      "epoch": 3.833820605777102,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 16390
    },
    {
      "epoch": 3.8340544965501113,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 16391
    },
    {
      "epoch": 3.83428838732312,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 16392
    },
    {
      "epoch": 3.834522278096129,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7495,
      "step": 16393
    },
    {
      "epoch": 3.834756168869138,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5355,
      "step": 16394
    },
    {
      "epoch": 3.834990059642147,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6361,
      "step": 16395
    },
    {
      "epoch": 3.835223950415156,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7923,
      "step": 16396
    },
    {
      "epoch": 3.8354578411881652,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 16397
    },
    {
      "epoch": 3.835691731961174,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6868,
      "step": 16398
    },
    {
      "epoch": 3.835925622734183,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 16399
    },
    {
      "epoch": 3.836159513507192,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 16400
    },
    {
      "epoch": 3.836159513507192,
      "eval_runtime": 4.6504,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 16400
    },
    {
      "epoch": 3.836393404280201,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 16401
    },
    {
      "epoch": 3.8366272950532103,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 16402
    },
    {
      "epoch": 3.836861185826219,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 16403
    },
    {
      "epoch": 3.837095076599228,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 16404
    },
    {
      "epoch": 3.8373289673722373,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 16405
    },
    {
      "epoch": 3.837562858145246,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 16406
    },
    {
      "epoch": 3.8377967489182554,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 16407
    },
    {
      "epoch": 3.8380306396912642,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 16408
    },
    {
      "epoch": 3.838264530464273,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 16409
    },
    {
      "epoch": 3.8384984212372824,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6785,
      "step": 16410
    },
    {
      "epoch": 3.838732312010291,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 16411
    },
    {
      "epoch": 3.8389662027833005,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 16412
    },
    {
      "epoch": 3.8392000935563093,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 16413
    },
    {
      "epoch": 3.839433984329318,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 16414
    },
    {
      "epoch": 3.839667875102327,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3832,
      "step": 16415
    },
    {
      "epoch": 3.8399017658753363,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 16416
    },
    {
      "epoch": 3.840135656648345,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 16417
    },
    {
      "epoch": 3.8403695474213544,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 16418
    },
    {
      "epoch": 3.8406034381943632,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 16419
    },
    {
      "epoch": 3.840837328967372,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 16420
    },
    {
      "epoch": 3.8410712197403813,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 16421
    },
    {
      "epoch": 3.84130511051339,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.5459,
      "step": 16422
    },
    {
      "epoch": 3.8415390012863995,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 16423
    },
    {
      "epoch": 3.8417728920594083,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 16424
    },
    {
      "epoch": 3.842006782832417,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6312,
      "step": 16425
    },
    {
      "epoch": 3.8422406736054264,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 16426
    },
    {
      "epoch": 3.8424745643784353,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3535,
      "step": 16427
    },
    {
      "epoch": 3.8427084551514445,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 16428
    },
    {
      "epoch": 3.8429423459244534,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7772,
      "step": 16429
    },
    {
      "epoch": 3.8431762366974622,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 16430
    },
    {
      "epoch": 3.843410127470471,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 16431
    },
    {
      "epoch": 3.8436440182434803,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 16432
    },
    {
      "epoch": 3.843877909016489,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 16433
    },
    {
      "epoch": 3.8441117997894985,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.623,
      "step": 16434
    },
    {
      "epoch": 3.8443456905625073,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 16435
    },
    {
      "epoch": 3.844579581335516,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 16436
    },
    {
      "epoch": 3.8448134721085254,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 16437
    },
    {
      "epoch": 3.8450473628815343,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 16438
    },
    {
      "epoch": 3.8452812536545435,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8347,
      "step": 16439
    },
    {
      "epoch": 3.8455151444275524,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 16440
    },
    {
      "epoch": 3.845749035200561,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 16441
    },
    {
      "epoch": 3.8459829259735705,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8726,
      "step": 16442
    },
    {
      "epoch": 3.8462168167465793,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 16443
    },
    {
      "epoch": 3.8464507075195886,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 16444
    },
    {
      "epoch": 3.8466845982925975,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3327,
      "step": 16445
    },
    {
      "epoch": 3.8469184890656063,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7625,
      "step": 16446
    },
    {
      "epoch": 3.847152379838615,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6986,
      "step": 16447
    },
    {
      "epoch": 3.8473862706116244,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 16448
    },
    {
      "epoch": 3.8476201613846333,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 16449
    },
    {
      "epoch": 3.8478540521576425,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 16450
    },
    {
      "epoch": 3.8480879429306514,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.23,
      "step": 16451
    },
    {
      "epoch": 3.84832183370366,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 16452
    },
    {
      "epoch": 3.8485557244766695,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 16453
    },
    {
      "epoch": 3.8487896152496783,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 16454
    },
    {
      "epoch": 3.8490235060226876,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 16455
    },
    {
      "epoch": 3.8492573967956965,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 16456
    },
    {
      "epoch": 3.8494912875687053,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 16457
    },
    {
      "epoch": 3.8497251783417146,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 16458
    },
    {
      "epoch": 3.8499590691147234,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 16459
    },
    {
      "epoch": 3.8501929598877327,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 16460
    },
    {
      "epoch": 3.8504268506607415,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6573,
      "step": 16461
    },
    {
      "epoch": 3.8506607414337504,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4711,
      "step": 16462
    },
    {
      "epoch": 3.850894632206759,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 16463
    },
    {
      "epoch": 3.8511285229797685,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 16464
    },
    {
      "epoch": 3.8513624137527773,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 16465
    },
    {
      "epoch": 3.8515963045257866,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 16466
    },
    {
      "epoch": 3.8518301952987954,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 16467
    },
    {
      "epoch": 3.8520640860718043,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 16468
    },
    {
      "epoch": 3.8522979768448136,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 16469
    },
    {
      "epoch": 3.8525318676178224,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 16470
    },
    {
      "epoch": 3.8527657583908317,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8385,
      "step": 16471
    },
    {
      "epoch": 3.8529996491638405,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 16472
    },
    {
      "epoch": 3.8532335399368494,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 16473
    },
    {
      "epoch": 3.8534674307098586,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 16474
    },
    {
      "epoch": 3.8537013214828675,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 16475
    },
    {
      "epoch": 3.8539352122558768,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 16476
    },
    {
      "epoch": 3.8541691030288856,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 16477
    },
    {
      "epoch": 3.8544029938018944,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 16478
    },
    {
      "epoch": 3.8546368845749033,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 16479
    },
    {
      "epoch": 3.8548707753479126,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7135,
      "step": 16480
    },
    {
      "epoch": 3.8551046661209214,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 16481
    },
    {
      "epoch": 3.8553385568939307,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 16482
    },
    {
      "epoch": 3.8555724476669395,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5638,
      "step": 16483
    },
    {
      "epoch": 3.8558063384399484,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 16484
    },
    {
      "epoch": 3.8560402292129576,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7115,
      "step": 16485
    },
    {
      "epoch": 3.8562741199859665,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8303,
      "step": 16486
    },
    {
      "epoch": 3.8565080107589758,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2241,
      "step": 16487
    },
    {
      "epoch": 3.8567419015319846,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 16488
    },
    {
      "epoch": 3.8569757923049934,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 16489
    },
    {
      "epoch": 3.8572096830780027,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 16490
    },
    {
      "epoch": 3.8574435738510116,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3565,
      "step": 16491
    },
    {
      "epoch": 3.857677464624021,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 16492
    },
    {
      "epoch": 3.8579113553970297,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.5422,
      "step": 16493
    },
    {
      "epoch": 3.8581452461700385,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 16494
    },
    {
      "epoch": 3.8583791369430473,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 16495
    },
    {
      "epoch": 3.8586130277160566,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 16496
    },
    {
      "epoch": 3.8588469184890655,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6501,
      "step": 16497
    },
    {
      "epoch": 3.8590808092620748,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 16498
    },
    {
      "epoch": 3.8593147000350836,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 16499
    },
    {
      "epoch": 3.8595485908080924,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5541,
      "step": 16500
    },
    {
      "epoch": 3.8595485908080924,
      "eval_runtime": 4.6289,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 16500
    },
    {
      "epoch": 3.8597824815811017,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 16501
    },
    {
      "epoch": 3.8600163723541105,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 16502
    },
    {
      "epoch": 3.86025026312712,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 16503
    },
    {
      "epoch": 3.8604841539001287,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5959,
      "step": 16504
    },
    {
      "epoch": 3.8607180446731375,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 16505
    },
    {
      "epoch": 3.860951935446147,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5773,
      "step": 16506
    },
    {
      "epoch": 3.8611858262191556,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 16507
    },
    {
      "epoch": 3.861419716992165,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8528,
      "step": 16508
    },
    {
      "epoch": 3.8616536077651737,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 16509
    },
    {
      "epoch": 3.8618874985381826,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 16510
    },
    {
      "epoch": 3.8621213893111914,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 16511
    },
    {
      "epoch": 3.8623552800842007,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 16512
    },
    {
      "epoch": 3.8625891708572095,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 16513
    },
    {
      "epoch": 3.862823061630219,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 16514
    },
    {
      "epoch": 3.8630569524032277,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 16515
    },
    {
      "epoch": 3.8632908431762365,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 16516
    },
    {
      "epoch": 3.863524733949246,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 16517
    },
    {
      "epoch": 3.8637586247222546,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7333,
      "step": 16518
    },
    {
      "epoch": 3.863992515495264,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 16519
    },
    {
      "epoch": 3.8642264062682727,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 16520
    },
    {
      "epoch": 3.8644602970412816,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 16521
    },
    {
      "epoch": 3.864694187814291,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 16522
    },
    {
      "epoch": 3.8649280785872997,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 16523
    },
    {
      "epoch": 3.865161969360309,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 16524
    },
    {
      "epoch": 3.865395860133318,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 16525
    },
    {
      "epoch": 3.8656297509063267,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 16526
    },
    {
      "epoch": 3.8658636416793355,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 16527
    },
    {
      "epoch": 3.8660975324523448,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7161,
      "step": 16528
    },
    {
      "epoch": 3.8663314232253536,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 16529
    },
    {
      "epoch": 3.866565313998363,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 16530
    },
    {
      "epoch": 3.8667992047713717,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 16531
    },
    {
      "epoch": 3.8670330955443806,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 16532
    },
    {
      "epoch": 3.86726698631739,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 16533
    },
    {
      "epoch": 3.8675008770903987,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 16534
    },
    {
      "epoch": 3.867734767863408,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 16535
    },
    {
      "epoch": 3.867968658636417,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 16536
    },
    {
      "epoch": 3.8682025494094257,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6651,
      "step": 16537
    },
    {
      "epoch": 3.868436440182435,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 16538
    },
    {
      "epoch": 3.8686703309554438,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 16539
    },
    {
      "epoch": 3.868904221728453,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 16540
    },
    {
      "epoch": 3.869138112501462,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 16541
    },
    {
      "epoch": 3.8693720032744707,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 16542
    },
    {
      "epoch": 3.8696058940474796,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 16543
    },
    {
      "epoch": 3.869839784820489,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 16544
    },
    {
      "epoch": 3.8700736755934977,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.3665,
      "step": 16545
    },
    {
      "epoch": 3.870307566366507,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 16546
    },
    {
      "epoch": 3.870541457139516,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7999,
      "step": 16547
    },
    {
      "epoch": 3.8707753479125246,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 16548
    },
    {
      "epoch": 3.871009238685534,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 16549
    },
    {
      "epoch": 3.8712431294585428,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5285,
      "step": 16550
    },
    {
      "epoch": 3.871477020231552,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7772,
      "step": 16551
    },
    {
      "epoch": 3.871710911004561,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 16552
    },
    {
      "epoch": 3.8719448017775697,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 16553
    },
    {
      "epoch": 3.872178692550579,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1457,
      "step": 16554
    },
    {
      "epoch": 3.872412583323588,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 16555
    },
    {
      "epoch": 3.872646474096597,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 16556
    },
    {
      "epoch": 3.872880364869606,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 16557
    },
    {
      "epoch": 3.873114255642615,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 16558
    },
    {
      "epoch": 3.8733481464156236,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 16559
    },
    {
      "epoch": 3.873582037188633,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 16560
    },
    {
      "epoch": 3.8738159279616418,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 16561
    },
    {
      "epoch": 3.874049818734651,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 16562
    },
    {
      "epoch": 3.87428370950766,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 16563
    },
    {
      "epoch": 3.8745176002806687,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 16564
    },
    {
      "epoch": 3.874751491053678,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 16565
    },
    {
      "epoch": 3.874985381826687,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 16566
    },
    {
      "epoch": 3.875219272599696,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 16567
    },
    {
      "epoch": 3.875453163372705,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7049,
      "step": 16568
    },
    {
      "epoch": 3.875687054145714,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 16569
    },
    {
      "epoch": 3.875920944918723,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 16570
    },
    {
      "epoch": 3.876154835691732,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6639,
      "step": 16571
    },
    {
      "epoch": 3.876388726464741,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 16572
    },
    {
      "epoch": 3.87662261723775,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 16573
    },
    {
      "epoch": 3.876856508010759,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 16574
    },
    {
      "epoch": 3.8770903987837677,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 16575
    },
    {
      "epoch": 3.877324289556777,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 16576
    },
    {
      "epoch": 3.877558180329786,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 16577
    },
    {
      "epoch": 3.877792071102795,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.838,
      "step": 16578
    },
    {
      "epoch": 3.878025961875804,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7369,
      "step": 16579
    },
    {
      "epoch": 3.878259852648813,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 16580
    },
    {
      "epoch": 3.878493743421822,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 16581
    },
    {
      "epoch": 3.878727634194831,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 2.1867,
      "step": 16582
    },
    {
      "epoch": 3.87896152496784,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7503,
      "step": 16583
    },
    {
      "epoch": 3.879195415740849,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 16584
    },
    {
      "epoch": 3.879429306513858,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 16585
    },
    {
      "epoch": 3.879663197286867,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 16586
    },
    {
      "epoch": 3.879897088059876,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 16587
    },
    {
      "epoch": 3.8801309788328853,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 16588
    },
    {
      "epoch": 3.880364869605894,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 16589
    },
    {
      "epoch": 3.880598760378903,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 16590
    },
    {
      "epoch": 3.880832651151912,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7059,
      "step": 16591
    },
    {
      "epoch": 3.881066541924921,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 16592
    },
    {
      "epoch": 3.88130043269793,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 16593
    },
    {
      "epoch": 3.881534323470939,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 16594
    },
    {
      "epoch": 3.881768214243948,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 16595
    },
    {
      "epoch": 3.882002105016957,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 16596
    },
    {
      "epoch": 3.882235995789966,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 16597
    },
    {
      "epoch": 3.882469886562975,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 16598
    },
    {
      "epoch": 3.8827037773359843,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 16599
    },
    {
      "epoch": 3.882937668108993,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 16600
    },
    {
      "epoch": 3.882937668108993,
      "eval_runtime": 4.6464,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 16600
    },
    {
      "epoch": 3.883171558882002,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 16601
    },
    {
      "epoch": 3.883405449655011,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 16602
    },
    {
      "epoch": 3.88363934042802,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 16603
    },
    {
      "epoch": 3.8838732312010293,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 16604
    },
    {
      "epoch": 3.884107121974038,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 16605
    },
    {
      "epoch": 3.884341012747047,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 16606
    },
    {
      "epoch": 3.8845749035200563,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 16607
    },
    {
      "epoch": 3.884808794293065,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5208,
      "step": 16608
    },
    {
      "epoch": 3.8850426850660744,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 16609
    },
    {
      "epoch": 3.8852765758390833,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 16610
    },
    {
      "epoch": 3.885510466612092,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 16611
    },
    {
      "epoch": 3.885744357385101,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5844,
      "step": 16612
    },
    {
      "epoch": 3.88597824815811,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 16613
    },
    {
      "epoch": 3.886212138931119,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 16614
    },
    {
      "epoch": 3.8864460297041283,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 16615
    },
    {
      "epoch": 3.886679920477137,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7474,
      "step": 16616
    },
    {
      "epoch": 3.886913811250146,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 16617
    },
    {
      "epoch": 3.8871477020231553,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 16618
    },
    {
      "epoch": 3.887381592796164,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 2.3884,
      "step": 16619
    },
    {
      "epoch": 3.8876154835691734,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 16620
    },
    {
      "epoch": 3.8878493743421823,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 16621
    },
    {
      "epoch": 3.888083265115191,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6964,
      "step": 16622
    },
    {
      "epoch": 3.8883171558882004,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7075,
      "step": 16623
    },
    {
      "epoch": 3.888551046661209,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6032,
      "step": 16624
    },
    {
      "epoch": 3.8887849374342185,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 16625
    },
    {
      "epoch": 3.8890188282072273,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.7068,
      "step": 16626
    },
    {
      "epoch": 3.889252718980236,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 16627
    },
    {
      "epoch": 3.889486609753245,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 16628
    },
    {
      "epoch": 3.8897205005262543,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9221,
      "step": 16629
    },
    {
      "epoch": 3.889954391299263,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 16630
    },
    {
      "epoch": 3.8901882820722724,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 16631
    },
    {
      "epoch": 3.8904221728452812,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 16632
    },
    {
      "epoch": 3.89065606361829,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 16633
    },
    {
      "epoch": 3.8908899543912994,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 16634
    },
    {
      "epoch": 3.891123845164308,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7161,
      "step": 16635
    },
    {
      "epoch": 3.8913577359373175,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8416,
      "step": 16636
    },
    {
      "epoch": 3.8915916267103263,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 16637
    },
    {
      "epoch": 3.891825517483335,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 16638
    },
    {
      "epoch": 3.8920594082563444,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 16639
    },
    {
      "epoch": 3.8922932990293533,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 16640
    },
    {
      "epoch": 3.8925271898023626,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 16641
    },
    {
      "epoch": 3.8927610805753714,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 16642
    },
    {
      "epoch": 3.8929949713483802,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 16643
    },
    {
      "epoch": 3.893228862121389,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 16644
    },
    {
      "epoch": 3.8934627528943984,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 16645
    },
    {
      "epoch": 3.893696643667407,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 16646
    },
    {
      "epoch": 3.8939305344404165,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 16647
    },
    {
      "epoch": 3.8941644252134253,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 16648
    },
    {
      "epoch": 3.894398315986434,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 16649
    },
    {
      "epoch": 3.8946322067594434,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 16650
    },
    {
      "epoch": 3.8948660975324523,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 16651
    },
    {
      "epoch": 3.8950999883054616,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 16652
    },
    {
      "epoch": 3.8953338790784704,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0599,
      "step": 16653
    },
    {
      "epoch": 3.8955677698514792,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 16654
    },
    {
      "epoch": 3.8958016606244885,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3406,
      "step": 16655
    },
    {
      "epoch": 3.8960355513974974,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 16656
    },
    {
      "epoch": 3.8962694421705066,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 16657
    },
    {
      "epoch": 3.8965033329435155,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.988,
      "step": 16658
    },
    {
      "epoch": 3.8967372237165243,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 16659
    },
    {
      "epoch": 3.896971114489533,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 16660
    },
    {
      "epoch": 3.8972050052625424,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8007,
      "step": 16661
    },
    {
      "epoch": 3.8974388960355513,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 16662
    },
    {
      "epoch": 3.8976727868085606,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7789,
      "step": 16663
    },
    {
      "epoch": 3.8979066775815694,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.5741,
      "step": 16664
    },
    {
      "epoch": 3.8981405683545782,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 16665
    },
    {
      "epoch": 3.8983744591275875,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 16666
    },
    {
      "epoch": 3.8986083499005963,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 16667
    },
    {
      "epoch": 3.8988422406736056,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 16668
    },
    {
      "epoch": 3.8990761314466145,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 16669
    },
    {
      "epoch": 3.8993100222196233,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 16670
    },
    {
      "epoch": 3.8995439129926326,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 16671
    },
    {
      "epoch": 3.8997778037656414,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 16672
    },
    {
      "epoch": 3.9000116945386507,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7023,
      "step": 16673
    },
    {
      "epoch": 3.9002455853116595,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 16674
    },
    {
      "epoch": 3.9004794760846684,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 16675
    },
    {
      "epoch": 3.9007133668576772,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 16676
    },
    {
      "epoch": 3.9009472576306865,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 16677
    },
    {
      "epoch": 3.9011811484036953,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 16678
    },
    {
      "epoch": 3.9014150391767046,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 16679
    },
    {
      "epoch": 3.9016489299497135,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 16680
    },
    {
      "epoch": 3.9018828207227223,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 16681
    },
    {
      "epoch": 3.9021167114957316,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 16682
    },
    {
      "epoch": 3.9023506022687404,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 16683
    },
    {
      "epoch": 3.9025844930417497,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 16684
    },
    {
      "epoch": 3.9028183838147585,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 16685
    },
    {
      "epoch": 3.9030522745877674,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8815,
      "step": 16686
    },
    {
      "epoch": 3.9032861653607767,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6681,
      "step": 16687
    },
    {
      "epoch": 3.9035200561337855,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 16688
    },
    {
      "epoch": 3.903753946906795,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 16689
    },
    {
      "epoch": 3.9039878376798036,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 16690
    },
    {
      "epoch": 3.9042217284528125,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 16691
    },
    {
      "epoch": 3.9044556192258213,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 16692
    },
    {
      "epoch": 3.9046895099988306,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 16693
    },
    {
      "epoch": 3.9049234007718394,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 16694
    },
    {
      "epoch": 3.9051572915448487,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 16695
    },
    {
      "epoch": 3.9053911823178575,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 16696
    },
    {
      "epoch": 3.9056250730908664,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 16697
    },
    {
      "epoch": 3.9058589638638757,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 16698
    },
    {
      "epoch": 3.9060928546368845,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 16699
    },
    {
      "epoch": 3.9063267454098938,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 16700
    },
    {
      "epoch": 3.9063267454098938,
      "eval_runtime": 4.59,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 16700
    },
    {
      "epoch": 3.9065606361829026,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 16701
    },
    {
      "epoch": 3.9067945269559115,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2432,
      "step": 16702
    },
    {
      "epoch": 3.9070284177289207,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 16703
    },
    {
      "epoch": 3.9072623085019296,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 16704
    },
    {
      "epoch": 3.907496199274939,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 16705
    },
    {
      "epoch": 3.9077300900479477,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 16706
    },
    {
      "epoch": 3.9079639808209565,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 16707
    },
    {
      "epoch": 3.9081978715939654,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 16708
    },
    {
      "epoch": 3.9084317623669746,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 16709
    },
    {
      "epoch": 3.9086656531399835,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 16710
    },
    {
      "epoch": 3.9088995439129928,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 16711
    },
    {
      "epoch": 3.9091334346860016,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 16712
    },
    {
      "epoch": 3.9093673254590104,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4729,
      "step": 16713
    },
    {
      "epoch": 3.9096012162320197,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 16714
    },
    {
      "epoch": 3.9098351070050286,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 16715
    },
    {
      "epoch": 3.910068997778038,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 16716
    },
    {
      "epoch": 3.9103028885510467,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 16717
    },
    {
      "epoch": 3.9105367793240555,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 16718
    },
    {
      "epoch": 3.910770670097065,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.4813,
      "step": 16719
    },
    {
      "epoch": 3.9110045608700736,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 16720
    },
    {
      "epoch": 3.911238451643083,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 16721
    },
    {
      "epoch": 3.9114723424160918,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 16722
    },
    {
      "epoch": 3.9117062331891006,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 16723
    },
    {
      "epoch": 3.9119401239621094,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 16724
    },
    {
      "epoch": 3.9121740147351187,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7803,
      "step": 16725
    },
    {
      "epoch": 3.9124079055081276,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 16726
    },
    {
      "epoch": 3.912641796281137,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 16727
    },
    {
      "epoch": 3.9128756870541457,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6997,
      "step": 16728
    },
    {
      "epoch": 3.9131095778271545,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 16729
    },
    {
      "epoch": 3.913343468600164,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6365,
      "step": 16730
    },
    {
      "epoch": 3.9135773593731726,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6029,
      "step": 16731
    },
    {
      "epoch": 3.913811250146182,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 16732
    },
    {
      "epoch": 3.9140451409191908,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9786,
      "step": 16733
    },
    {
      "epoch": 3.9142790316921996,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 16734
    },
    {
      "epoch": 3.914512922465209,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 16735
    },
    {
      "epoch": 3.9147468132382177,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 16736
    },
    {
      "epoch": 3.914980704011227,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 16737
    },
    {
      "epoch": 3.915214594784236,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 16738
    },
    {
      "epoch": 3.9154484855572447,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7914,
      "step": 16739
    },
    {
      "epoch": 3.9156823763302535,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.489,
      "step": 16740
    },
    {
      "epoch": 3.915916267103263,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 16741
    },
    {
      "epoch": 3.9161501578762716,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 16742
    },
    {
      "epoch": 3.916384048649281,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 16743
    },
    {
      "epoch": 3.9166179394222898,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 16744
    },
    {
      "epoch": 3.9168518301952986,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 16745
    },
    {
      "epoch": 3.917085720968308,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 16746
    },
    {
      "epoch": 3.9173196117413167,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 16747
    },
    {
      "epoch": 3.917553502514326,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6249,
      "step": 16748
    },
    {
      "epoch": 3.917787393287335,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 16749
    },
    {
      "epoch": 3.9180212840603437,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5878,
      "step": 16750
    },
    {
      "epoch": 3.918255174833353,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 16751
    },
    {
      "epoch": 3.918489065606362,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 16752
    },
    {
      "epoch": 3.918722956379371,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 16753
    },
    {
      "epoch": 3.91895684715238,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 16754
    },
    {
      "epoch": 3.9191907379253887,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7173,
      "step": 16755
    },
    {
      "epoch": 3.9194246286983976,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3931,
      "step": 16756
    },
    {
      "epoch": 3.919658519471407,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 16757
    },
    {
      "epoch": 3.9198924102444157,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 16758
    },
    {
      "epoch": 3.920126301017425,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 16759
    },
    {
      "epoch": 3.920360191790434,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 16760
    },
    {
      "epoch": 3.9205940825634427,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 16761
    },
    {
      "epoch": 3.920827973336452,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 16762
    },
    {
      "epoch": 3.921061864109461,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 16763
    },
    {
      "epoch": 3.92129575488247,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 16764
    },
    {
      "epoch": 3.921529645655479,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6579,
      "step": 16765
    },
    {
      "epoch": 3.9217635364284877,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6579,
      "step": 16766
    },
    {
      "epoch": 3.921997427201497,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 16767
    },
    {
      "epoch": 3.922231317974506,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 16768
    },
    {
      "epoch": 3.922465208747515,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 16769
    },
    {
      "epoch": 3.922699099520524,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 16770
    },
    {
      "epoch": 3.922932990293533,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5913,
      "step": 16771
    },
    {
      "epoch": 3.9231668810665417,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 16772
    },
    {
      "epoch": 3.923400771839551,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 16773
    },
    {
      "epoch": 3.9236346626125598,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 16774
    },
    {
      "epoch": 3.923868553385569,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 16775
    },
    {
      "epoch": 3.924102444158578,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6507,
      "step": 16776
    },
    {
      "epoch": 3.9243363349315867,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 16777
    },
    {
      "epoch": 3.924570225704596,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 16778
    },
    {
      "epoch": 3.924804116477605,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7845,
      "step": 16779
    },
    {
      "epoch": 3.925038007250614,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 16780
    },
    {
      "epoch": 3.925271898023623,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 16781
    },
    {
      "epoch": 3.925505788796632,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 16782
    },
    {
      "epoch": 3.925739679569641,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 16783
    },
    {
      "epoch": 3.92597357034265,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 16784
    },
    {
      "epoch": 3.926207461115659,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 16785
    },
    {
      "epoch": 3.926441351888668,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 16786
    },
    {
      "epoch": 3.926675242661677,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 16787
    },
    {
      "epoch": 3.9269091334346857,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 16788
    },
    {
      "epoch": 3.927143024207695,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 16789
    },
    {
      "epoch": 3.927376914980704,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 16790
    },
    {
      "epoch": 3.927610805753713,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6546,
      "step": 16791
    },
    {
      "epoch": 3.927844696526722,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 16792
    },
    {
      "epoch": 3.928078587299731,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 16793
    },
    {
      "epoch": 3.92831247807274,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 16794
    },
    {
      "epoch": 3.928546368845749,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6536,
      "step": 16795
    },
    {
      "epoch": 3.928780259618758,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 16796
    },
    {
      "epoch": 3.929014150391767,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 16797
    },
    {
      "epoch": 3.929248041164776,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 16798
    },
    {
      "epoch": 3.929481931937785,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 16799
    },
    {
      "epoch": 3.929715822710794,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8308,
      "step": 16800
    },
    {
      "epoch": 3.929715822710794,
      "eval_runtime": 4.6026,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 16800
    },
    {
      "epoch": 3.9299497134838033,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 16801
    },
    {
      "epoch": 3.930183604256812,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 16802
    },
    {
      "epoch": 3.930417495029821,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 16803
    },
    {
      "epoch": 3.9306513858028302,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 16804
    },
    {
      "epoch": 3.930885276575839,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 16805
    },
    {
      "epoch": 3.9311191673488484,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 16806
    },
    {
      "epoch": 3.931353058121857,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 16807
    },
    {
      "epoch": 3.931586948894866,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 16808
    },
    {
      "epoch": 3.931820839667875,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 16809
    },
    {
      "epoch": 3.932054730440884,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 16810
    },
    {
      "epoch": 3.932288621213893,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 16811
    },
    {
      "epoch": 3.9325225119869023,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8093,
      "step": 16812
    },
    {
      "epoch": 3.932756402759911,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 16813
    },
    {
      "epoch": 3.93299029353292,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 16814
    },
    {
      "epoch": 3.9332241843059292,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 16815
    },
    {
      "epoch": 3.933458075078938,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7643,
      "step": 16816
    },
    {
      "epoch": 3.9336919658519474,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 16817
    },
    {
      "epoch": 3.933925856624956,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 16818
    },
    {
      "epoch": 3.934159747397965,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.886,
      "step": 16819
    },
    {
      "epoch": 3.9343936381709743,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 16820
    },
    {
      "epoch": 3.934627528943983,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0296,
      "step": 16821
    },
    {
      "epoch": 3.9348614197169924,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 16822
    },
    {
      "epoch": 3.9350953104900013,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 16823
    },
    {
      "epoch": 3.93532920126301,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 16824
    },
    {
      "epoch": 3.935563092036019,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 16825
    },
    {
      "epoch": 3.9357969828090282,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 16826
    },
    {
      "epoch": 3.936030873582037,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 16827
    },
    {
      "epoch": 3.9362647643550464,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6075,
      "step": 16828
    },
    {
      "epoch": 3.936498655128055,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 16829
    },
    {
      "epoch": 3.936732545901064,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 16830
    },
    {
      "epoch": 3.9369664366740733,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 16831
    },
    {
      "epoch": 3.937200327447082,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8257,
      "step": 16832
    },
    {
      "epoch": 3.9374342182200914,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 16833
    },
    {
      "epoch": 3.9376681089931003,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 16834
    },
    {
      "epoch": 3.937901999766109,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 16835
    },
    {
      "epoch": 3.9381358905391184,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 16836
    },
    {
      "epoch": 3.9383697813121272,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 16837
    },
    {
      "epoch": 3.9386036720851365,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7128,
      "step": 16838
    },
    {
      "epoch": 3.9388375628581453,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6155,
      "step": 16839
    },
    {
      "epoch": 3.939071453631154,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 16840
    },
    {
      "epoch": 3.939305344404163,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 16841
    },
    {
      "epoch": 3.9395392351771723,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 16842
    },
    {
      "epoch": 3.939773125950181,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 16843
    },
    {
      "epoch": 3.9400070167231904,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 16844
    },
    {
      "epoch": 3.9402409074961993,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 16845
    },
    {
      "epoch": 3.940474798269208,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 16846
    },
    {
      "epoch": 3.9407086890422174,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 16847
    },
    {
      "epoch": 3.940942579815226,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 16848
    },
    {
      "epoch": 3.9411764705882355,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4807,
      "step": 16849
    },
    {
      "epoch": 3.9414103613612443,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 16850
    },
    {
      "epoch": 3.941644252134253,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7455,
      "step": 16851
    },
    {
      "epoch": 3.9418781429072625,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 16852
    },
    {
      "epoch": 3.9421120336802713,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 16853
    },
    {
      "epoch": 3.9423459244532806,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 16854
    },
    {
      "epoch": 3.9425798152262894,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5821,
      "step": 16855
    },
    {
      "epoch": 3.9428137059992983,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 16856
    },
    {
      "epoch": 3.943047596772307,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 16857
    },
    {
      "epoch": 3.9432814875453164,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 16858
    },
    {
      "epoch": 3.943515378318325,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 16859
    },
    {
      "epoch": 3.9437492690913345,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 16860
    },
    {
      "epoch": 3.9439831598643433,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 16861
    },
    {
      "epoch": 3.944217050637352,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 16862
    },
    {
      "epoch": 3.9444509414103615,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 16863
    },
    {
      "epoch": 3.9446848321833703,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7401,
      "step": 16864
    },
    {
      "epoch": 3.9449187229563796,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 16865
    },
    {
      "epoch": 3.9451526137293884,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8721,
      "step": 16866
    },
    {
      "epoch": 3.9453865045023973,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 16867
    },
    {
      "epoch": 3.9456203952754065,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 16868
    },
    {
      "epoch": 3.9458542860484154,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 16869
    },
    {
      "epoch": 3.9460881768214247,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 16870
    },
    {
      "epoch": 3.9463220675944335,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 16871
    },
    {
      "epoch": 3.9465559583674423,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 16872
    },
    {
      "epoch": 3.946789849140451,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 16873
    },
    {
      "epoch": 3.9470237399134604,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 16874
    },
    {
      "epoch": 3.9472576306864693,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5934,
      "step": 16875
    },
    {
      "epoch": 3.9474915214594786,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 16876
    },
    {
      "epoch": 3.9477254122324874,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 16877
    },
    {
      "epoch": 3.9479593030054962,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7954,
      "step": 16878
    },
    {
      "epoch": 3.9481931937785055,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 16879
    },
    {
      "epoch": 3.9484270845515144,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6862,
      "step": 16880
    },
    {
      "epoch": 3.9486609753245236,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 16881
    },
    {
      "epoch": 3.9488948660975325,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 16882
    },
    {
      "epoch": 3.9491287568705413,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 16883
    },
    {
      "epoch": 3.9493626476435506,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 16884
    },
    {
      "epoch": 3.9495965384165594,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7154,
      "step": 16885
    },
    {
      "epoch": 3.9498304291895687,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6618,
      "step": 16886
    },
    {
      "epoch": 3.9500643199625776,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 16887
    },
    {
      "epoch": 3.9502982107355864,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 16888
    },
    {
      "epoch": 3.9505321015085952,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 16889
    },
    {
      "epoch": 3.9507659922816045,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 16890
    },
    {
      "epoch": 3.9509998830546134,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 16891
    },
    {
      "epoch": 3.9512337738276226,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 16892
    },
    {
      "epoch": 3.9514676646006315,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 16893
    },
    {
      "epoch": 3.9517015553736403,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6855,
      "step": 16894
    },
    {
      "epoch": 3.9519354461466496,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5824,
      "step": 16895
    },
    {
      "epoch": 3.9521693369196584,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 16896
    },
    {
      "epoch": 3.9524032276926677,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 16897
    },
    {
      "epoch": 3.9526371184656766,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7725,
      "step": 16898
    },
    {
      "epoch": 3.9528710092386854,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 16899
    },
    {
      "epoch": 3.9531049000116947,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 16900
    },
    {
      "epoch": 3.9531049000116947,
      "eval_runtime": 4.6134,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 16900
    },
    {
      "epoch": 3.9533387907847035,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 16901
    },
    {
      "epoch": 3.953572681557713,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 16902
    },
    {
      "epoch": 3.9538065723307216,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6663,
      "step": 16903
    },
    {
      "epoch": 3.9540404631037305,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 16904
    },
    {
      "epoch": 3.9542743538767393,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 16905
    },
    {
      "epoch": 3.9545082446497486,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 16906
    },
    {
      "epoch": 3.9547421354227574,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 16907
    },
    {
      "epoch": 3.9549760261957667,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 16908
    },
    {
      "epoch": 3.9552099169687756,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 16909
    },
    {
      "epoch": 3.9554438077417844,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 16910
    },
    {
      "epoch": 3.9556776985147937,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 16911
    },
    {
      "epoch": 3.9559115892878025,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3441,
      "step": 16912
    },
    {
      "epoch": 3.956145480060812,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 16913
    },
    {
      "epoch": 3.9563793708338206,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 16914
    },
    {
      "epoch": 3.9566132616068295,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 16915
    },
    {
      "epoch": 3.9568471523798388,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 16916
    },
    {
      "epoch": 3.9570810431528476,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 16917
    },
    {
      "epoch": 3.957314933925857,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7453,
      "step": 16918
    },
    {
      "epoch": 3.9575488246988657,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 16919
    },
    {
      "epoch": 3.9577827154718745,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 16920
    },
    {
      "epoch": 3.9580166062448834,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7723,
      "step": 16921
    },
    {
      "epoch": 3.9582504970178927,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6653,
      "step": 16922
    },
    {
      "epoch": 3.9584843877909015,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 16923
    },
    {
      "epoch": 3.958718278563911,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 16924
    },
    {
      "epoch": 3.9589521693369196,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 16925
    },
    {
      "epoch": 3.9591860601099285,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 16926
    },
    {
      "epoch": 3.9594199508829377,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6116,
      "step": 16927
    },
    {
      "epoch": 3.9596538416559466,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 16928
    },
    {
      "epoch": 3.959887732428956,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 16929
    },
    {
      "epoch": 3.9601216232019647,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 16930
    },
    {
      "epoch": 3.9603555139749735,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4702,
      "step": 16931
    },
    {
      "epoch": 3.960589404747983,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 16932
    },
    {
      "epoch": 3.9608232955209917,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 16933
    },
    {
      "epoch": 3.961057186294001,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 16934
    },
    {
      "epoch": 3.96129107706701,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 16935
    },
    {
      "epoch": 3.9615249678400186,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 16936
    },
    {
      "epoch": 3.9617588586130275,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7489,
      "step": 16937
    },
    {
      "epoch": 3.9619927493860367,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 16938
    },
    {
      "epoch": 3.9622266401590456,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 16939
    },
    {
      "epoch": 3.962460530932055,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 16940
    },
    {
      "epoch": 3.9626944217050637,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 16941
    },
    {
      "epoch": 3.9629283124780725,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 16942
    },
    {
      "epoch": 3.963162203251082,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 16943
    },
    {
      "epoch": 3.9633960940240907,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 16944
    },
    {
      "epoch": 3.9636299847971,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 16945
    },
    {
      "epoch": 3.9638638755701088,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 16946
    },
    {
      "epoch": 3.9640977663431176,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 16947
    },
    {
      "epoch": 3.964331657116127,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 16948
    },
    {
      "epoch": 3.9645655478891357,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 16949
    },
    {
      "epoch": 3.964799438662145,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.532,
      "step": 16950
    },
    {
      "epoch": 3.965033329435154,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3673,
      "step": 16951
    },
    {
      "epoch": 3.9652672202081627,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 16952
    },
    {
      "epoch": 3.9655011109811715,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3224,
      "step": 16953
    },
    {
      "epoch": 3.965735001754181,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 16954
    },
    {
      "epoch": 3.9659688925271896,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8178,
      "step": 16955
    },
    {
      "epoch": 3.966202783300199,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 16956
    },
    {
      "epoch": 3.9664366740732078,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7019,
      "step": 16957
    },
    {
      "epoch": 3.9666705648462166,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 16958
    },
    {
      "epoch": 3.966904455619226,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7868,
      "step": 16959
    },
    {
      "epoch": 3.9671383463922347,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2165,
      "step": 16960
    },
    {
      "epoch": 3.967372237165244,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 16961
    },
    {
      "epoch": 3.967606127938253,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 16962
    },
    {
      "epoch": 3.9678400187112617,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 16963
    },
    {
      "epoch": 3.968073909484271,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 16964
    },
    {
      "epoch": 3.96830780025728,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 16965
    },
    {
      "epoch": 3.968541691030289,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 16966
    },
    {
      "epoch": 3.968775581803298,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 16967
    },
    {
      "epoch": 3.9690094725763068,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 16968
    },
    {
      "epoch": 3.9692433633493156,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6352,
      "step": 16969
    },
    {
      "epoch": 3.969477254122325,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 16970
    },
    {
      "epoch": 3.9697111448953337,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5515,
      "step": 16971
    },
    {
      "epoch": 3.969945035668343,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6563,
      "step": 16972
    },
    {
      "epoch": 3.970178926441352,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.572,
      "step": 16973
    },
    {
      "epoch": 3.9704128172143607,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 16974
    },
    {
      "epoch": 3.97064670798737,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 16975
    },
    {
      "epoch": 3.970880598760379,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 16976
    },
    {
      "epoch": 3.971114489533388,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 16977
    },
    {
      "epoch": 3.971348380306397,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 16978
    },
    {
      "epoch": 3.9715822710794058,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 16979
    },
    {
      "epoch": 3.971816161852415,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 16980
    },
    {
      "epoch": 3.972050052625424,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 16981
    },
    {
      "epoch": 3.972283943398433,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 16982
    },
    {
      "epoch": 3.972517834171442,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 16983
    },
    {
      "epoch": 3.972751724944451,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6731,
      "step": 16984
    },
    {
      "epoch": 3.9729856157174597,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 16985
    },
    {
      "epoch": 3.973219506490469,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 16986
    },
    {
      "epoch": 3.973453397263478,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 16987
    },
    {
      "epoch": 3.973687288036487,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 16988
    },
    {
      "epoch": 3.973921178809496,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 16989
    },
    {
      "epoch": 3.9741550695825048,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.6654,
      "step": 16990
    },
    {
      "epoch": 3.974388960355514,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 16991
    },
    {
      "epoch": 3.974622851128523,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 16992
    },
    {
      "epoch": 3.974856741901532,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6608,
      "step": 16993
    },
    {
      "epoch": 3.975090632674541,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 16994
    },
    {
      "epoch": 3.97532452344755,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 16995
    },
    {
      "epoch": 3.975558414220559,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 16996
    },
    {
      "epoch": 3.975792304993568,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 16997
    },
    {
      "epoch": 3.9760261957665772,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.533,
      "step": 16998
    },
    {
      "epoch": 3.976260086539586,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.225,
      "step": 16999
    },
    {
      "epoch": 3.976493977312595,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 17000
    },
    {
      "epoch": 3.976493977312595,
      "eval_runtime": 4.5858,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 17000
    },
    {
      "epoch": 3.9767278680856037,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 17001
    },
    {
      "epoch": 3.976961758858613,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 17002
    },
    {
      "epoch": 3.9771956496316223,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 17003
    },
    {
      "epoch": 3.977429540404631,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 17004
    },
    {
      "epoch": 3.97766343117764,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 17005
    },
    {
      "epoch": 3.977897321950649,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6653,
      "step": 17006
    },
    {
      "epoch": 3.978131212723658,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 17007
    },
    {
      "epoch": 3.978365103496667,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 17008
    },
    {
      "epoch": 3.9785989942696762,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 17009
    },
    {
      "epoch": 3.978832885042685,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 17010
    },
    {
      "epoch": 3.979066775815694,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 17011
    },
    {
      "epoch": 3.979300666588703,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 17012
    },
    {
      "epoch": 3.979534557361712,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 17013
    },
    {
      "epoch": 3.9797684481347213,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7589,
      "step": 17014
    },
    {
      "epoch": 3.98000233890773,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 17015
    },
    {
      "epoch": 3.980236229680739,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 17016
    },
    {
      "epoch": 3.9804701204537483,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 17017
    },
    {
      "epoch": 3.980704011226757,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 17018
    },
    {
      "epoch": 3.9809379019997664,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 17019
    },
    {
      "epoch": 3.981171792772775,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 17020
    },
    {
      "epoch": 3.981405683545784,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 17021
    },
    {
      "epoch": 3.981639574318793,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 17022
    },
    {
      "epoch": 3.981873465091802,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 17023
    },
    {
      "epoch": 3.982107355864811,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7266,
      "step": 17024
    },
    {
      "epoch": 3.9823412466378203,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 17025
    },
    {
      "epoch": 3.982575137410829,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 17026
    },
    {
      "epoch": 3.982809028183838,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 17027
    },
    {
      "epoch": 3.9830429189568473,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6668,
      "step": 17028
    },
    {
      "epoch": 3.983276809729856,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 17029
    },
    {
      "epoch": 3.9835107005028654,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7727,
      "step": 17030
    },
    {
      "epoch": 3.983744591275874,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 17031
    },
    {
      "epoch": 3.983978482048883,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 17032
    },
    {
      "epoch": 3.9842123728218923,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.475,
      "step": 17033
    },
    {
      "epoch": 3.984446263594901,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 17034
    },
    {
      "epoch": 3.9846801543679105,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 17035
    },
    {
      "epoch": 3.9849140451409193,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 17036
    },
    {
      "epoch": 3.985147935913928,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 17037
    },
    {
      "epoch": 3.985381826686937,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 17038
    },
    {
      "epoch": 3.9856157174599463,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 17039
    },
    {
      "epoch": 3.985849608232955,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 17040
    },
    {
      "epoch": 3.9860834990059644,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 17041
    },
    {
      "epoch": 3.986317389778973,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6559,
      "step": 17042
    },
    {
      "epoch": 3.986551280551982,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 17043
    },
    {
      "epoch": 3.9867851713249913,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 17044
    },
    {
      "epoch": 3.987019062098,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6323,
      "step": 17045
    },
    {
      "epoch": 3.9872529528710094,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 17046
    },
    {
      "epoch": 3.9874868436440183,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 17047
    },
    {
      "epoch": 3.987720734417027,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 17048
    },
    {
      "epoch": 3.9879546251900364,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 17049
    },
    {
      "epoch": 3.9881885159630452,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 17050
    },
    {
      "epoch": 3.9884224067360545,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6623,
      "step": 17051
    },
    {
      "epoch": 3.9886562975090634,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 17052
    },
    {
      "epoch": 3.988890188282072,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 17053
    },
    {
      "epoch": 3.989124079055081,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 17054
    },
    {
      "epoch": 3.9893579698280903,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6785,
      "step": 17055
    },
    {
      "epoch": 3.989591860601099,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 17056
    },
    {
      "epoch": 3.9898257513741084,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3713,
      "step": 17057
    },
    {
      "epoch": 3.9900596421471173,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 17058
    },
    {
      "epoch": 3.990293532920126,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 17059
    },
    {
      "epoch": 3.9905274236931354,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 17060
    },
    {
      "epoch": 3.9907613144661442,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 17061
    },
    {
      "epoch": 3.9909952052391535,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 17062
    },
    {
      "epoch": 3.9912290960121624,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8242,
      "step": 17063
    },
    {
      "epoch": 3.991462986785171,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 17064
    },
    {
      "epoch": 3.9916968775581805,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 17065
    },
    {
      "epoch": 3.9919307683311893,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 17066
    },
    {
      "epoch": 3.9921646591041986,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6906,
      "step": 17067
    },
    {
      "epoch": 3.9923985498772074,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 17068
    },
    {
      "epoch": 3.9926324406502163,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7592,
      "step": 17069
    },
    {
      "epoch": 3.992866331423225,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 17070
    },
    {
      "epoch": 3.9931002221962344,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 17071
    },
    {
      "epoch": 3.9933341129692432,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 17072
    },
    {
      "epoch": 3.9935680037422525,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 17073
    },
    {
      "epoch": 3.9938018945152614,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 17074
    },
    {
      "epoch": 3.99403578528827,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 17075
    },
    {
      "epoch": 3.9942696760612795,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 17076
    },
    {
      "epoch": 3.9945035668342883,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 17077
    },
    {
      "epoch": 3.9947374576072976,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 17078
    },
    {
      "epoch": 3.9949713483803064,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8988,
      "step": 17079
    },
    {
      "epoch": 3.9952052391533153,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 17080
    },
    {
      "epoch": 3.9954391299263246,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 17081
    },
    {
      "epoch": 3.9956730206993334,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7381,
      "step": 17082
    },
    {
      "epoch": 3.9959069114723427,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6861,
      "step": 17083
    },
    {
      "epoch": 3.9961408022453515,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 17084
    },
    {
      "epoch": 3.9963746930183603,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 17085
    },
    {
      "epoch": 3.996608583791369,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6632,
      "step": 17086
    },
    {
      "epoch": 3.9968424745643785,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7105,
      "step": 17087
    },
    {
      "epoch": 3.9970763653373873,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 17088
    },
    {
      "epoch": 3.9973102561103966,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 17089
    },
    {
      "epoch": 3.9975441468834054,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 17090
    },
    {
      "epoch": 3.9977780376564143,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 17091
    },
    {
      "epoch": 3.9980119284294235,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 17092
    },
    {
      "epoch": 3.9982458192024324,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 17093
    },
    {
      "epoch": 3.9984797099754417,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 17094
    },
    {
      "epoch": 3.9987136007484505,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 17095
    },
    {
      "epoch": 3.9989474915214593,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 17096
    },
    {
      "epoch": 3.9991813822944686,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 17097
    },
    {
      "epoch": 3.9994152730674775,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 17098
    },
    {
      "epoch": 3.9996491638404867,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 17099
    },
    {
      "epoch": 3.9998830546134956,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6267,
      "step": 17100
    },
    {
      "epoch": 3.9998830546134956,
      "eval_runtime": 4.6376,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 17100
    },
    {
      "epoch": 4.000233890773009,
      "grad_norm": 9.4375,
      "learning_rate": 3e-05,
      "loss": 3.9604,
      "step": 17101
    },
    {
      "epoch": 4.000467781546018,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 17102
    },
    {
      "epoch": 4.000701672319027,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 17103
    },
    {
      "epoch": 4.000935563092036,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6446,
      "step": 17104
    },
    {
      "epoch": 4.001169453865045,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6205,
      "step": 17105
    },
    {
      "epoch": 4.001403344638054,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7295,
      "step": 17106
    },
    {
      "epoch": 4.001637235411063,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5163,
      "step": 17107
    },
    {
      "epoch": 4.0018711261840725,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 17108
    },
    {
      "epoch": 4.002105016957081,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 17109
    },
    {
      "epoch": 4.00233890773009,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6353,
      "step": 17110
    },
    {
      "epoch": 4.002572798503099,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.525,
      "step": 17111
    },
    {
      "epoch": 4.002806689276108,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.525,
      "step": 17112
    },
    {
      "epoch": 4.003040580049117,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6733,
      "step": 17113
    },
    {
      "epoch": 4.003274470822126,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 17114
    },
    {
      "epoch": 4.003508361595135,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 17115
    },
    {
      "epoch": 4.003742252368144,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 17116
    },
    {
      "epoch": 4.003976143141153,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 17117
    },
    {
      "epoch": 4.004210033914162,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4604,
      "step": 17118
    },
    {
      "epoch": 4.0044439246871715,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 17119
    },
    {
      "epoch": 4.00467781546018,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6817,
      "step": 17120
    },
    {
      "epoch": 4.004911706233189,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 17121
    },
    {
      "epoch": 4.005145597006198,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.475,
      "step": 17122
    },
    {
      "epoch": 4.005379487779207,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6794,
      "step": 17123
    },
    {
      "epoch": 4.0056133785522166,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.556,
      "step": 17124
    },
    {
      "epoch": 4.005847269325225,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7349,
      "step": 17125
    },
    {
      "epoch": 4.006081160098234,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 17126
    },
    {
      "epoch": 4.006315050871243,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8305,
      "step": 17127
    },
    {
      "epoch": 4.006548941644252,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 17128
    },
    {
      "epoch": 4.006782832417261,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.177,
      "step": 17129
    },
    {
      "epoch": 4.0070167231902705,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6746,
      "step": 17130
    },
    {
      "epoch": 4.007250613963279,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7345,
      "step": 17131
    },
    {
      "epoch": 4.007484504736288,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5212,
      "step": 17132
    },
    {
      "epoch": 4.007718395509297,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 17133
    },
    {
      "epoch": 4.007952286282306,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 17134
    },
    {
      "epoch": 4.0081861770553155,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 17135
    },
    {
      "epoch": 4.008420067828324,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 17136
    },
    {
      "epoch": 4.008653958601333,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.651,
      "step": 17137
    },
    {
      "epoch": 4.008887849374342,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 17138
    },
    {
      "epoch": 4.009121740147351,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 17139
    },
    {
      "epoch": 4.009355630920361,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 17140
    },
    {
      "epoch": 4.0095895216933695,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 17141
    },
    {
      "epoch": 4.009823412466378,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.72,
      "step": 17142
    },
    {
      "epoch": 4.010057303239387,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9787,
      "step": 17143
    },
    {
      "epoch": 4.010291194012396,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 17144
    },
    {
      "epoch": 4.010525084785405,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 17145
    },
    {
      "epoch": 4.0107589755584145,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9503,
      "step": 17146
    },
    {
      "epoch": 4.010992866331423,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6864,
      "step": 17147
    },
    {
      "epoch": 4.011226757104432,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 17148
    },
    {
      "epoch": 4.011460647877441,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 17149
    },
    {
      "epoch": 4.01169453865045,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 17150
    },
    {
      "epoch": 4.01192842942346,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 17151
    },
    {
      "epoch": 4.0121623201964685,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8027,
      "step": 17152
    },
    {
      "epoch": 4.012396210969477,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 17153
    },
    {
      "epoch": 4.012630101742486,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7461,
      "step": 17154
    },
    {
      "epoch": 4.012863992515495,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 17155
    },
    {
      "epoch": 4.013097883288505,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 17156
    },
    {
      "epoch": 4.0133317740615135,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7573,
      "step": 17157
    },
    {
      "epoch": 4.013565664834522,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 17158
    },
    {
      "epoch": 4.013799555607531,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6434,
      "step": 17159
    },
    {
      "epoch": 4.01403344638054,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8053,
      "step": 17160
    },
    {
      "epoch": 4.014267337153549,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5055,
      "step": 17161
    },
    {
      "epoch": 4.014501227926559,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 17162
    },
    {
      "epoch": 4.0147351186995675,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 17163
    },
    {
      "epoch": 4.014969009472576,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 17164
    },
    {
      "epoch": 4.015202900245585,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8657,
      "step": 17165
    },
    {
      "epoch": 4.015436791018594,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 17166
    },
    {
      "epoch": 4.015670681791604,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7339,
      "step": 17167
    },
    {
      "epoch": 4.0159045725646125,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 17168
    },
    {
      "epoch": 4.016138463337621,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 17169
    },
    {
      "epoch": 4.01637235411063,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 17170
    },
    {
      "epoch": 4.016606244883639,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7191,
      "step": 17171
    },
    {
      "epoch": 4.016840135656649,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 17172
    },
    {
      "epoch": 4.017074026429658,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7244,
      "step": 17173
    },
    {
      "epoch": 4.0173079172026664,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5352,
      "step": 17174
    },
    {
      "epoch": 4.017541807975675,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 17175
    },
    {
      "epoch": 4.017775698748684,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 17176
    },
    {
      "epoch": 4.018009589521693,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 17177
    },
    {
      "epoch": 4.018243480294703,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7771,
      "step": 17178
    },
    {
      "epoch": 4.0184773710677115,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 17179
    },
    {
      "epoch": 4.01871126184072,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4804,
      "step": 17180
    },
    {
      "epoch": 4.018945152613729,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 17181
    },
    {
      "epoch": 4.019179043386738,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 17182
    },
    {
      "epoch": 4.019412934159748,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 17183
    },
    {
      "epoch": 4.019646824932757,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6497,
      "step": 17184
    },
    {
      "epoch": 4.019880715705765,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 17185
    },
    {
      "epoch": 4.020114606478774,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 17186
    },
    {
      "epoch": 4.020348497251783,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5374,
      "step": 17187
    },
    {
      "epoch": 4.020582388024793,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6283,
      "step": 17188
    },
    {
      "epoch": 4.020816278797802,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 17189
    },
    {
      "epoch": 4.0210501695708105,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6732,
      "step": 17190
    },
    {
      "epoch": 4.021284060343819,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 17191
    },
    {
      "epoch": 4.021517951116828,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 17192
    },
    {
      "epoch": 4.021751841889837,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 17193
    },
    {
      "epoch": 4.021985732662847,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.5225,
      "step": 17194
    },
    {
      "epoch": 4.022219623435856,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 17195
    },
    {
      "epoch": 4.022453514208864,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 17196
    },
    {
      "epoch": 4.022687404981873,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 17197
    },
    {
      "epoch": 4.022921295754882,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 17198
    },
    {
      "epoch": 4.023155186527892,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 17199
    },
    {
      "epoch": 4.023389077300901,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5189,
      "step": 17200
    },
    {
      "epoch": 4.023389077300901,
      "eval_runtime": 4.638,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 17200
    },
    {
      "epoch": 4.0236229680739095,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 17201
    },
    {
      "epoch": 4.023856858846918,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 17202
    },
    {
      "epoch": 4.024090749619927,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 17203
    },
    {
      "epoch": 4.024324640392937,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 17204
    },
    {
      "epoch": 4.024558531165946,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 17205
    },
    {
      "epoch": 4.024792421938955,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 17206
    },
    {
      "epoch": 4.025026312711963,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7108,
      "step": 17207
    },
    {
      "epoch": 4.025260203484972,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 17208
    },
    {
      "epoch": 4.025494094257981,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8101,
      "step": 17209
    },
    {
      "epoch": 4.025727985030991,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 17210
    },
    {
      "epoch": 4.025961875804,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 17211
    },
    {
      "epoch": 4.0261957665770085,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 17212
    },
    {
      "epoch": 4.026429657350017,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6743,
      "step": 17213
    },
    {
      "epoch": 4.026663548123026,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 17214
    },
    {
      "epoch": 4.026897438896036,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 17215
    },
    {
      "epoch": 4.027131329669045,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 17216
    },
    {
      "epoch": 4.027365220442054,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6873,
      "step": 17217
    },
    {
      "epoch": 4.027599111215062,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 17218
    },
    {
      "epoch": 4.027833001988071,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 17219
    },
    {
      "epoch": 4.028066892761081,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6322,
      "step": 17220
    },
    {
      "epoch": 4.02830078353409,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 17221
    },
    {
      "epoch": 4.028534674307099,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 17222
    },
    {
      "epoch": 4.0287685650801075,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 17223
    },
    {
      "epoch": 4.029002455853116,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 17224
    },
    {
      "epoch": 4.029236346626125,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 17225
    },
    {
      "epoch": 4.029470237399135,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 17226
    },
    {
      "epoch": 4.029704128172144,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6679,
      "step": 17227
    },
    {
      "epoch": 4.029938018945153,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.758,
      "step": 17228
    },
    {
      "epoch": 4.030171909718161,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 17229
    },
    {
      "epoch": 4.03040580049117,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 17230
    },
    {
      "epoch": 4.03063969126418,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 17231
    },
    {
      "epoch": 4.030873582037189,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 17232
    },
    {
      "epoch": 4.031107472810198,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 17233
    },
    {
      "epoch": 4.0313413635832065,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 17234
    },
    {
      "epoch": 4.031575254356215,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6937,
      "step": 17235
    },
    {
      "epoch": 4.031809145129225,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7166,
      "step": 17236
    },
    {
      "epoch": 4.032043035902234,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 17237
    },
    {
      "epoch": 4.032276926675243,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7666,
      "step": 17238
    },
    {
      "epoch": 4.032510817448252,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 17239
    },
    {
      "epoch": 4.03274470822126,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 17240
    },
    {
      "epoch": 4.032978598994269,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 17241
    },
    {
      "epoch": 4.033212489767279,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 17242
    },
    {
      "epoch": 4.033446380540288,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7191,
      "step": 17243
    },
    {
      "epoch": 4.033680271313297,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.4421,
      "step": 17244
    },
    {
      "epoch": 4.0339141620863055,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.4285,
      "step": 17245
    },
    {
      "epoch": 4.034148052859314,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8151,
      "step": 17246
    },
    {
      "epoch": 4.034381943632324,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 17247
    },
    {
      "epoch": 4.034615834405333,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 17248
    },
    {
      "epoch": 4.034849725178342,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7454,
      "step": 17249
    },
    {
      "epoch": 4.035083615951351,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.5915,
      "step": 17250
    },
    {
      "epoch": 4.035317506724359,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 17251
    },
    {
      "epoch": 4.035551397497369,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 17252
    },
    {
      "epoch": 4.035785288270378,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.6285,
      "step": 17253
    },
    {
      "epoch": 4.036019179043387,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6967,
      "step": 17254
    },
    {
      "epoch": 4.036253069816396,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 17255
    },
    {
      "epoch": 4.0364869605894045,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6442,
      "step": 17256
    },
    {
      "epoch": 4.036720851362413,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 17257
    },
    {
      "epoch": 4.036954742135423,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6526,
      "step": 17258
    },
    {
      "epoch": 4.037188632908432,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 17259
    },
    {
      "epoch": 4.037422523681441,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 17260
    },
    {
      "epoch": 4.03765641445445,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6976,
      "step": 17261
    },
    {
      "epoch": 4.037890305227458,
      "grad_norm": 17.5,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 17262
    },
    {
      "epoch": 4.038124196000468,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.76,
      "step": 17263
    },
    {
      "epoch": 4.038358086773477,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 17264
    },
    {
      "epoch": 4.038591977546486,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4457,
      "step": 17265
    },
    {
      "epoch": 4.038825868319495,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5868,
      "step": 17266
    },
    {
      "epoch": 4.0390597590925035,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 17267
    },
    {
      "epoch": 4.039293649865513,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 17268
    },
    {
      "epoch": 4.039527540638522,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 17269
    },
    {
      "epoch": 4.039761431411531,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 17270
    },
    {
      "epoch": 4.03999532218454,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9506,
      "step": 17271
    },
    {
      "epoch": 4.0402292129575486,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 17272
    },
    {
      "epoch": 4.040463103730557,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 17273
    },
    {
      "epoch": 4.040696994503567,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3217,
      "step": 17274
    },
    {
      "epoch": 4.040930885276576,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 17275
    },
    {
      "epoch": 4.041164776049585,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 17276
    },
    {
      "epoch": 4.041398666822594,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.4403,
      "step": 17277
    },
    {
      "epoch": 4.0416325575956025,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 17278
    },
    {
      "epoch": 4.041866448368612,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 17279
    },
    {
      "epoch": 4.042100339141621,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 17280
    },
    {
      "epoch": 4.04233422991463,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5229,
      "step": 17281
    },
    {
      "epoch": 4.042568120687639,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 17282
    },
    {
      "epoch": 4.0428020114606475,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 17283
    },
    {
      "epoch": 4.043035902233657,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6181,
      "step": 17284
    },
    {
      "epoch": 4.043269793006666,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 17285
    },
    {
      "epoch": 4.043503683779675,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6621,
      "step": 17286
    },
    {
      "epoch": 4.043737574552684,
      "grad_norm": 2.484375,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 17287
    },
    {
      "epoch": 4.043971465325693,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 17288
    },
    {
      "epoch": 4.044205356098702,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 17289
    },
    {
      "epoch": 4.044439246871711,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 17290
    },
    {
      "epoch": 4.04467313764472,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7128,
      "step": 17291
    },
    {
      "epoch": 4.044907028417729,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 17292
    },
    {
      "epoch": 4.045140919190738,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5956,
      "step": 17293
    },
    {
      "epoch": 4.0453748099637465,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 17294
    },
    {
      "epoch": 4.045608700736756,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 17295
    },
    {
      "epoch": 4.045842591509765,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 17296
    },
    {
      "epoch": 4.046076482282774,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8108,
      "step": 17297
    },
    {
      "epoch": 4.046310373055783,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.5889,
      "step": 17298
    },
    {
      "epoch": 4.046544263828792,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 17299
    },
    {
      "epoch": 4.046778154601801,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 17300
    },
    {
      "epoch": 4.046778154601801,
      "eval_runtime": 4.6263,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 17300
    },
    {
      "epoch": 4.04701204537481,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6591,
      "step": 17301
    },
    {
      "epoch": 4.047245936147819,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 17302
    },
    {
      "epoch": 4.047479826920828,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 17303
    },
    {
      "epoch": 4.047713717693837,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 17304
    },
    {
      "epoch": 4.0479476084668455,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 17305
    },
    {
      "epoch": 4.048181499239855,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6726,
      "step": 17306
    },
    {
      "epoch": 4.048415390012864,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 17307
    },
    {
      "epoch": 4.048649280785873,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 17308
    },
    {
      "epoch": 4.048883171558882,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 17309
    },
    {
      "epoch": 4.049117062331891,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 17310
    },
    {
      "epoch": 4.0493509531049,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 17311
    },
    {
      "epoch": 4.049584843877909,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 17312
    },
    {
      "epoch": 4.049818734650918,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5773,
      "step": 17313
    },
    {
      "epoch": 4.050052625423927,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 17314
    },
    {
      "epoch": 4.050286516196936,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 17315
    },
    {
      "epoch": 4.050520406969945,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 17316
    },
    {
      "epoch": 4.050754297742954,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 17317
    },
    {
      "epoch": 4.050988188515963,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 17318
    },
    {
      "epoch": 4.051222079288972,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 17319
    },
    {
      "epoch": 4.051455970061981,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 17320
    },
    {
      "epoch": 4.0516898608349905,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 17321
    },
    {
      "epoch": 4.051923751607999,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.616,
      "step": 17322
    },
    {
      "epoch": 4.052157642381008,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 17323
    },
    {
      "epoch": 4.052391533154017,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6833,
      "step": 17324
    },
    {
      "epoch": 4.052625423927026,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 17325
    },
    {
      "epoch": 4.052859314700035,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 17326
    },
    {
      "epoch": 4.053093205473044,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 17327
    },
    {
      "epoch": 4.053327096246053,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 17328
    },
    {
      "epoch": 4.053560987019062,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 17329
    },
    {
      "epoch": 4.053794877792071,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 17330
    },
    {
      "epoch": 4.05402876856508,
      "grad_norm": 10.1875,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 17331
    },
    {
      "epoch": 4.0542626593380895,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 17332
    },
    {
      "epoch": 4.054496550111098,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 17333
    },
    {
      "epoch": 4.054730440884107,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.5412,
      "step": 17334
    },
    {
      "epoch": 4.054964331657116,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 17335
    },
    {
      "epoch": 4.055198222430125,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7933,
      "step": 17336
    },
    {
      "epoch": 4.055432113203135,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 17337
    },
    {
      "epoch": 4.055666003976143,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 17338
    },
    {
      "epoch": 4.055899894749152,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 17339
    },
    {
      "epoch": 4.056133785522161,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 17340
    },
    {
      "epoch": 4.05636767629517,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 17341
    },
    {
      "epoch": 4.056601567068179,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 17342
    },
    {
      "epoch": 4.0568354578411885,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 17343
    },
    {
      "epoch": 4.057069348614197,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 17344
    },
    {
      "epoch": 4.057303239387206,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.525,
      "step": 17345
    },
    {
      "epoch": 4.057537130160215,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 17346
    },
    {
      "epoch": 4.057771020933224,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 17347
    },
    {
      "epoch": 4.058004911706234,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.4595,
      "step": 17348
    },
    {
      "epoch": 4.058238802479242,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8047,
      "step": 17349
    },
    {
      "epoch": 4.058472693252251,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 17350
    },
    {
      "epoch": 4.05870658402526,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 17351
    },
    {
      "epoch": 4.058940474798269,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 17352
    },
    {
      "epoch": 4.059174365571279,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7604,
      "step": 17353
    },
    {
      "epoch": 4.0594082563442875,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 17354
    },
    {
      "epoch": 4.059642147117296,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7724,
      "step": 17355
    },
    {
      "epoch": 4.059876037890305,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 17356
    },
    {
      "epoch": 4.060109928663314,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6114,
      "step": 17357
    },
    {
      "epoch": 4.060343819436323,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.779,
      "step": 17358
    },
    {
      "epoch": 4.060577710209333,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 17359
    },
    {
      "epoch": 4.060811600982341,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 17360
    },
    {
      "epoch": 4.06104549175535,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 17361
    },
    {
      "epoch": 4.061279382528359,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 17362
    },
    {
      "epoch": 4.061513273301368,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 17363
    },
    {
      "epoch": 4.061747164074378,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.4328,
      "step": 17364
    },
    {
      "epoch": 4.0619810548473865,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.4754,
      "step": 17365
    },
    {
      "epoch": 4.062214945620395,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 17366
    },
    {
      "epoch": 4.062448836393404,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 17367
    },
    {
      "epoch": 4.062682727166413,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 17368
    },
    {
      "epoch": 4.062916617939423,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 17369
    },
    {
      "epoch": 4.0631505087124316,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.447,
      "step": 17370
    },
    {
      "epoch": 4.06338439948544,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 17371
    },
    {
      "epoch": 4.063618290258449,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 17372
    },
    {
      "epoch": 4.063852181031458,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 17373
    },
    {
      "epoch": 4.064086071804467,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 17374
    },
    {
      "epoch": 4.064319962577477,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 17375
    },
    {
      "epoch": 4.0645538533504855,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 17376
    },
    {
      "epoch": 4.064787744123494,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5826,
      "step": 17377
    },
    {
      "epoch": 4.065021634896503,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6544,
      "step": 17378
    },
    {
      "epoch": 4.065255525669512,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 17379
    },
    {
      "epoch": 4.065489416442522,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.667,
      "step": 17380
    },
    {
      "epoch": 4.0657233072155305,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 17381
    },
    {
      "epoch": 4.065957197988539,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 17382
    },
    {
      "epoch": 4.066191088761548,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 17383
    },
    {
      "epoch": 4.066424979534557,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 17384
    },
    {
      "epoch": 4.066658870307567,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 17385
    },
    {
      "epoch": 4.066892761080576,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 17386
    },
    {
      "epoch": 4.0671266518535845,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 17387
    },
    {
      "epoch": 4.067360542626593,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 17388
    },
    {
      "epoch": 4.067594433399602,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8322,
      "step": 17389
    },
    {
      "epoch": 4.067828324172611,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.6535,
      "step": 17390
    },
    {
      "epoch": 4.068062214945621,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 17391
    },
    {
      "epoch": 4.0682961057186295,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 17392
    },
    {
      "epoch": 4.068529996491638,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 17393
    },
    {
      "epoch": 4.068763887264647,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8902,
      "step": 17394
    },
    {
      "epoch": 4.068997778037656,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 17395
    },
    {
      "epoch": 4.069231668810666,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 17396
    },
    {
      "epoch": 4.069465559583675,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 17397
    },
    {
      "epoch": 4.0696994503566835,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6584,
      "step": 17398
    },
    {
      "epoch": 4.069933341129692,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 17399
    },
    {
      "epoch": 4.070167231902701,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 17400
    },
    {
      "epoch": 4.070167231902701,
      "eval_runtime": 4.6069,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 17400
    },
    {
      "epoch": 4.070401122675711,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 17401
    },
    {
      "epoch": 4.07063501344872,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 17402
    },
    {
      "epoch": 4.0708689042217285,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 17403
    },
    {
      "epoch": 4.071102794994737,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 17404
    },
    {
      "epoch": 4.071336685767746,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 17405
    },
    {
      "epoch": 4.071570576540755,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 17406
    },
    {
      "epoch": 4.071804467313765,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7932,
      "step": 17407
    },
    {
      "epoch": 4.072038358086774,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6018,
      "step": 17408
    },
    {
      "epoch": 4.0722722488597825,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 17409
    },
    {
      "epoch": 4.072506139632791,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6053,
      "step": 17410
    },
    {
      "epoch": 4.0727400304058,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 17411
    },
    {
      "epoch": 4.07297392117881,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 17412
    },
    {
      "epoch": 4.073207811951819,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 17413
    },
    {
      "epoch": 4.0734417027248275,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 17414
    },
    {
      "epoch": 4.073675593497836,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 17415
    },
    {
      "epoch": 4.073909484270845,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 17416
    },
    {
      "epoch": 4.074143375043855,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5516,
      "step": 17417
    },
    {
      "epoch": 4.074377265816864,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 17418
    },
    {
      "epoch": 4.074611156589873,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 17419
    },
    {
      "epoch": 4.0748450473628814,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7146,
      "step": 17420
    },
    {
      "epoch": 4.07507893813589,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 17421
    },
    {
      "epoch": 4.075312828908899,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 17422
    },
    {
      "epoch": 4.075546719681909,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 17423
    },
    {
      "epoch": 4.075780610454918,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6489,
      "step": 17424
    },
    {
      "epoch": 4.0760145012279265,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 17425
    },
    {
      "epoch": 4.076248392000935,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 17426
    },
    {
      "epoch": 4.076482282773944,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5647,
      "step": 17427
    },
    {
      "epoch": 4.076716173546954,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 17428
    },
    {
      "epoch": 4.076950064319963,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5689,
      "step": 17429
    },
    {
      "epoch": 4.077183955092972,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5741,
      "step": 17430
    },
    {
      "epoch": 4.07741784586598,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 17431
    },
    {
      "epoch": 4.077651736638989,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 17432
    },
    {
      "epoch": 4.077885627411999,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7737,
      "step": 17433
    },
    {
      "epoch": 4.078119518185008,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 17434
    },
    {
      "epoch": 4.078353408958017,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6413,
      "step": 17435
    },
    {
      "epoch": 4.0785872997310255,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 17436
    },
    {
      "epoch": 4.078821190504034,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 17437
    },
    {
      "epoch": 4.079055081277043,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 17438
    },
    {
      "epoch": 4.079288972050053,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 17439
    },
    {
      "epoch": 4.079522862823062,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5757,
      "step": 17440
    },
    {
      "epoch": 4.079756753596071,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 17441
    },
    {
      "epoch": 4.079990644369079,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 17442
    },
    {
      "epoch": 4.080224535142088,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 17443
    },
    {
      "epoch": 4.080458425915098,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 17444
    },
    {
      "epoch": 4.080692316688107,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 17445
    },
    {
      "epoch": 4.080926207461116,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 17446
    },
    {
      "epoch": 4.0811600982341245,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5756,
      "step": 17447
    },
    {
      "epoch": 4.081393989007133,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 17448
    },
    {
      "epoch": 4.081627879780143,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 17449
    },
    {
      "epoch": 4.081861770553152,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 17450
    },
    {
      "epoch": 4.082095661326161,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.6506,
      "step": 17451
    },
    {
      "epoch": 4.08232955209917,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 17452
    },
    {
      "epoch": 4.082563442872178,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6796,
      "step": 17453
    },
    {
      "epoch": 4.082797333645187,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 17454
    },
    {
      "epoch": 4.083031224418197,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 17455
    },
    {
      "epoch": 4.083265115191206,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6545,
      "step": 17456
    },
    {
      "epoch": 4.083499005964215,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.717,
      "step": 17457
    },
    {
      "epoch": 4.0837328967372235,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 17458
    },
    {
      "epoch": 4.083966787510232,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5789,
      "step": 17459
    },
    {
      "epoch": 4.084200678283242,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 17460
    },
    {
      "epoch": 4.084434569056251,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7277,
      "step": 17461
    },
    {
      "epoch": 4.08466845982926,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.7025,
      "step": 17462
    },
    {
      "epoch": 4.084902350602269,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 17463
    },
    {
      "epoch": 4.085136241375277,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 17464
    },
    {
      "epoch": 4.085370132148287,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 17465
    },
    {
      "epoch": 4.085604022921296,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6651,
      "step": 17466
    },
    {
      "epoch": 4.085837913694305,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 17467
    },
    {
      "epoch": 4.086071804467314,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 17468
    },
    {
      "epoch": 4.0863056952403225,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 17469
    },
    {
      "epoch": 4.086539586013331,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 17470
    },
    {
      "epoch": 4.086773476786341,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.4678,
      "step": 17471
    },
    {
      "epoch": 4.08700736755935,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5449,
      "step": 17472
    },
    {
      "epoch": 4.087241258332359,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5214,
      "step": 17473
    },
    {
      "epoch": 4.087475149105368,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 17474
    },
    {
      "epoch": 4.087709039878376,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 17475
    },
    {
      "epoch": 4.087942930651386,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 17476
    },
    {
      "epoch": 4.088176821424395,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 17477
    },
    {
      "epoch": 4.088410712197404,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 17478
    },
    {
      "epoch": 4.088644602970413,
      "grad_norm": 11.3125,
      "learning_rate": 3e-05,
      "loss": 2.4167,
      "step": 17479
    },
    {
      "epoch": 4.0888784937434215,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.65,
      "step": 17480
    },
    {
      "epoch": 4.089112384516431,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 17481
    },
    {
      "epoch": 4.08934627528944,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8061,
      "step": 17482
    },
    {
      "epoch": 4.089580166062449,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2854,
      "step": 17483
    },
    {
      "epoch": 4.089814056835458,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 17484
    },
    {
      "epoch": 4.090047947608467,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 17485
    },
    {
      "epoch": 4.090281838381475,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 17486
    },
    {
      "epoch": 4.090515729154485,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 17487
    },
    {
      "epoch": 4.090749619927494,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6474,
      "step": 17488
    },
    {
      "epoch": 4.090983510700503,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4014,
      "step": 17489
    },
    {
      "epoch": 4.091217401473512,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 17490
    },
    {
      "epoch": 4.0914512922465205,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7151,
      "step": 17491
    },
    {
      "epoch": 4.09168518301953,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7845,
      "step": 17492
    },
    {
      "epoch": 4.091919073792539,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 17493
    },
    {
      "epoch": 4.092152964565548,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 17494
    },
    {
      "epoch": 4.092386855338557,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 17495
    },
    {
      "epoch": 4.092620746111566,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7398,
      "step": 17496
    },
    {
      "epoch": 4.092854636884575,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 17497
    },
    {
      "epoch": 4.093088527657584,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 17498
    },
    {
      "epoch": 4.093322418430593,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0575,
      "step": 17499
    },
    {
      "epoch": 4.093556309203602,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 17500
    },
    {
      "epoch": 4.093556309203602,
      "eval_runtime": 4.6168,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 17500
    },
    {
      "epoch": 4.093790199976611,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 17501
    },
    {
      "epoch": 4.09402409074962,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 17502
    },
    {
      "epoch": 4.094257981522629,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 17503
    },
    {
      "epoch": 4.094491872295638,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2532,
      "step": 17504
    },
    {
      "epoch": 4.094725763068647,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 17505
    },
    {
      "epoch": 4.094959653841656,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 17506
    },
    {
      "epoch": 4.095193544614665,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6985,
      "step": 17507
    },
    {
      "epoch": 4.095427435387674,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 17508
    },
    {
      "epoch": 4.095661326160683,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 17509
    },
    {
      "epoch": 4.095895216933692,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 17510
    },
    {
      "epoch": 4.096129107706701,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 17511
    },
    {
      "epoch": 4.09636299847971,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 17512
    },
    {
      "epoch": 4.096596889252719,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 17513
    },
    {
      "epoch": 4.096830780025728,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 17514
    },
    {
      "epoch": 4.097064670798737,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8838,
      "step": 17515
    },
    {
      "epoch": 4.097298561571746,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 17516
    },
    {
      "epoch": 4.097532452344755,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6009,
      "step": 17517
    },
    {
      "epoch": 4.0977663431177636,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7932,
      "step": 17518
    },
    {
      "epoch": 4.098000233890773,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 17519
    },
    {
      "epoch": 4.098234124663782,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 17520
    },
    {
      "epoch": 4.098468015436791,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 17521
    },
    {
      "epoch": 4.0987019062098,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 17522
    },
    {
      "epoch": 4.098935796982809,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 17523
    },
    {
      "epoch": 4.099169687755818,
      "grad_norm": 2.546875,
      "learning_rate": 3e-05,
      "loss": 1.4389,
      "step": 17524
    },
    {
      "epoch": 4.099403578528827,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6572,
      "step": 17525
    },
    {
      "epoch": 4.099637469301836,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 17526
    },
    {
      "epoch": 4.099871360074845,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 17527
    },
    {
      "epoch": 4.100105250847854,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 17528
    },
    {
      "epoch": 4.100339141620863,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 17529
    },
    {
      "epoch": 4.100573032393872,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6862,
      "step": 17530
    },
    {
      "epoch": 4.100806923166881,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 17531
    },
    {
      "epoch": 4.10104081393989,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 17532
    },
    {
      "epoch": 4.101274704712899,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 17533
    },
    {
      "epoch": 4.1015085954859085,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 17534
    },
    {
      "epoch": 4.101742486258917,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 17535
    },
    {
      "epoch": 4.101976377031926,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8122,
      "step": 17536
    },
    {
      "epoch": 4.102210267804935,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6847,
      "step": 17537
    },
    {
      "epoch": 4.102444158577944,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 17538
    },
    {
      "epoch": 4.102678049350953,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 17539
    },
    {
      "epoch": 4.102911940123962,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 17540
    },
    {
      "epoch": 4.103145830896971,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 17541
    },
    {
      "epoch": 4.10337972166998,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 17542
    },
    {
      "epoch": 4.103613612442989,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.8124,
      "step": 17543
    },
    {
      "epoch": 4.103847503215998,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 17544
    },
    {
      "epoch": 4.1040813939890075,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 17545
    },
    {
      "epoch": 4.104315284762016,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 17546
    },
    {
      "epoch": 4.104549175535025,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 17547
    },
    {
      "epoch": 4.104783066308034,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 17548
    },
    {
      "epoch": 4.105016957081043,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7201,
      "step": 17549
    },
    {
      "epoch": 4.105250847854053,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 17550
    },
    {
      "epoch": 4.105484738627061,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 17551
    },
    {
      "epoch": 4.10571862940007,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 17552
    },
    {
      "epoch": 4.105952520173079,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 17553
    },
    {
      "epoch": 4.106186410946088,
      "grad_norm": 7.71875,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 17554
    },
    {
      "epoch": 4.106420301719097,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 17555
    },
    {
      "epoch": 4.1066541924921065,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 17556
    },
    {
      "epoch": 4.106888083265115,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 17557
    },
    {
      "epoch": 4.107121974038124,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 17558
    },
    {
      "epoch": 4.107355864811133,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5753,
      "step": 17559
    },
    {
      "epoch": 4.107589755584142,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 17560
    },
    {
      "epoch": 4.107823646357152,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 17561
    },
    {
      "epoch": 4.10805753713016,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 17562
    },
    {
      "epoch": 4.108291427903169,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 17563
    },
    {
      "epoch": 4.108525318676178,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 17564
    },
    {
      "epoch": 4.108759209449187,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 17565
    },
    {
      "epoch": 4.108993100222197,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 17566
    },
    {
      "epoch": 4.1092269909952055,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 17567
    },
    {
      "epoch": 4.109460881768214,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 17568
    },
    {
      "epoch": 4.109694772541223,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6164,
      "step": 17569
    },
    {
      "epoch": 4.109928663314232,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6447,
      "step": 17570
    },
    {
      "epoch": 4.110162554087241,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 17571
    },
    {
      "epoch": 4.110396444860251,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 17572
    },
    {
      "epoch": 4.110630335633259,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8043,
      "step": 17573
    },
    {
      "epoch": 4.110864226406268,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 17574
    },
    {
      "epoch": 4.111098117179277,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 17575
    },
    {
      "epoch": 4.111332007952286,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 17576
    },
    {
      "epoch": 4.111565898725296,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7936,
      "step": 17577
    },
    {
      "epoch": 4.1117997894983045,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7724,
      "step": 17578
    },
    {
      "epoch": 4.112033680271313,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 17579
    },
    {
      "epoch": 4.112267571044322,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 17580
    },
    {
      "epoch": 4.112501461817331,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5905,
      "step": 17581
    },
    {
      "epoch": 4.112735352590341,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 17582
    },
    {
      "epoch": 4.11296924336335,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 17583
    },
    {
      "epoch": 4.113203134136358,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 17584
    },
    {
      "epoch": 4.113437024909367,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 17585
    },
    {
      "epoch": 4.113670915682376,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 17586
    },
    {
      "epoch": 4.113904806455385,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 17587
    },
    {
      "epoch": 4.114138697228395,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0112,
      "step": 17588
    },
    {
      "epoch": 4.1143725880014035,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 17589
    },
    {
      "epoch": 4.114606478774412,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 17590
    },
    {
      "epoch": 4.114840369547421,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 17591
    },
    {
      "epoch": 4.11507426032043,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 17592
    },
    {
      "epoch": 4.11530815109344,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 17593
    },
    {
      "epoch": 4.115542041866449,
      "grad_norm": 8.6875,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 17594
    },
    {
      "epoch": 4.115775932639457,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6454,
      "step": 17595
    },
    {
      "epoch": 4.116009823412466,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 17596
    },
    {
      "epoch": 4.116243714185475,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5978,
      "step": 17597
    },
    {
      "epoch": 4.116477604958485,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7273,
      "step": 17598
    },
    {
      "epoch": 4.116711495731494,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5322,
      "step": 17599
    },
    {
      "epoch": 4.1169453865045025,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 17600
    },
    {
      "epoch": 4.1169453865045025,
      "eval_runtime": 4.6167,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 17600
    },
    {
      "epoch": 4.117179277277511,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.741,
      "step": 17601
    },
    {
      "epoch": 4.11741316805052,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5533,
      "step": 17602
    },
    {
      "epoch": 4.117647058823529,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 17603
    },
    {
      "epoch": 4.117880949596539,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 17604
    },
    {
      "epoch": 4.118114840369548,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7013,
      "step": 17605
    },
    {
      "epoch": 4.118348731142556,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 17606
    },
    {
      "epoch": 4.118582621915565,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 17607
    },
    {
      "epoch": 4.118816512688574,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 17608
    },
    {
      "epoch": 4.119050403461584,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 17609
    },
    {
      "epoch": 4.119284294234593,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 17610
    },
    {
      "epoch": 4.1195181850076015,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 17611
    },
    {
      "epoch": 4.11975207578061,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 17612
    },
    {
      "epoch": 4.119985966553619,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 17613
    },
    {
      "epoch": 4.120219857326629,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 17614
    },
    {
      "epoch": 4.120453748099638,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 17615
    },
    {
      "epoch": 4.1206876388726466,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 17616
    },
    {
      "epoch": 4.120921529645655,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0649,
      "step": 17617
    },
    {
      "epoch": 4.121155420418664,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.557,
      "step": 17618
    },
    {
      "epoch": 4.121389311191673,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 17619
    },
    {
      "epoch": 4.121623201964683,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 17620
    },
    {
      "epoch": 4.121857092737692,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5872,
      "step": 17621
    },
    {
      "epoch": 4.1220909835107005,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 17622
    },
    {
      "epoch": 4.122324874283709,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 17623
    },
    {
      "epoch": 4.122558765056718,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9313,
      "step": 17624
    },
    {
      "epoch": 4.122792655829728,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 17625
    },
    {
      "epoch": 4.123026546602737,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 17626
    },
    {
      "epoch": 4.1232604373757455,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 17627
    },
    {
      "epoch": 4.123494328148754,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5559,
      "step": 17628
    },
    {
      "epoch": 4.123728218921763,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 17629
    },
    {
      "epoch": 4.123962109694773,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 17630
    },
    {
      "epoch": 4.124196000467782,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7415,
      "step": 17631
    },
    {
      "epoch": 4.124429891240791,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 17632
    },
    {
      "epoch": 4.1246637820137995,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 17633
    },
    {
      "epoch": 4.124897672786808,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 17634
    },
    {
      "epoch": 4.125131563559817,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9147,
      "step": 17635
    },
    {
      "epoch": 4.125365454332827,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 17636
    },
    {
      "epoch": 4.125599345105836,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 17637
    },
    {
      "epoch": 4.1258332358788445,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 17638
    },
    {
      "epoch": 4.126067126651853,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.3952,
      "step": 17639
    },
    {
      "epoch": 4.126301017424862,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5842,
      "step": 17640
    },
    {
      "epoch": 4.126534908197872,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 17641
    },
    {
      "epoch": 4.126768798970881,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 17642
    },
    {
      "epoch": 4.12700268974389,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 17643
    },
    {
      "epoch": 4.1272365805168985,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 17644
    },
    {
      "epoch": 4.127470471289907,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 17645
    },
    {
      "epoch": 4.127704362062917,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 17646
    },
    {
      "epoch": 4.127938252835926,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 17647
    },
    {
      "epoch": 4.128172143608935,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 17648
    },
    {
      "epoch": 4.1284060343819435,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 17649
    },
    {
      "epoch": 4.128639925154952,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 17650
    },
    {
      "epoch": 4.128873815927962,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 17651
    },
    {
      "epoch": 4.129107706700971,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 17652
    },
    {
      "epoch": 4.12934159747398,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 17653
    },
    {
      "epoch": 4.129575488246989,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.6145,
      "step": 17654
    },
    {
      "epoch": 4.1298093790199975,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 17655
    },
    {
      "epoch": 4.130043269793006,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 17656
    },
    {
      "epoch": 4.130277160566016,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 17657
    },
    {
      "epoch": 4.130511051339025,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7198,
      "step": 17658
    },
    {
      "epoch": 4.130744942112034,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 17659
    },
    {
      "epoch": 4.1309788328850425,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8415,
      "step": 17660
    },
    {
      "epoch": 4.131212723658051,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 17661
    },
    {
      "epoch": 4.131446614431061,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6117,
      "step": 17662
    },
    {
      "epoch": 4.13168050520407,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 17663
    },
    {
      "epoch": 4.131914395977079,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0727,
      "step": 17664
    },
    {
      "epoch": 4.132148286750088,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 17665
    },
    {
      "epoch": 4.1323821775230964,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 17666
    },
    {
      "epoch": 4.132616068296105,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 17667
    },
    {
      "epoch": 4.132849959069115,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 17668
    },
    {
      "epoch": 4.133083849842124,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5749,
      "step": 17669
    },
    {
      "epoch": 4.133317740615133,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6515,
      "step": 17670
    },
    {
      "epoch": 4.1335516313881415,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 17671
    },
    {
      "epoch": 4.13378552216115,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 17672
    },
    {
      "epoch": 4.13401941293416,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 17673
    },
    {
      "epoch": 4.134253303707169,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7407,
      "step": 17674
    },
    {
      "epoch": 4.134487194480178,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 17675
    },
    {
      "epoch": 4.134721085253187,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 17676
    },
    {
      "epoch": 4.134954976026195,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 17677
    },
    {
      "epoch": 4.135188866799205,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 17678
    },
    {
      "epoch": 4.135422757572214,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 17679
    },
    {
      "epoch": 4.135656648345223,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 17680
    },
    {
      "epoch": 4.135890539118232,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6584,
      "step": 17681
    },
    {
      "epoch": 4.1361244298912405,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 17682
    },
    {
      "epoch": 4.13635832066425,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5488,
      "step": 17683
    },
    {
      "epoch": 4.136592211437259,
      "grad_norm": 9.5,
      "learning_rate": 3e-05,
      "loss": 1.5481,
      "step": 17684
    },
    {
      "epoch": 4.136826102210268,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 17685
    },
    {
      "epoch": 4.137059992983277,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6826,
      "step": 17686
    },
    {
      "epoch": 4.137293883756286,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 17687
    },
    {
      "epoch": 4.137527774529294,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 17688
    },
    {
      "epoch": 4.137761665302304,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 17689
    },
    {
      "epoch": 4.137995556075313,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 17690
    },
    {
      "epoch": 4.138229446848322,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5427,
      "step": 17691
    },
    {
      "epoch": 4.138463337621331,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 17692
    },
    {
      "epoch": 4.1386972283943395,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 17693
    },
    {
      "epoch": 4.138931119167349,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 17694
    },
    {
      "epoch": 4.139165009940358,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 17695
    },
    {
      "epoch": 4.139398900713367,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8027,
      "step": 17696
    },
    {
      "epoch": 4.139632791486376,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 17697
    },
    {
      "epoch": 4.139866682259385,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.2805,
      "step": 17698
    },
    {
      "epoch": 4.140100573032393,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 17699
    },
    {
      "epoch": 4.140334463805403,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 17700
    },
    {
      "epoch": 4.140334463805403,
      "eval_runtime": 4.6895,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 17700
    },
    {
      "epoch": 4.140568354578412,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4939,
      "step": 17701
    },
    {
      "epoch": 4.140802245351421,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0791,
      "step": 17702
    },
    {
      "epoch": 4.14103613612443,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 17703
    },
    {
      "epoch": 4.1412700268974385,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7251,
      "step": 17704
    },
    {
      "epoch": 4.141503917670448,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7292,
      "step": 17705
    },
    {
      "epoch": 4.141737808443457,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 17706
    },
    {
      "epoch": 4.141971699216466,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 17707
    },
    {
      "epoch": 4.142205589989475,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 17708
    },
    {
      "epoch": 4.142439480762484,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 17709
    },
    {
      "epoch": 4.142673371535493,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 17710
    },
    {
      "epoch": 4.142907262308502,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5267,
      "step": 17711
    },
    {
      "epoch": 4.143141153081511,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4434,
      "step": 17712
    },
    {
      "epoch": 4.14337504385452,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 17713
    },
    {
      "epoch": 4.143608934627529,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6343,
      "step": 17714
    },
    {
      "epoch": 4.143842825400538,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 17715
    },
    {
      "epoch": 4.144076716173547,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7008,
      "step": 17716
    },
    {
      "epoch": 4.144310606946556,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 17717
    },
    {
      "epoch": 4.144544497719565,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5076,
      "step": 17718
    },
    {
      "epoch": 4.144778388492574,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 17719
    },
    {
      "epoch": 4.145012279265583,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 17720
    },
    {
      "epoch": 4.145246170038592,
      "grad_norm": 9.0625,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 17721
    },
    {
      "epoch": 4.145480060811601,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 17722
    },
    {
      "epoch": 4.14571395158461,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 17723
    },
    {
      "epoch": 4.145947842357619,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 17724
    },
    {
      "epoch": 4.146181733130628,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 17725
    },
    {
      "epoch": 4.146415623903637,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 17726
    },
    {
      "epoch": 4.146649514676646,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 17727
    },
    {
      "epoch": 4.146883405449655,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 17728
    },
    {
      "epoch": 4.147117296222664,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 17729
    },
    {
      "epoch": 4.147351186995673,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 17730
    },
    {
      "epoch": 4.147585077768682,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 17731
    },
    {
      "epoch": 4.147818968541691,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6538,
      "step": 17732
    },
    {
      "epoch": 4.1480528593147,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7767,
      "step": 17733
    },
    {
      "epoch": 4.148286750087709,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 17734
    },
    {
      "epoch": 4.148520640860718,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6494,
      "step": 17735
    },
    {
      "epoch": 4.148754531633727,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.489,
      "step": 17736
    },
    {
      "epoch": 4.148988422406736,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 17737
    },
    {
      "epoch": 4.149222313179745,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6182,
      "step": 17738
    },
    {
      "epoch": 4.149456203952754,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 17739
    },
    {
      "epoch": 4.149690094725763,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7067,
      "step": 17740
    },
    {
      "epoch": 4.149923985498772,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.585,
      "step": 17741
    },
    {
      "epoch": 4.1501578762717815,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 17742
    },
    {
      "epoch": 4.15039176704479,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 17743
    },
    {
      "epoch": 4.150625657817799,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 17744
    },
    {
      "epoch": 4.150859548590808,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6941,
      "step": 17745
    },
    {
      "epoch": 4.151093439363817,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 17746
    },
    {
      "epoch": 4.1513273301368265,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 17747
    },
    {
      "epoch": 4.151561220909835,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 17748
    },
    {
      "epoch": 4.151795111682844,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 17749
    },
    {
      "epoch": 4.152029002455853,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 17750
    },
    {
      "epoch": 4.152262893228862,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 17751
    },
    {
      "epoch": 4.152496784001871,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 17752
    },
    {
      "epoch": 4.1527306747748804,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 17753
    },
    {
      "epoch": 4.152964565547889,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7156,
      "step": 17754
    },
    {
      "epoch": 4.153198456320898,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 17755
    },
    {
      "epoch": 4.153432347093907,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 17756
    },
    {
      "epoch": 4.153666237866916,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 17757
    },
    {
      "epoch": 4.1539001286399255,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 17758
    },
    {
      "epoch": 4.154134019412934,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 17759
    },
    {
      "epoch": 4.154367910185943,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 17760
    },
    {
      "epoch": 4.154601800958952,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 17761
    },
    {
      "epoch": 4.154835691731961,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6078,
      "step": 17762
    },
    {
      "epoch": 4.155069582504971,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 17763
    },
    {
      "epoch": 4.155303473277979,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6723,
      "step": 17764
    },
    {
      "epoch": 4.155537364050988,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 17765
    },
    {
      "epoch": 4.155771254823997,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 17766
    },
    {
      "epoch": 4.156005145597006,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7527,
      "step": 17767
    },
    {
      "epoch": 4.156239036370015,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5871,
      "step": 17768
    },
    {
      "epoch": 4.1564729271430245,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.565,
      "step": 17769
    },
    {
      "epoch": 4.156706817916033,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 17770
    },
    {
      "epoch": 4.156940708689042,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 17771
    },
    {
      "epoch": 4.157174599462051,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 17772
    },
    {
      "epoch": 4.15740849023506,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 17773
    },
    {
      "epoch": 4.15764238100807,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 17774
    },
    {
      "epoch": 4.157876271781078,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6804,
      "step": 17775
    },
    {
      "epoch": 4.158110162554087,
      "grad_norm": 16.5,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 17776
    },
    {
      "epoch": 4.158344053327096,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6019,
      "step": 17777
    },
    {
      "epoch": 4.158577944100105,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 17778
    },
    {
      "epoch": 4.158811834873115,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 17779
    },
    {
      "epoch": 4.1590457256461235,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 17780
    },
    {
      "epoch": 4.159279616419132,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 17781
    },
    {
      "epoch": 4.159513507192141,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 17782
    },
    {
      "epoch": 4.15974739796515,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 17783
    },
    {
      "epoch": 4.159981288738159,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6646,
      "step": 17784
    },
    {
      "epoch": 4.160215179511169,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 17785
    },
    {
      "epoch": 4.160449070284177,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 17786
    },
    {
      "epoch": 4.160682961057186,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8169,
      "step": 17787
    },
    {
      "epoch": 4.160916851830195,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6594,
      "step": 17788
    },
    {
      "epoch": 4.161150742603204,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 17789
    },
    {
      "epoch": 4.161384633376214,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 17790
    },
    {
      "epoch": 4.1616185241492225,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 17791
    },
    {
      "epoch": 4.161852414922231,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4095,
      "step": 17792
    },
    {
      "epoch": 4.16208630569524,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5628,
      "step": 17793
    },
    {
      "epoch": 4.162320196468249,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.4852,
      "step": 17794
    },
    {
      "epoch": 4.162554087241259,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 17795
    },
    {
      "epoch": 4.162787978014268,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 17796
    },
    {
      "epoch": 4.163021868787276,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 17797
    },
    {
      "epoch": 4.163255759560285,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6,
      "step": 17798
    },
    {
      "epoch": 4.163489650333294,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7207,
      "step": 17799
    },
    {
      "epoch": 4.163723541106303,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.3971,
      "step": 17800
    },
    {
      "epoch": 4.163723541106303,
      "eval_runtime": 4.6408,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 17800
    },
    {
      "epoch": 4.163957431879313,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6909,
      "step": 17801
    },
    {
      "epoch": 4.1641913226523215,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 17802
    },
    {
      "epoch": 4.16442521342533,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 17803
    },
    {
      "epoch": 4.164659104198339,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5729,
      "step": 17804
    },
    {
      "epoch": 4.164892994971348,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 17805
    },
    {
      "epoch": 4.165126885744358,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 17806
    },
    {
      "epoch": 4.165360776517367,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 17807
    },
    {
      "epoch": 4.165594667290375,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5579,
      "step": 17808
    },
    {
      "epoch": 4.165828558063384,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8242,
      "step": 17809
    },
    {
      "epoch": 4.166062448836393,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7782,
      "step": 17810
    },
    {
      "epoch": 4.166296339609403,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7423,
      "step": 17811
    },
    {
      "epoch": 4.166530230382412,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 17812
    },
    {
      "epoch": 4.1667641211554205,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 17813
    },
    {
      "epoch": 4.166998011928429,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 17814
    },
    {
      "epoch": 4.167231902701438,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 17815
    },
    {
      "epoch": 4.167465793474447,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 17816
    },
    {
      "epoch": 4.167699684247457,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 17817
    },
    {
      "epoch": 4.167933575020466,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 17818
    },
    {
      "epoch": 4.168167465793474,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 17819
    },
    {
      "epoch": 4.168401356566483,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 17820
    },
    {
      "epoch": 4.168635247339492,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.3311,
      "step": 17821
    },
    {
      "epoch": 4.168869138112502,
      "grad_norm": 10.4375,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 17822
    },
    {
      "epoch": 4.169103028885511,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 17823
    },
    {
      "epoch": 4.1693369196585195,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5221,
      "step": 17824
    },
    {
      "epoch": 4.169570810431528,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.4624,
      "step": 17825
    },
    {
      "epoch": 4.169804701204537,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 17826
    },
    {
      "epoch": 4.170038591977547,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 17827
    },
    {
      "epoch": 4.170272482750556,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 17828
    },
    {
      "epoch": 4.170506373523565,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.2997,
      "step": 17829
    },
    {
      "epoch": 4.170740264296573,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 17830
    },
    {
      "epoch": 4.170974155069582,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5619,
      "step": 17831
    },
    {
      "epoch": 4.171208045842591,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 17832
    },
    {
      "epoch": 4.171441936615601,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8945,
      "step": 17833
    },
    {
      "epoch": 4.17167582738861,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 17834
    },
    {
      "epoch": 4.1719097181616185,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 17835
    },
    {
      "epoch": 4.172143608934627,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 17836
    },
    {
      "epoch": 4.172377499707636,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5003,
      "step": 17837
    },
    {
      "epoch": 4.172611390480646,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 17838
    },
    {
      "epoch": 4.172845281253655,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 17839
    },
    {
      "epoch": 4.173079172026664,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7112,
      "step": 17840
    },
    {
      "epoch": 4.173313062799672,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 17841
    },
    {
      "epoch": 4.173546953572681,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7782,
      "step": 17842
    },
    {
      "epoch": 4.173780844345691,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 17843
    },
    {
      "epoch": 4.1740147351187,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 17844
    },
    {
      "epoch": 4.174248625891709,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7442,
      "step": 17845
    },
    {
      "epoch": 4.1744825166647175,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.541,
      "step": 17846
    },
    {
      "epoch": 4.174716407437726,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.591,
      "step": 17847
    },
    {
      "epoch": 4.174950298210735,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 17848
    },
    {
      "epoch": 4.175184188983745,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 17849
    },
    {
      "epoch": 4.175418079756754,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2222,
      "step": 17850
    },
    {
      "epoch": 4.175651970529763,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7041,
      "step": 17851
    },
    {
      "epoch": 4.175885861302771,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 17852
    },
    {
      "epoch": 4.17611975207578,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 17853
    },
    {
      "epoch": 4.17635364284879,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.7115,
      "step": 17854
    },
    {
      "epoch": 4.176587533621799,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6399,
      "step": 17855
    },
    {
      "epoch": 4.176821424394808,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 17856
    },
    {
      "epoch": 4.1770553151678165,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6775,
      "step": 17857
    },
    {
      "epoch": 4.177289205940825,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 17858
    },
    {
      "epoch": 4.177523096713835,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 17859
    },
    {
      "epoch": 4.177756987486844,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 17860
    },
    {
      "epoch": 4.177990878259853,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 17861
    },
    {
      "epoch": 4.1782247690328616,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4886,
      "step": 17862
    },
    {
      "epoch": 4.17845865980587,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6919,
      "step": 17863
    },
    {
      "epoch": 4.17869255057888,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9676,
      "step": 17864
    },
    {
      "epoch": 4.178926441351889,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6655,
      "step": 17865
    },
    {
      "epoch": 4.179160332124898,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 17866
    },
    {
      "epoch": 4.179394222897907,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 17867
    },
    {
      "epoch": 4.1796281136709155,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 17868
    },
    {
      "epoch": 4.179862004443924,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 17869
    },
    {
      "epoch": 4.180095895216934,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 17870
    },
    {
      "epoch": 4.180329785989943,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6251,
      "step": 17871
    },
    {
      "epoch": 4.180563676762952,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 17872
    },
    {
      "epoch": 4.1807975675359605,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 17873
    },
    {
      "epoch": 4.181031458308969,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 17874
    },
    {
      "epoch": 4.181265349081979,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 17875
    },
    {
      "epoch": 4.181499239854988,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7339,
      "step": 17876
    },
    {
      "epoch": 4.181733130627997,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 17877
    },
    {
      "epoch": 4.181967021401006,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 17878
    },
    {
      "epoch": 4.1822009121740145,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 17879
    },
    {
      "epoch": 4.182434802947023,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 17880
    },
    {
      "epoch": 4.182668693720033,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 17881
    },
    {
      "epoch": 4.182902584493042,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5725,
      "step": 17882
    },
    {
      "epoch": 4.183136475266051,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 17883
    },
    {
      "epoch": 4.1833703660390595,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 17884
    },
    {
      "epoch": 4.183604256812068,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 17885
    },
    {
      "epoch": 4.183838147585078,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8001,
      "step": 17886
    },
    {
      "epoch": 4.184072038358087,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 17887
    },
    {
      "epoch": 4.184305929131096,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 17888
    },
    {
      "epoch": 4.184539819904105,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6068,
      "step": 17889
    },
    {
      "epoch": 4.1847737106771135,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 17890
    },
    {
      "epoch": 4.185007601450123,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4388,
      "step": 17891
    },
    {
      "epoch": 4.185241492223132,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 17892
    },
    {
      "epoch": 4.185475382996141,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2683,
      "step": 17893
    },
    {
      "epoch": 4.18570927376915,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 17894
    },
    {
      "epoch": 4.1859431645421585,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 17895
    },
    {
      "epoch": 4.186177055315168,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 17896
    },
    {
      "epoch": 4.186410946088177,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7792,
      "step": 17897
    },
    {
      "epoch": 4.186644836861186,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 17898
    },
    {
      "epoch": 4.186878727634195,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7381,
      "step": 17899
    },
    {
      "epoch": 4.187112618407204,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7831,
      "step": 17900
    },
    {
      "epoch": 4.187112618407204,
      "eval_runtime": 4.6157,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 17900
    },
    {
      "epoch": 4.1873465091802125,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 17901
    },
    {
      "epoch": 4.187580399953222,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 17902
    },
    {
      "epoch": 4.187814290726231,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2006,
      "step": 17903
    },
    {
      "epoch": 4.18804818149924,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 17904
    },
    {
      "epoch": 4.188282072272249,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 17905
    },
    {
      "epoch": 4.1885159630452575,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.5163,
      "step": 17906
    },
    {
      "epoch": 4.188749853818267,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.4419,
      "step": 17907
    },
    {
      "epoch": 4.188983744591276,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 17908
    },
    {
      "epoch": 4.189217635364285,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 17909
    },
    {
      "epoch": 4.189451526137294,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 17910
    },
    {
      "epoch": 4.189685416910303,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 17911
    },
    {
      "epoch": 4.1899193076833114,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 17912
    },
    {
      "epoch": 4.190153198456321,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 17913
    },
    {
      "epoch": 4.19038708922933,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 17914
    },
    {
      "epoch": 4.190620980002339,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7321,
      "step": 17915
    },
    {
      "epoch": 4.190854870775348,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 17916
    },
    {
      "epoch": 4.1910887615483565,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 17917
    },
    {
      "epoch": 4.191322652321366,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 17918
    },
    {
      "epoch": 4.191556543094375,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 17919
    },
    {
      "epoch": 4.191790433867384,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6958,
      "step": 17920
    },
    {
      "epoch": 4.192024324640393,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 17921
    },
    {
      "epoch": 4.192258215413402,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 17922
    },
    {
      "epoch": 4.192492106186411,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 17923
    },
    {
      "epoch": 4.19272599695942,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 17924
    },
    {
      "epoch": 4.192959887732429,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 17925
    },
    {
      "epoch": 4.193193778505438,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 17926
    },
    {
      "epoch": 4.193427669278447,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 17927
    },
    {
      "epoch": 4.193661560051456,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 17928
    },
    {
      "epoch": 4.193895450824465,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 17929
    },
    {
      "epoch": 4.194129341597474,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 17930
    },
    {
      "epoch": 4.194363232370483,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 17931
    },
    {
      "epoch": 4.194597123143492,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 17932
    },
    {
      "epoch": 4.194831013916501,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 17933
    },
    {
      "epoch": 4.19506490468951,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.7745,
      "step": 17934
    },
    {
      "epoch": 4.195298795462519,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 17935
    },
    {
      "epoch": 4.195532686235528,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8415,
      "step": 17936
    },
    {
      "epoch": 4.195766577008537,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 17937
    },
    {
      "epoch": 4.196000467781546,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 17938
    },
    {
      "epoch": 4.196234358554555,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 17939
    },
    {
      "epoch": 4.196468249327564,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6846,
      "step": 17940
    },
    {
      "epoch": 4.196702140100573,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8451,
      "step": 17941
    },
    {
      "epoch": 4.196936030873582,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6698,
      "step": 17942
    },
    {
      "epoch": 4.197169921646591,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 17943
    },
    {
      "epoch": 4.1974038124196,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 17944
    },
    {
      "epoch": 4.197637703192609,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 17945
    },
    {
      "epoch": 4.197871593965618,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7267,
      "step": 17946
    },
    {
      "epoch": 4.198105484738627,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5777,
      "step": 17947
    },
    {
      "epoch": 4.198339375511636,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 17948
    },
    {
      "epoch": 4.198573266284645,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 17949
    },
    {
      "epoch": 4.198807157057654,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 17950
    },
    {
      "epoch": 4.199041047830663,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 17951
    },
    {
      "epoch": 4.199274938603672,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 17952
    },
    {
      "epoch": 4.199508829376681,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6925,
      "step": 17953
    },
    {
      "epoch": 4.19974272014969,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 17954
    },
    {
      "epoch": 4.1999766109226995,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.4208,
      "step": 17955
    },
    {
      "epoch": 4.200210501695708,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 17956
    },
    {
      "epoch": 4.200444392468717,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 17957
    },
    {
      "epoch": 4.200678283241726,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 17958
    },
    {
      "epoch": 4.200912174014735,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 17959
    },
    {
      "epoch": 4.2011460647877445,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 17960
    },
    {
      "epoch": 4.201379955560753,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 17961
    },
    {
      "epoch": 4.201613846333762,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 17962
    },
    {
      "epoch": 4.201847737106771,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 17963
    },
    {
      "epoch": 4.20208162787978,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 17964
    },
    {
      "epoch": 4.202315518652789,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5799,
      "step": 17965
    },
    {
      "epoch": 4.2025494094257985,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 17966
    },
    {
      "epoch": 4.202783300198807,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8369,
      "step": 17967
    },
    {
      "epoch": 4.203017190971816,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 17968
    },
    {
      "epoch": 4.203251081744825,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 17969
    },
    {
      "epoch": 4.203484972517834,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.509,
      "step": 17970
    },
    {
      "epoch": 4.2037188632908435,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 17971
    },
    {
      "epoch": 4.203952754063852,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 17972
    },
    {
      "epoch": 4.204186644836861,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 17973
    },
    {
      "epoch": 4.20442053560987,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 17974
    },
    {
      "epoch": 4.204654426382879,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 17975
    },
    {
      "epoch": 4.204888317155889,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 17976
    },
    {
      "epoch": 4.2051222079288975,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 17977
    },
    {
      "epoch": 4.205356098701906,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6124,
      "step": 17978
    },
    {
      "epoch": 4.205589989474915,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4676,
      "step": 17979
    },
    {
      "epoch": 4.205823880247924,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 17980
    },
    {
      "epoch": 4.206057771020933,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 17981
    },
    {
      "epoch": 4.2062916617939425,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 17982
    },
    {
      "epoch": 4.206525552566951,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 17983
    },
    {
      "epoch": 4.20675944333996,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 17984
    },
    {
      "epoch": 4.206993334112969,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 17985
    },
    {
      "epoch": 4.207227224885978,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5457,
      "step": 17986
    },
    {
      "epoch": 4.207461115658988,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 17987
    },
    {
      "epoch": 4.2076950064319965,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7058,
      "step": 17988
    },
    {
      "epoch": 4.207928897205005,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 17989
    },
    {
      "epoch": 4.208162787978014,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 17990
    },
    {
      "epoch": 4.208396678751023,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 17991
    },
    {
      "epoch": 4.208630569524033,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 17992
    },
    {
      "epoch": 4.2088644602970415,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5099,
      "step": 17993
    },
    {
      "epoch": 4.20909835107005,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 17994
    },
    {
      "epoch": 4.209332241843059,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 17995
    },
    {
      "epoch": 4.209566132616068,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 17996
    },
    {
      "epoch": 4.209800023389077,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 17997
    },
    {
      "epoch": 4.210033914162087,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.6493,
      "step": 17998
    },
    {
      "epoch": 4.2102678049350954,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 17999
    },
    {
      "epoch": 4.210501695708104,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 18000
    },
    {
      "epoch": 4.210501695708104,
      "eval_runtime": 4.6412,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 18000
    },
    {
      "epoch": 4.210735586481113,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 18001
    },
    {
      "epoch": 4.210969477254122,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 18002
    },
    {
      "epoch": 4.211203368027132,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 18003
    },
    {
      "epoch": 4.2114372588001405,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5812,
      "step": 18004
    },
    {
      "epoch": 4.211671149573149,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 18005
    },
    {
      "epoch": 4.211905040346158,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 18006
    },
    {
      "epoch": 4.212138931119167,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 18007
    },
    {
      "epoch": 4.212372821892177,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 18008
    },
    {
      "epoch": 4.212606712665186,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7372,
      "step": 18009
    },
    {
      "epoch": 4.212840603438194,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8553,
      "step": 18010
    },
    {
      "epoch": 4.213074494211203,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 18011
    },
    {
      "epoch": 4.213308384984212,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 18012
    },
    {
      "epoch": 4.213542275757221,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 18013
    },
    {
      "epoch": 4.213776166530231,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 18014
    },
    {
      "epoch": 4.2140100573032395,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 18015
    },
    {
      "epoch": 4.214243948076248,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 18016
    },
    {
      "epoch": 4.214477838849257,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 18017
    },
    {
      "epoch": 4.214711729622266,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.528,
      "step": 18018
    },
    {
      "epoch": 4.214945620395276,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 18019
    },
    {
      "epoch": 4.215179511168285,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 18020
    },
    {
      "epoch": 4.215413401941293,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 18021
    },
    {
      "epoch": 4.215647292714302,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 18022
    },
    {
      "epoch": 4.215881183487311,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 18023
    },
    {
      "epoch": 4.216115074260321,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 18024
    },
    {
      "epoch": 4.21634896503333,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 18025
    },
    {
      "epoch": 4.2165828558063385,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 18026
    },
    {
      "epoch": 4.216816746579347,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4196,
      "step": 18027
    },
    {
      "epoch": 4.217050637352356,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 18028
    },
    {
      "epoch": 4.217284528125365,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 18029
    },
    {
      "epoch": 4.217518418898375,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 18030
    },
    {
      "epoch": 4.217752309671384,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 18031
    },
    {
      "epoch": 4.217986200444392,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 18032
    },
    {
      "epoch": 4.218220091217401,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7939,
      "step": 18033
    },
    {
      "epoch": 4.21845398199041,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 18034
    },
    {
      "epoch": 4.21868787276342,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 18035
    },
    {
      "epoch": 4.218921763536429,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 18036
    },
    {
      "epoch": 4.2191556543094375,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.685,
      "step": 18037
    },
    {
      "epoch": 4.219389545082446,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 18038
    },
    {
      "epoch": 4.219623435855455,
      "grad_norm": 14.0625,
      "learning_rate": 3e-05,
      "loss": 2.395,
      "step": 18039
    },
    {
      "epoch": 4.219857326628465,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9103,
      "step": 18040
    },
    {
      "epoch": 4.220091217401474,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 18041
    },
    {
      "epoch": 4.220325108174483,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 18042
    },
    {
      "epoch": 4.220558998947491,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 18043
    },
    {
      "epoch": 4.2207928897205,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 18044
    },
    {
      "epoch": 4.22102678049351,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 18045
    },
    {
      "epoch": 4.221260671266519,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 18046
    },
    {
      "epoch": 4.221494562039528,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.5952,
      "step": 18047
    },
    {
      "epoch": 4.2217284528125365,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 18048
    },
    {
      "epoch": 4.221962343585545,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 18049
    },
    {
      "epoch": 4.222196234358554,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6491,
      "step": 18050
    },
    {
      "epoch": 4.222430125131564,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 18051
    },
    {
      "epoch": 4.222664015904573,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 18052
    },
    {
      "epoch": 4.222897906677582,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7698,
      "step": 18053
    },
    {
      "epoch": 4.22313179745059,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8608,
      "step": 18054
    },
    {
      "epoch": 4.223365688223599,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.0943,
      "step": 18055
    },
    {
      "epoch": 4.223599578996609,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 18056
    },
    {
      "epoch": 4.223833469769618,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 18057
    },
    {
      "epoch": 4.224067360542627,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 18058
    },
    {
      "epoch": 4.2243012513156355,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 18059
    },
    {
      "epoch": 4.224535142088644,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 18060
    },
    {
      "epoch": 4.224769032861653,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 18061
    },
    {
      "epoch": 4.225002923634663,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 18062
    },
    {
      "epoch": 4.225236814407672,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.714,
      "step": 18063
    },
    {
      "epoch": 4.225470705180681,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5216,
      "step": 18064
    },
    {
      "epoch": 4.225704595953689,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 18065
    },
    {
      "epoch": 4.225938486726698,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 18066
    },
    {
      "epoch": 4.226172377499708,
      "grad_norm": 11.75,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 18067
    },
    {
      "epoch": 4.226406268272717,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 18068
    },
    {
      "epoch": 4.226640159045726,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 18069
    },
    {
      "epoch": 4.2268740498187345,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 18070
    },
    {
      "epoch": 4.227107940591743,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 18071
    },
    {
      "epoch": 4.227341831364753,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 18072
    },
    {
      "epoch": 4.227575722137762,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 18073
    },
    {
      "epoch": 4.227809612910771,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 18074
    },
    {
      "epoch": 4.22804350368378,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5986,
      "step": 18075
    },
    {
      "epoch": 4.228277394456788,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 18076
    },
    {
      "epoch": 4.228511285229798,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 18077
    },
    {
      "epoch": 4.228745176002807,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.5973,
      "step": 18078
    },
    {
      "epoch": 4.228979066775816,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 18079
    },
    {
      "epoch": 4.229212957548825,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 18080
    },
    {
      "epoch": 4.2294468483218335,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.2712,
      "step": 18081
    },
    {
      "epoch": 4.229680739094842,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5659,
      "step": 18082
    },
    {
      "epoch": 4.229914629867852,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 18083
    },
    {
      "epoch": 4.230148520640861,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6727,
      "step": 18084
    },
    {
      "epoch": 4.23038241141387,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7252,
      "step": 18085
    },
    {
      "epoch": 4.230616302186879,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 18086
    },
    {
      "epoch": 4.230850192959887,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 18087
    },
    {
      "epoch": 4.231084083732897,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6771,
      "step": 18088
    },
    {
      "epoch": 4.231317974505906,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 18089
    },
    {
      "epoch": 4.231551865278915,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.814,
      "step": 18090
    },
    {
      "epoch": 4.231785756051924,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6948,
      "step": 18091
    },
    {
      "epoch": 4.2320196468249325,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 18092
    },
    {
      "epoch": 4.232253537597941,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7231,
      "step": 18093
    },
    {
      "epoch": 4.232487428370951,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6693,
      "step": 18094
    },
    {
      "epoch": 4.23272131914396,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 18095
    },
    {
      "epoch": 4.232955209916969,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6243,
      "step": 18096
    },
    {
      "epoch": 4.233189100689978,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6943,
      "step": 18097
    },
    {
      "epoch": 4.233422991462986,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 18098
    },
    {
      "epoch": 4.233656882235996,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 18099
    },
    {
      "epoch": 4.233890773009005,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 18100
    },
    {
      "epoch": 4.233890773009005,
      "eval_runtime": 4.6926,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 18100
    },
    {
      "epoch": 4.234124663782014,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5558,
      "step": 18101
    },
    {
      "epoch": 4.234358554555023,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7854,
      "step": 18102
    },
    {
      "epoch": 4.2345924453280315,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 18103
    },
    {
      "epoch": 4.234826336101041,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 18104
    },
    {
      "epoch": 4.23506022687405,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 18105
    },
    {
      "epoch": 4.235294117647059,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 18106
    },
    {
      "epoch": 4.235528008420068,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 18107
    },
    {
      "epoch": 4.2357618991930766,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8929,
      "step": 18108
    },
    {
      "epoch": 4.235995789966086,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 18109
    },
    {
      "epoch": 4.236229680739095,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 18110
    },
    {
      "epoch": 4.236463571512104,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 18111
    },
    {
      "epoch": 4.236697462285113,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6858,
      "step": 18112
    },
    {
      "epoch": 4.236931353058122,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7079,
      "step": 18113
    },
    {
      "epoch": 4.2371652438311305,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7146,
      "step": 18114
    },
    {
      "epoch": 4.23739913460414,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4066,
      "step": 18115
    },
    {
      "epoch": 4.237633025377149,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 18116
    },
    {
      "epoch": 4.237866916150158,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 18117
    },
    {
      "epoch": 4.238100806923167,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3375,
      "step": 18118
    },
    {
      "epoch": 4.2383346976961755,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.747,
      "step": 18119
    },
    {
      "epoch": 4.238568588469185,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 18120
    },
    {
      "epoch": 4.238802479242194,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 18121
    },
    {
      "epoch": 4.239036370015203,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5196,
      "step": 18122
    },
    {
      "epoch": 4.239270260788212,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7454,
      "step": 18123
    },
    {
      "epoch": 4.239504151561221,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 18124
    },
    {
      "epoch": 4.2397380423342295,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5826,
      "step": 18125
    },
    {
      "epoch": 4.239971933107239,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 18126
    },
    {
      "epoch": 4.240205823880248,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 18127
    },
    {
      "epoch": 4.240439714653257,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 18128
    },
    {
      "epoch": 4.240673605426266,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6782,
      "step": 18129
    },
    {
      "epoch": 4.2409074961992745,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 18130
    },
    {
      "epoch": 4.241141386972284,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 18131
    },
    {
      "epoch": 4.241375277745293,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 18132
    },
    {
      "epoch": 4.241609168518302,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 18133
    },
    {
      "epoch": 4.241843059291311,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 18134
    },
    {
      "epoch": 4.24207695006432,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 18135
    },
    {
      "epoch": 4.242310840837329,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 18136
    },
    {
      "epoch": 4.242544731610338,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 18137
    },
    {
      "epoch": 4.242778622383347,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 18138
    },
    {
      "epoch": 4.243012513156356,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 18139
    },
    {
      "epoch": 4.243246403929365,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 18140
    },
    {
      "epoch": 4.243480294702374,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 18141
    },
    {
      "epoch": 4.243714185475383,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 18142
    },
    {
      "epoch": 4.243948076248392,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 18143
    },
    {
      "epoch": 4.244181967021401,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6818,
      "step": 18144
    },
    {
      "epoch": 4.24441585779441,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 18145
    },
    {
      "epoch": 4.244649748567419,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 18146
    },
    {
      "epoch": 4.244883639340428,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 18147
    },
    {
      "epoch": 4.245117530113437,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 18148
    },
    {
      "epoch": 4.245351420886446,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6336,
      "step": 18149
    },
    {
      "epoch": 4.245585311659455,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 18150
    },
    {
      "epoch": 4.245819202432464,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 18151
    },
    {
      "epoch": 4.246053093205473,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 18152
    },
    {
      "epoch": 4.246286983978482,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 18153
    },
    {
      "epoch": 4.246520874751491,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5331,
      "step": 18154
    },
    {
      "epoch": 4.2467547655245,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5184,
      "step": 18155
    },
    {
      "epoch": 4.246988656297509,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6596,
      "step": 18156
    },
    {
      "epoch": 4.2472225470705185,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4742,
      "step": 18157
    },
    {
      "epoch": 4.247456437843527,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2115,
      "step": 18158
    },
    {
      "epoch": 4.247690328616536,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 18159
    },
    {
      "epoch": 4.247924219389545,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 18160
    },
    {
      "epoch": 4.248158110162554,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6229,
      "step": 18161
    },
    {
      "epoch": 4.248392000935563,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.71,
      "step": 18162
    },
    {
      "epoch": 4.248625891708572,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8122,
      "step": 18163
    },
    {
      "epoch": 4.248859782481581,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 18164
    },
    {
      "epoch": 4.24909367325459,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7273,
      "step": 18165
    },
    {
      "epoch": 4.249327564027599,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9075,
      "step": 18166
    },
    {
      "epoch": 4.249561454800608,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 18167
    },
    {
      "epoch": 4.2497953455736175,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 18168
    },
    {
      "epoch": 4.250029236346626,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6064,
      "step": 18169
    },
    {
      "epoch": 4.250263127119635,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5121,
      "step": 18170
    },
    {
      "epoch": 4.250497017892644,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 18171
    },
    {
      "epoch": 4.250730908665653,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 18172
    },
    {
      "epoch": 4.250964799438663,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6874,
      "step": 18173
    },
    {
      "epoch": 4.251198690211671,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 18174
    },
    {
      "epoch": 4.25143258098468,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 18175
    },
    {
      "epoch": 4.251666471757689,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6485,
      "step": 18176
    },
    {
      "epoch": 4.251900362530698,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 18177
    },
    {
      "epoch": 4.252134253303707,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 18178
    },
    {
      "epoch": 4.2523681440767165,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7066,
      "step": 18179
    },
    {
      "epoch": 4.252602034849725,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 18180
    },
    {
      "epoch": 4.252835925622734,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 18181
    },
    {
      "epoch": 4.253069816395743,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 18182
    },
    {
      "epoch": 4.253303707168752,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 18183
    },
    {
      "epoch": 4.253537597941762,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 18184
    },
    {
      "epoch": 4.25377148871477,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 18185
    },
    {
      "epoch": 4.254005379487779,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 18186
    },
    {
      "epoch": 4.254239270260788,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7048,
      "step": 18187
    },
    {
      "epoch": 4.254473161033797,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5776,
      "step": 18188
    },
    {
      "epoch": 4.254707051806806,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8801,
      "step": 18189
    },
    {
      "epoch": 4.2549409425798155,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 18190
    },
    {
      "epoch": 4.255174833352824,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 18191
    },
    {
      "epoch": 4.255408724125833,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6249,
      "step": 18192
    },
    {
      "epoch": 4.255642614898842,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 18193
    },
    {
      "epoch": 4.255876505671852,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 18194
    },
    {
      "epoch": 4.2561103964448606,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 18195
    },
    {
      "epoch": 4.256344287217869,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6742,
      "step": 18196
    },
    {
      "epoch": 4.256578177990878,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6937,
      "step": 18197
    },
    {
      "epoch": 4.256812068763887,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.538,
      "step": 18198
    },
    {
      "epoch": 4.257045959536896,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 18199
    },
    {
      "epoch": 4.257279850309906,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 18200
    },
    {
      "epoch": 4.257279850309906,
      "eval_runtime": 4.6034,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 18200
    },
    {
      "epoch": 4.2575137410829145,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 18201
    },
    {
      "epoch": 4.257747631855923,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4896,
      "step": 18202
    },
    {
      "epoch": 4.257981522628932,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 18203
    },
    {
      "epoch": 4.258215413401941,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6065,
      "step": 18204
    },
    {
      "epoch": 4.258449304174951,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 18205
    },
    {
      "epoch": 4.2586831949479595,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 18206
    },
    {
      "epoch": 4.258917085720968,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 18207
    },
    {
      "epoch": 4.259150976493977,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 18208
    },
    {
      "epoch": 4.259384867266986,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6607,
      "step": 18209
    },
    {
      "epoch": 4.259618758039995,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 18210
    },
    {
      "epoch": 4.259852648813005,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 18211
    },
    {
      "epoch": 4.2600865395860135,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 18212
    },
    {
      "epoch": 4.260320430359022,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 18213
    },
    {
      "epoch": 4.260554321132031,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.487,
      "step": 18214
    },
    {
      "epoch": 4.26078821190504,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 18215
    },
    {
      "epoch": 4.26102210267805,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 18216
    },
    {
      "epoch": 4.2612559934510585,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7007,
      "step": 18217
    },
    {
      "epoch": 4.261489884224067,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 18218
    },
    {
      "epoch": 4.261723774997076,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6065,
      "step": 18219
    },
    {
      "epoch": 4.261957665770085,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 18220
    },
    {
      "epoch": 4.262191556543095,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 18221
    },
    {
      "epoch": 4.262425447316104,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7086,
      "step": 18222
    },
    {
      "epoch": 4.2626593380891125,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 18223
    },
    {
      "epoch": 4.262893228862121,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4493,
      "step": 18224
    },
    {
      "epoch": 4.26312711963513,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8308,
      "step": 18225
    },
    {
      "epoch": 4.26336101040814,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6341,
      "step": 18226
    },
    {
      "epoch": 4.263594901181149,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.4328,
      "step": 18227
    },
    {
      "epoch": 4.2638287919541575,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7033,
      "step": 18228
    },
    {
      "epoch": 4.264062682727166,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 18229
    },
    {
      "epoch": 4.264296573500175,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 18230
    },
    {
      "epoch": 4.264530464273184,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6221,
      "step": 18231
    },
    {
      "epoch": 4.264764355046194,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 18232
    },
    {
      "epoch": 4.264998245819203,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 18233
    },
    {
      "epoch": 4.2652321365922115,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4778,
      "step": 18234
    },
    {
      "epoch": 4.26546602736522,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 18235
    },
    {
      "epoch": 4.265699918138229,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.5454,
      "step": 18236
    },
    {
      "epoch": 4.265933808911239,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 18237
    },
    {
      "epoch": 4.266167699684248,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 18238
    },
    {
      "epoch": 4.2664015904572565,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 18239
    },
    {
      "epoch": 4.266635481230265,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6812,
      "step": 18240
    },
    {
      "epoch": 4.266869372003274,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6908,
      "step": 18241
    },
    {
      "epoch": 4.267103262776283,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 18242
    },
    {
      "epoch": 4.267337153549293,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6178,
      "step": 18243
    },
    {
      "epoch": 4.267571044322302,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 18244
    },
    {
      "epoch": 4.2678049350953104,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 18245
    },
    {
      "epoch": 4.268038825868319,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 18246
    },
    {
      "epoch": 4.268272716641328,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1734,
      "step": 18247
    },
    {
      "epoch": 4.268506607414338,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 18248
    },
    {
      "epoch": 4.268740498187347,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 18249
    },
    {
      "epoch": 4.2689743889603555,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 18250
    },
    {
      "epoch": 4.269208279733364,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6734,
      "step": 18251
    },
    {
      "epoch": 4.269442170506373,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6796,
      "step": 18252
    },
    {
      "epoch": 4.269676061279383,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 18253
    },
    {
      "epoch": 4.269909952052392,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 18254
    },
    {
      "epoch": 4.270143842825401,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 18255
    },
    {
      "epoch": 4.270377733598409,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 18256
    },
    {
      "epoch": 4.270611624371418,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 18257
    },
    {
      "epoch": 4.270845515144428,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 18258
    },
    {
      "epoch": 4.271079405917437,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6868,
      "step": 18259
    },
    {
      "epoch": 4.271313296690446,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 18260
    },
    {
      "epoch": 4.2715471874634545,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 18261
    },
    {
      "epoch": 4.271781078236463,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 18262
    },
    {
      "epoch": 4.272014969009472,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 18263
    },
    {
      "epoch": 4.272248859782482,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 18264
    },
    {
      "epoch": 4.272482750555491,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 18265
    },
    {
      "epoch": 4.2727166413285,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5578,
      "step": 18266
    },
    {
      "epoch": 4.272950532101508,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7572,
      "step": 18267
    },
    {
      "epoch": 4.273184422874517,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 18268
    },
    {
      "epoch": 4.273418313647527,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 18269
    },
    {
      "epoch": 4.273652204420536,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 18270
    },
    {
      "epoch": 4.273886095193545,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9556,
      "step": 18271
    },
    {
      "epoch": 4.2741199859665535,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6352,
      "step": 18272
    },
    {
      "epoch": 4.274353876739562,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 18273
    },
    {
      "epoch": 4.274587767512571,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 18274
    },
    {
      "epoch": 4.274821658285581,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 18275
    },
    {
      "epoch": 4.27505554905859,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.6202,
      "step": 18276
    },
    {
      "epoch": 4.275289439831599,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 18277
    },
    {
      "epoch": 4.275523330604607,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8153,
      "step": 18278
    },
    {
      "epoch": 4.275757221377616,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 18279
    },
    {
      "epoch": 4.275991112150626,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 18280
    },
    {
      "epoch": 4.276225002923635,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 18281
    },
    {
      "epoch": 4.276458893696644,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 18282
    },
    {
      "epoch": 4.2766927844696525,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 18283
    },
    {
      "epoch": 4.276926675242661,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 18284
    },
    {
      "epoch": 4.277160566015671,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 18285
    },
    {
      "epoch": 4.27739445678868,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 18286
    },
    {
      "epoch": 4.277628347561689,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 18287
    },
    {
      "epoch": 4.277862238334698,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 18288
    },
    {
      "epoch": 4.278096129107706,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 18289
    },
    {
      "epoch": 4.278330019880716,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 18290
    },
    {
      "epoch": 4.278563910653725,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 18291
    },
    {
      "epoch": 4.278797801426734,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 18292
    },
    {
      "epoch": 4.279031692199743,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7165,
      "step": 18293
    },
    {
      "epoch": 4.2792655829727515,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 18294
    },
    {
      "epoch": 4.27949947374576,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 18295
    },
    {
      "epoch": 4.27973336451877,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 18296
    },
    {
      "epoch": 4.279967255291779,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6539,
      "step": 18297
    },
    {
      "epoch": 4.280201146064788,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 18298
    },
    {
      "epoch": 4.280435036837797,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5182,
      "step": 18299
    },
    {
      "epoch": 4.280668927610805,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 18300
    },
    {
      "epoch": 4.280668927610805,
      "eval_runtime": 4.5862,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 18300
    },
    {
      "epoch": 4.280902818383815,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 18301
    },
    {
      "epoch": 4.281136709156824,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 18302
    },
    {
      "epoch": 4.281370599929833,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 18303
    },
    {
      "epoch": 4.281604490702842,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 18304
    },
    {
      "epoch": 4.2818383814758505,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 18305
    },
    {
      "epoch": 4.282072272248859,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 18306
    },
    {
      "epoch": 4.282306163021869,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 18307
    },
    {
      "epoch": 4.282540053794878,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.739,
      "step": 18308
    },
    {
      "epoch": 4.282773944567887,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6066,
      "step": 18309
    },
    {
      "epoch": 4.283007835340896,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 18310
    },
    {
      "epoch": 4.283241726113904,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 18311
    },
    {
      "epoch": 4.283475616886914,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 18312
    },
    {
      "epoch": 4.283709507659923,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 18313
    },
    {
      "epoch": 4.283943398432932,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 18314
    },
    {
      "epoch": 4.284177289205941,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8029,
      "step": 18315
    },
    {
      "epoch": 4.2844111799789495,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6031,
      "step": 18316
    },
    {
      "epoch": 4.284645070751959,
      "grad_norm": 2.421875,
      "learning_rate": 3e-05,
      "loss": 1.4532,
      "step": 18317
    },
    {
      "epoch": 4.284878961524968,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 18318
    },
    {
      "epoch": 4.285112852297977,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 18319
    },
    {
      "epoch": 4.285346743070986,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 18320
    },
    {
      "epoch": 4.285580633843995,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7741,
      "step": 18321
    },
    {
      "epoch": 4.285814524617004,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6409,
      "step": 18322
    },
    {
      "epoch": 4.286048415390013,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 18323
    },
    {
      "epoch": 4.286282306163022,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 18324
    },
    {
      "epoch": 4.286516196936031,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1514,
      "step": 18325
    },
    {
      "epoch": 4.28675008770904,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5753,
      "step": 18326
    },
    {
      "epoch": 4.2869839784820485,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 18327
    },
    {
      "epoch": 4.287217869255058,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 18328
    },
    {
      "epoch": 4.287451760028067,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 18329
    },
    {
      "epoch": 4.287685650801076,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7737,
      "step": 18330
    },
    {
      "epoch": 4.287919541574085,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 18331
    },
    {
      "epoch": 4.288153432347094,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 18332
    },
    {
      "epoch": 4.288387323120103,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 18333
    },
    {
      "epoch": 4.288621213893112,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 18334
    },
    {
      "epoch": 4.288855104666121,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.708,
      "step": 18335
    },
    {
      "epoch": 4.28908899543913,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 18336
    },
    {
      "epoch": 4.289322886212139,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5962,
      "step": 18337
    },
    {
      "epoch": 4.2895567769851475,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 18338
    },
    {
      "epoch": 4.289790667758157,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 18339
    },
    {
      "epoch": 4.290024558531166,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5535,
      "step": 18340
    },
    {
      "epoch": 4.290258449304175,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 18341
    },
    {
      "epoch": 4.290492340077184,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 18342
    },
    {
      "epoch": 4.290726230850193,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 18343
    },
    {
      "epoch": 4.290960121623202,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 18344
    },
    {
      "epoch": 4.291194012396211,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 18345
    },
    {
      "epoch": 4.29142790316922,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 18346
    },
    {
      "epoch": 4.291661793942229,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 18347
    },
    {
      "epoch": 4.291895684715238,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 18348
    },
    {
      "epoch": 4.292129575488247,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9303,
      "step": 18349
    },
    {
      "epoch": 4.292363466261256,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 18350
    },
    {
      "epoch": 4.292597357034265,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5285,
      "step": 18351
    },
    {
      "epoch": 4.292831247807274,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 18352
    },
    {
      "epoch": 4.293065138580283,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 18353
    },
    {
      "epoch": 4.293299029353292,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 18354
    },
    {
      "epoch": 4.293532920126301,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 18355
    },
    {
      "epoch": 4.29376681089931,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2724,
      "step": 18356
    },
    {
      "epoch": 4.294000701672319,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 18357
    },
    {
      "epoch": 4.294234592445328,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 18358
    },
    {
      "epoch": 4.294468483218337,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 18359
    },
    {
      "epoch": 4.294702373991346,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.305,
      "step": 18360
    },
    {
      "epoch": 4.294936264764355,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 18361
    },
    {
      "epoch": 4.295170155537364,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 18362
    },
    {
      "epoch": 4.295404046310373,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5369,
      "step": 18363
    },
    {
      "epoch": 4.295637937083382,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 18364
    },
    {
      "epoch": 4.295871827856391,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 18365
    },
    {
      "epoch": 4.2961057186294,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 18366
    },
    {
      "epoch": 4.296339609402409,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 18367
    },
    {
      "epoch": 4.296573500175418,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 18368
    },
    {
      "epoch": 4.296807390948427,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 18369
    },
    {
      "epoch": 4.297041281721436,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 18370
    },
    {
      "epoch": 4.297275172494445,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7531,
      "step": 18371
    },
    {
      "epoch": 4.297509063267454,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 18372
    },
    {
      "epoch": 4.297742954040463,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 18373
    },
    {
      "epoch": 4.297976844813472,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 18374
    },
    {
      "epoch": 4.298210735586481,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 18375
    },
    {
      "epoch": 4.29844462635949,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 18376
    },
    {
      "epoch": 4.298678517132499,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 18377
    },
    {
      "epoch": 4.298912407905508,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8007,
      "step": 18378
    },
    {
      "epoch": 4.299146298678517,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 18379
    },
    {
      "epoch": 4.299380189451526,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 18380
    },
    {
      "epoch": 4.2996140802245355,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 18381
    },
    {
      "epoch": 4.299847970997544,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 18382
    },
    {
      "epoch": 4.300081861770553,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6933,
      "step": 18383
    },
    {
      "epoch": 4.300315752543562,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8201,
      "step": 18384
    },
    {
      "epoch": 4.300549643316571,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 18385
    },
    {
      "epoch": 4.300783534089581,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 18386
    },
    {
      "epoch": 4.301017424862589,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 18387
    },
    {
      "epoch": 4.301251315635598,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8846,
      "step": 18388
    },
    {
      "epoch": 4.301485206408607,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 18389
    },
    {
      "epoch": 4.301719097181616,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 18390
    },
    {
      "epoch": 4.301952987954625,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 18391
    },
    {
      "epoch": 4.3021868787276345,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 18392
    },
    {
      "epoch": 4.302420769500643,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 18393
    },
    {
      "epoch": 4.302654660273652,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.7165,
      "step": 18394
    },
    {
      "epoch": 4.302888551046661,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 18395
    },
    {
      "epoch": 4.30312244181967,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3545,
      "step": 18396
    },
    {
      "epoch": 4.30335633259268,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 18397
    },
    {
      "epoch": 4.303590223365688,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5208,
      "step": 18398
    },
    {
      "epoch": 4.303824114138697,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7214,
      "step": 18399
    },
    {
      "epoch": 4.304058004911706,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1923,
      "step": 18400
    },
    {
      "epoch": 4.304058004911706,
      "eval_runtime": 4.6055,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 18400
    },
    {
      "epoch": 4.304291895684715,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 18401
    },
    {
      "epoch": 4.304525786457725,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 18402
    },
    {
      "epoch": 4.3047596772307335,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.3932,
      "step": 18403
    },
    {
      "epoch": 4.304993568003742,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 18404
    },
    {
      "epoch": 4.305227458776751,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4437,
      "step": 18405
    },
    {
      "epoch": 4.30546134954976,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 18406
    },
    {
      "epoch": 4.30569524032277,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 18407
    },
    {
      "epoch": 4.305929131095779,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 18408
    },
    {
      "epoch": 4.306163021868787,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7987,
      "step": 18409
    },
    {
      "epoch": 4.306396912641796,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 18410
    },
    {
      "epoch": 4.306630803414805,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 18411
    },
    {
      "epoch": 4.306864694187814,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 18412
    },
    {
      "epoch": 4.307098584960824,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.5728,
      "step": 18413
    },
    {
      "epoch": 4.3073324757338325,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4588,
      "step": 18414
    },
    {
      "epoch": 4.307566366506841,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 18415
    },
    {
      "epoch": 4.30780025727985,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7298,
      "step": 18416
    },
    {
      "epoch": 4.308034148052859,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 18417
    },
    {
      "epoch": 4.308268038825869,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7624,
      "step": 18418
    },
    {
      "epoch": 4.308501929598878,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 18419
    },
    {
      "epoch": 4.308735820371886,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 18420
    },
    {
      "epoch": 4.308969711144895,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 18421
    },
    {
      "epoch": 4.309203601917904,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 18422
    },
    {
      "epoch": 4.309437492690913,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 18423
    },
    {
      "epoch": 4.309671383463923,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6105,
      "step": 18424
    },
    {
      "epoch": 4.3099052742369315,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 18425
    },
    {
      "epoch": 4.31013916500994,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6507,
      "step": 18426
    },
    {
      "epoch": 4.310373055782949,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 18427
    },
    {
      "epoch": 4.310606946555958,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 18428
    },
    {
      "epoch": 4.310840837328968,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 18429
    },
    {
      "epoch": 4.311074728101977,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 18430
    },
    {
      "epoch": 4.311308618874985,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.718,
      "step": 18431
    },
    {
      "epoch": 4.311542509647994,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 18432
    },
    {
      "epoch": 4.311776400421003,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 18433
    },
    {
      "epoch": 4.312010291194013,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 18434
    },
    {
      "epoch": 4.312244181967022,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 18435
    },
    {
      "epoch": 4.3124780727400305,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 18436
    },
    {
      "epoch": 4.312711963513039,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 18437
    },
    {
      "epoch": 4.312945854286048,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8711,
      "step": 18438
    },
    {
      "epoch": 4.313179745059058,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 18439
    },
    {
      "epoch": 4.313413635832067,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 18440
    },
    {
      "epoch": 4.3136475266050756,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 18441
    },
    {
      "epoch": 4.313881417378084,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 18442
    },
    {
      "epoch": 4.314115308151093,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6606,
      "step": 18443
    },
    {
      "epoch": 4.314349198924102,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 18444
    },
    {
      "epoch": 4.314583089697112,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 18445
    },
    {
      "epoch": 4.314816980470121,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7862,
      "step": 18446
    },
    {
      "epoch": 4.3150508712431295,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 18447
    },
    {
      "epoch": 4.315284762016138,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 18448
    },
    {
      "epoch": 4.315518652789147,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 18449
    },
    {
      "epoch": 4.315752543562157,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 18450
    },
    {
      "epoch": 4.315986434335166,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 18451
    },
    {
      "epoch": 4.3162203251081745,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 18452
    },
    {
      "epoch": 4.316454215881183,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7608,
      "step": 18453
    },
    {
      "epoch": 4.316688106654192,
      "grad_norm": 14.0,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 18454
    },
    {
      "epoch": 4.316921997427201,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7147,
      "step": 18455
    },
    {
      "epoch": 4.317155888200211,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 18456
    },
    {
      "epoch": 4.31738977897322,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7699,
      "step": 18457
    },
    {
      "epoch": 4.3176236697462285,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 18458
    },
    {
      "epoch": 4.317857560519237,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.5097,
      "step": 18459
    },
    {
      "epoch": 4.318091451292246,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 18460
    },
    {
      "epoch": 4.318325342065256,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.236,
      "step": 18461
    },
    {
      "epoch": 4.318559232838265,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 18462
    },
    {
      "epoch": 4.3187931236112735,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 18463
    },
    {
      "epoch": 4.319027014384282,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 18464
    },
    {
      "epoch": 4.319260905157291,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8532,
      "step": 18465
    },
    {
      "epoch": 4.319494795930301,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 18466
    },
    {
      "epoch": 4.31972868670331,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8132,
      "step": 18467
    },
    {
      "epoch": 4.319962577476319,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 18468
    },
    {
      "epoch": 4.3201964682493275,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 18469
    },
    {
      "epoch": 4.320430359022336,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 18470
    },
    {
      "epoch": 4.320664249795346,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 18471
    },
    {
      "epoch": 4.320898140568355,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 18472
    },
    {
      "epoch": 4.321132031341364,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 18473
    },
    {
      "epoch": 4.3213659221143725,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 18474
    },
    {
      "epoch": 4.321599812887381,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 18475
    },
    {
      "epoch": 4.32183370366039,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 18476
    },
    {
      "epoch": 4.3220675944334,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 18477
    },
    {
      "epoch": 4.322301485206409,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 18478
    },
    {
      "epoch": 4.322535375979418,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.4898,
      "step": 18479
    },
    {
      "epoch": 4.3227692667524265,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2967,
      "step": 18480
    },
    {
      "epoch": 4.323003157525435,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5589,
      "step": 18481
    },
    {
      "epoch": 4.323237048298445,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 18482
    },
    {
      "epoch": 4.323470939071454,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 18483
    },
    {
      "epoch": 4.323704829844463,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 18484
    },
    {
      "epoch": 4.3239387206174715,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 18485
    },
    {
      "epoch": 4.32417261139048,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 18486
    },
    {
      "epoch": 4.324406502163489,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 18487
    },
    {
      "epoch": 4.324640392936499,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 18488
    },
    {
      "epoch": 4.324874283709508,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 18489
    },
    {
      "epoch": 4.325108174482517,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 18490
    },
    {
      "epoch": 4.3253420652555254,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6935,
      "step": 18491
    },
    {
      "epoch": 4.325575956028534,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 18492
    },
    {
      "epoch": 4.325809846801544,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 18493
    },
    {
      "epoch": 4.326043737574553,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 18494
    },
    {
      "epoch": 4.326277628347562,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 18495
    },
    {
      "epoch": 4.3265115191205705,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 18496
    },
    {
      "epoch": 4.326745409893579,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6311,
      "step": 18497
    },
    {
      "epoch": 4.326979300666589,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 18498
    },
    {
      "epoch": 4.327213191439598,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 18499
    },
    {
      "epoch": 4.327447082212607,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5272,
      "step": 18500
    },
    {
      "epoch": 4.327447082212607,
      "eval_runtime": 4.6278,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 18500
    },
    {
      "epoch": 4.327680972985616,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 18501
    },
    {
      "epoch": 4.327914863758624,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 18502
    },
    {
      "epoch": 4.328148754531634,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6492,
      "step": 18503
    },
    {
      "epoch": 4.328382645304643,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 18504
    },
    {
      "epoch": 4.328616536077652,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 18505
    },
    {
      "epoch": 4.328850426850661,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 18506
    },
    {
      "epoch": 4.3290843176236695,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7474,
      "step": 18507
    },
    {
      "epoch": 4.329318208396678,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 18508
    },
    {
      "epoch": 4.329552099169688,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 18509
    },
    {
      "epoch": 4.329785989942697,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 18510
    },
    {
      "epoch": 4.330019880715706,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 18511
    },
    {
      "epoch": 4.330253771488715,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.6251,
      "step": 18512
    },
    {
      "epoch": 4.330487662261723,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 18513
    },
    {
      "epoch": 4.330721553034733,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6203,
      "step": 18514
    },
    {
      "epoch": 4.330955443807742,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 18515
    },
    {
      "epoch": 4.331189334580751,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 18516
    },
    {
      "epoch": 4.33142322535376,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 18517
    },
    {
      "epoch": 4.3316571161267685,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 18518
    },
    {
      "epoch": 4.331891006899777,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 18519
    },
    {
      "epoch": 4.332124897672787,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 18520
    },
    {
      "epoch": 4.332358788445796,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8347,
      "step": 18521
    },
    {
      "epoch": 4.332592679218805,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 18522
    },
    {
      "epoch": 4.332826569991814,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 18523
    },
    {
      "epoch": 4.333060460764822,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6623,
      "step": 18524
    },
    {
      "epoch": 4.333294351537832,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 18525
    },
    {
      "epoch": 4.333528242310841,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 18526
    },
    {
      "epoch": 4.33376213308385,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 18527
    },
    {
      "epoch": 4.333996023856859,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 18528
    },
    {
      "epoch": 4.3342299146298675,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8234,
      "step": 18529
    },
    {
      "epoch": 4.334463805402877,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5144,
      "step": 18530
    },
    {
      "epoch": 4.334697696175886,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 18531
    },
    {
      "epoch": 4.334931586948895,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 18532
    },
    {
      "epoch": 4.335165477721904,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 18533
    },
    {
      "epoch": 4.335399368494913,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 18534
    },
    {
      "epoch": 4.335633259267922,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.3855,
      "step": 18535
    },
    {
      "epoch": 4.335867150040931,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 18536
    },
    {
      "epoch": 4.33610104081394,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 18537
    },
    {
      "epoch": 4.336334931586949,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 18538
    },
    {
      "epoch": 4.336568822359958,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 18539
    },
    {
      "epoch": 4.3368027131329665,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 18540
    },
    {
      "epoch": 4.337036603905976,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 18541
    },
    {
      "epoch": 4.337270494678985,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 18542
    },
    {
      "epoch": 4.337504385451994,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 18543
    },
    {
      "epoch": 4.337738276225003,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 18544
    },
    {
      "epoch": 4.337972166998012,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 18545
    },
    {
      "epoch": 4.338206057771021,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8535,
      "step": 18546
    },
    {
      "epoch": 4.33843994854403,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 18547
    },
    {
      "epoch": 4.338673839317039,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 18548
    },
    {
      "epoch": 4.338907730090048,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 18549
    },
    {
      "epoch": 4.339141620863057,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 18550
    },
    {
      "epoch": 4.3393755116360655,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2131,
      "step": 18551
    },
    {
      "epoch": 4.339609402409075,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5599,
      "step": 18552
    },
    {
      "epoch": 4.339843293182084,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3037,
      "step": 18553
    },
    {
      "epoch": 4.340077183955093,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 18554
    },
    {
      "epoch": 4.340311074728102,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 18555
    },
    {
      "epoch": 4.340544965501111,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 18556
    },
    {
      "epoch": 4.34077885627412,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 18557
    },
    {
      "epoch": 4.341012747047129,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6724,
      "step": 18558
    },
    {
      "epoch": 4.341246637820138,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 18559
    },
    {
      "epoch": 4.341480528593147,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2443,
      "step": 18560
    },
    {
      "epoch": 4.341714419366156,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 18561
    },
    {
      "epoch": 4.341948310139165,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 18562
    },
    {
      "epoch": 4.342182200912174,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5204,
      "step": 18563
    },
    {
      "epoch": 4.342416091685183,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9322,
      "step": 18564
    },
    {
      "epoch": 4.342649982458192,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 18565
    },
    {
      "epoch": 4.342883873231201,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5193,
      "step": 18566
    },
    {
      "epoch": 4.3431177640042105,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 18567
    },
    {
      "epoch": 4.343351654777219,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 18568
    },
    {
      "epoch": 4.343585545550228,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 18569
    },
    {
      "epoch": 4.343819436323237,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 18570
    },
    {
      "epoch": 4.344053327096246,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 18571
    },
    {
      "epoch": 4.344287217869255,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 18572
    },
    {
      "epoch": 4.344521108642264,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 18573
    },
    {
      "epoch": 4.344754999415273,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6897,
      "step": 18574
    },
    {
      "epoch": 4.344988890188282,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 18575
    },
    {
      "epoch": 4.345222780961291,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6139,
      "step": 18576
    },
    {
      "epoch": 4.3454566717343,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5652,
      "step": 18577
    },
    {
      "epoch": 4.3456905625073095,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 18578
    },
    {
      "epoch": 4.345924453280318,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 18579
    },
    {
      "epoch": 4.346158344053327,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6094,
      "step": 18580
    },
    {
      "epoch": 4.346392234826336,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7651,
      "step": 18581
    },
    {
      "epoch": 4.346626125599345,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 18582
    },
    {
      "epoch": 4.346860016372354,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 18583
    },
    {
      "epoch": 4.347093907145363,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 18584
    },
    {
      "epoch": 4.347327797918372,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7416,
      "step": 18585
    },
    {
      "epoch": 4.347561688691381,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 18586
    },
    {
      "epoch": 4.34779557946439,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6099,
      "step": 18587
    },
    {
      "epoch": 4.348029470237399,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 18588
    },
    {
      "epoch": 4.3482633610104084,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 18589
    },
    {
      "epoch": 4.348497251783417,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 18590
    },
    {
      "epoch": 4.348731142556426,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7851,
      "step": 18591
    },
    {
      "epoch": 4.348965033329435,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2941,
      "step": 18592
    },
    {
      "epoch": 4.349198924102444,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 18593
    },
    {
      "epoch": 4.3494328148754535,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 18594
    },
    {
      "epoch": 4.349666705648462,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 18595
    },
    {
      "epoch": 4.349900596421471,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 18596
    },
    {
      "epoch": 4.35013448719448,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 18597
    },
    {
      "epoch": 4.350368377967489,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 18598
    },
    {
      "epoch": 4.350602268740499,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 18599
    },
    {
      "epoch": 4.350836159513507,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.747,
      "step": 18600
    },
    {
      "epoch": 4.350836159513507,
      "eval_runtime": 4.6104,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 18600
    },
    {
      "epoch": 4.351070050286516,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 18601
    },
    {
      "epoch": 4.351303941059525,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 18602
    },
    {
      "epoch": 4.351537831832534,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7854,
      "step": 18603
    },
    {
      "epoch": 4.351771722605543,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 18604
    },
    {
      "epoch": 4.3520056133785525,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 18605
    },
    {
      "epoch": 4.352239504151561,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 18606
    },
    {
      "epoch": 4.35247339492457,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 18607
    },
    {
      "epoch": 4.352707285697579,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 18608
    },
    {
      "epoch": 4.352941176470588,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 18609
    },
    {
      "epoch": 4.353175067243598,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.727,
      "step": 18610
    },
    {
      "epoch": 4.353408958016606,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 18611
    },
    {
      "epoch": 4.353642848789615,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 18612
    },
    {
      "epoch": 4.353876739562624,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 18613
    },
    {
      "epoch": 4.354110630335633,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 18614
    },
    {
      "epoch": 4.354344521108643,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7604,
      "step": 18615
    },
    {
      "epoch": 4.3545784118816515,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 18616
    },
    {
      "epoch": 4.35481230265466,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 18617
    },
    {
      "epoch": 4.355046193427669,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6244,
      "step": 18618
    },
    {
      "epoch": 4.355280084200678,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 18619
    },
    {
      "epoch": 4.355513974973688,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 18620
    },
    {
      "epoch": 4.355747865746697,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6055,
      "step": 18621
    },
    {
      "epoch": 4.355981756519705,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 18622
    },
    {
      "epoch": 4.356215647292714,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 18623
    },
    {
      "epoch": 4.356449538065723,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 18624
    },
    {
      "epoch": 4.356683428838732,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 18625
    },
    {
      "epoch": 4.356917319611742,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.3283,
      "step": 18626
    },
    {
      "epoch": 4.3571512103847505,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6914,
      "step": 18627
    },
    {
      "epoch": 4.357385101157759,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 18628
    },
    {
      "epoch": 4.357618991930768,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 18629
    },
    {
      "epoch": 4.357852882703777,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 18630
    },
    {
      "epoch": 4.358086773476787,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 18631
    },
    {
      "epoch": 4.358320664249796,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.3747,
      "step": 18632
    },
    {
      "epoch": 4.358554555022804,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 18633
    },
    {
      "epoch": 4.358788445795813,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 18634
    },
    {
      "epoch": 4.359022336568822,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 18635
    },
    {
      "epoch": 4.359256227341831,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 18636
    },
    {
      "epoch": 4.359490118114841,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 18637
    },
    {
      "epoch": 4.3597240088878495,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5889,
      "step": 18638
    },
    {
      "epoch": 4.359957899660858,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 18639
    },
    {
      "epoch": 4.360191790433867,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 18640
    },
    {
      "epoch": 4.360425681206876,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 18641
    },
    {
      "epoch": 4.360659571979886,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 18642
    },
    {
      "epoch": 4.360893462752895,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5715,
      "step": 18643
    },
    {
      "epoch": 4.361127353525903,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 18644
    },
    {
      "epoch": 4.361361244298912,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 18645
    },
    {
      "epoch": 4.361595135071921,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 18646
    },
    {
      "epoch": 4.361829025844931,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.3442,
      "step": 18647
    },
    {
      "epoch": 4.36206291661794,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 18648
    },
    {
      "epoch": 4.3622968073909485,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 18649
    },
    {
      "epoch": 4.362530698163957,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 18650
    },
    {
      "epoch": 4.362764588936966,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 18651
    },
    {
      "epoch": 4.362998479709976,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 18652
    },
    {
      "epoch": 4.363232370482985,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 18653
    },
    {
      "epoch": 4.363466261255994,
      "grad_norm": 9.6875,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 18654
    },
    {
      "epoch": 4.363700152029002,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 18655
    },
    {
      "epoch": 4.363934042802011,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 18656
    },
    {
      "epoch": 4.36416793357502,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 18657
    },
    {
      "epoch": 4.36440182434803,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7802,
      "step": 18658
    },
    {
      "epoch": 4.364635715121039,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6543,
      "step": 18659
    },
    {
      "epoch": 4.3648696058940475,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8524,
      "step": 18660
    },
    {
      "epoch": 4.365103496667056,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 18661
    },
    {
      "epoch": 4.365337387440065,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 18662
    },
    {
      "epoch": 4.365571278213075,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 18663
    },
    {
      "epoch": 4.365805168986084,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 18664
    },
    {
      "epoch": 4.366039059759093,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 18665
    },
    {
      "epoch": 4.366272950532101,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 18666
    },
    {
      "epoch": 4.36650684130511,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.533,
      "step": 18667
    },
    {
      "epoch": 4.366740732078119,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 18668
    },
    {
      "epoch": 4.366974622851129,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 18669
    },
    {
      "epoch": 4.367208513624138,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5882,
      "step": 18670
    },
    {
      "epoch": 4.3674424043971465,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 18671
    },
    {
      "epoch": 4.367676295170155,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 18672
    },
    {
      "epoch": 4.367910185943164,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 18673
    },
    {
      "epoch": 4.368144076716174,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 18674
    },
    {
      "epoch": 4.368377967489183,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 18675
    },
    {
      "epoch": 4.368611858262192,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7103,
      "step": 18676
    },
    {
      "epoch": 4.3688457490352,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.3543,
      "step": 18677
    },
    {
      "epoch": 4.369079639808209,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 18678
    },
    {
      "epoch": 4.369313530581219,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6956,
      "step": 18679
    },
    {
      "epoch": 4.369547421354228,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 18680
    },
    {
      "epoch": 4.369781312127237,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 18681
    },
    {
      "epoch": 4.3700152029002455,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 18682
    },
    {
      "epoch": 4.370249093673254,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.3735,
      "step": 18683
    },
    {
      "epoch": 4.370482984446264,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 18684
    },
    {
      "epoch": 4.370716875219273,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 18685
    },
    {
      "epoch": 4.370950765992282,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7186,
      "step": 18686
    },
    {
      "epoch": 4.3711846567652906,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 18687
    },
    {
      "epoch": 4.371418547538299,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 18688
    },
    {
      "epoch": 4.371652438311308,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 18689
    },
    {
      "epoch": 4.371886329084318,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 18690
    },
    {
      "epoch": 4.372120219857327,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8304,
      "step": 18691
    },
    {
      "epoch": 4.372354110630336,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 18692
    },
    {
      "epoch": 4.3725880014033445,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6925,
      "step": 18693
    },
    {
      "epoch": 4.372821892176353,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 18694
    },
    {
      "epoch": 4.373055782949363,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 18695
    },
    {
      "epoch": 4.373289673722372,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 18696
    },
    {
      "epoch": 4.373523564495381,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 18697
    },
    {
      "epoch": 4.3737574552683895,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8547,
      "step": 18698
    },
    {
      "epoch": 4.373991346041398,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 18699
    },
    {
      "epoch": 4.374225236814407,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 18700
    },
    {
      "epoch": 4.374225236814407,
      "eval_runtime": 4.6156,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 18700
    },
    {
      "epoch": 4.374459127587417,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 18701
    },
    {
      "epoch": 4.374693018360426,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5948,
      "step": 18702
    },
    {
      "epoch": 4.374926909133435,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 18703
    },
    {
      "epoch": 4.3751607999064435,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 18704
    },
    {
      "epoch": 4.375394690679452,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 18705
    },
    {
      "epoch": 4.375628581452462,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 18706
    },
    {
      "epoch": 4.375862472225471,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 18707
    },
    {
      "epoch": 4.37609636299848,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 18708
    },
    {
      "epoch": 4.3763302537714885,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 18709
    },
    {
      "epoch": 4.376564144544497,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 18710
    },
    {
      "epoch": 4.376798035317507,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7265,
      "step": 18711
    },
    {
      "epoch": 4.377031926090516,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 18712
    },
    {
      "epoch": 4.377265816863525,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 18713
    },
    {
      "epoch": 4.377499707636534,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 18714
    },
    {
      "epoch": 4.3777335984095425,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 18715
    },
    {
      "epoch": 4.377967489182552,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 18716
    },
    {
      "epoch": 4.378201379955561,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7209,
      "step": 18717
    },
    {
      "epoch": 4.37843527072857,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 18718
    },
    {
      "epoch": 4.378669161501579,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6568,
      "step": 18719
    },
    {
      "epoch": 4.3789030522745875,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 18720
    },
    {
      "epoch": 4.379136943047596,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 18721
    },
    {
      "epoch": 4.379370833820606,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 18722
    },
    {
      "epoch": 4.379604724593615,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.4827,
      "step": 18723
    },
    {
      "epoch": 4.379838615366624,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 18724
    },
    {
      "epoch": 4.380072506139633,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 18725
    },
    {
      "epoch": 4.3803063969126415,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 18726
    },
    {
      "epoch": 4.380540287685651,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5523,
      "step": 18727
    },
    {
      "epoch": 4.38077417845866,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9676,
      "step": 18728
    },
    {
      "epoch": 4.381008069231669,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5168,
      "step": 18729
    },
    {
      "epoch": 4.381241960004678,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 18730
    },
    {
      "epoch": 4.3814758507776865,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 18731
    },
    {
      "epoch": 4.381709741550695,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 18732
    },
    {
      "epoch": 4.381943632323705,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 18733
    },
    {
      "epoch": 4.382177523096714,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6126,
      "step": 18734
    },
    {
      "epoch": 4.382411413869723,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 18735
    },
    {
      "epoch": 4.382645304642732,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 18736
    },
    {
      "epoch": 4.3828791954157404,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.673,
      "step": 18737
    },
    {
      "epoch": 4.38311308618875,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 18738
    },
    {
      "epoch": 4.383346976961759,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 18739
    },
    {
      "epoch": 4.383580867734768,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5088,
      "step": 18740
    },
    {
      "epoch": 4.383814758507777,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.5768,
      "step": 18741
    },
    {
      "epoch": 4.3840486492807855,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 18742
    },
    {
      "epoch": 4.384282540053795,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 18743
    },
    {
      "epoch": 4.384516430826804,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 18744
    },
    {
      "epoch": 4.384750321599813,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 18745
    },
    {
      "epoch": 4.384984212372822,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 18746
    },
    {
      "epoch": 4.385218103145831,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 18747
    },
    {
      "epoch": 4.38545199391884,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 18748
    },
    {
      "epoch": 4.385685884691849,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5777,
      "step": 18749
    },
    {
      "epoch": 4.385919775464858,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 18750
    },
    {
      "epoch": 4.386153666237867,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.717,
      "step": 18751
    },
    {
      "epoch": 4.386387557010876,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7617,
      "step": 18752
    },
    {
      "epoch": 4.3866214477838845,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.57,
      "step": 18753
    },
    {
      "epoch": 4.386855338556894,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 18754
    },
    {
      "epoch": 4.387089229329903,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 18755
    },
    {
      "epoch": 4.387323120102912,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 18756
    },
    {
      "epoch": 4.387557010875921,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 18757
    },
    {
      "epoch": 4.38779090164893,
      "grad_norm": 10.75,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 18758
    },
    {
      "epoch": 4.388024792421939,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 18759
    },
    {
      "epoch": 4.388258683194948,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 18760
    },
    {
      "epoch": 4.388492573967957,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4728,
      "step": 18761
    },
    {
      "epoch": 4.388726464740966,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 18762
    },
    {
      "epoch": 4.388960355513975,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5714,
      "step": 18763
    },
    {
      "epoch": 4.3891942462869835,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 18764
    },
    {
      "epoch": 4.389428137059993,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4811,
      "step": 18765
    },
    {
      "epoch": 4.389662027833002,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 18766
    },
    {
      "epoch": 4.389895918606011,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 18767
    },
    {
      "epoch": 4.39012980937902,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 18768
    },
    {
      "epoch": 4.390363700152029,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 18769
    },
    {
      "epoch": 4.390597590925038,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 18770
    },
    {
      "epoch": 4.390831481698047,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.7064,
      "step": 18771
    },
    {
      "epoch": 4.391065372471056,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5934,
      "step": 18772
    },
    {
      "epoch": 4.391299263244065,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.456,
      "step": 18773
    },
    {
      "epoch": 4.391533154017074,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 18774
    },
    {
      "epoch": 4.391767044790083,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 18775
    },
    {
      "epoch": 4.392000935563092,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5857,
      "step": 18776
    },
    {
      "epoch": 4.392234826336101,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 18777
    },
    {
      "epoch": 4.39246871710911,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7038,
      "step": 18778
    },
    {
      "epoch": 4.392702607882119,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 18779
    },
    {
      "epoch": 4.3929364986551285,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 18780
    },
    {
      "epoch": 4.393170389428137,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 18781
    },
    {
      "epoch": 4.393404280201146,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6505,
      "step": 18782
    },
    {
      "epoch": 4.393638170974155,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.793,
      "step": 18783
    },
    {
      "epoch": 4.393872061747164,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 18784
    },
    {
      "epoch": 4.394105952520173,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7081,
      "step": 18785
    },
    {
      "epoch": 4.394339843293182,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 18786
    },
    {
      "epoch": 4.394573734066191,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5314,
      "step": 18787
    },
    {
      "epoch": 4.3948076248392,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 18788
    },
    {
      "epoch": 4.395041515612209,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 18789
    },
    {
      "epoch": 4.395275406385218,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 18790
    },
    {
      "epoch": 4.3955092971582275,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 18791
    },
    {
      "epoch": 4.395743187931236,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 18792
    },
    {
      "epoch": 4.395977078704245,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 18793
    },
    {
      "epoch": 4.396210969477254,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 18794
    },
    {
      "epoch": 4.396444860250263,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 18795
    },
    {
      "epoch": 4.3966787510232725,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6553,
      "step": 18796
    },
    {
      "epoch": 4.396912641796281,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 18797
    },
    {
      "epoch": 4.39714653256929,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8201,
      "step": 18798
    },
    {
      "epoch": 4.397380423342299,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 18799
    },
    {
      "epoch": 4.397614314115308,
      "grad_norm": 7.21875,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 18800
    },
    {
      "epoch": 4.397614314115308,
      "eval_runtime": 4.6196,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 18800
    },
    {
      "epoch": 4.397848204888318,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3136,
      "step": 18801
    },
    {
      "epoch": 4.3980820956613265,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 18802
    },
    {
      "epoch": 4.398315986434335,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6609,
      "step": 18803
    },
    {
      "epoch": 4.398549877207344,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7205,
      "step": 18804
    },
    {
      "epoch": 4.398783767980353,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 18805
    },
    {
      "epoch": 4.399017658753362,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 18806
    },
    {
      "epoch": 4.3992515495263715,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6699,
      "step": 18807
    },
    {
      "epoch": 4.39948544029938,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 18808
    },
    {
      "epoch": 4.399719331072389,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6055,
      "step": 18809
    },
    {
      "epoch": 4.399953221845398,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5771,
      "step": 18810
    },
    {
      "epoch": 4.400187112618407,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 18811
    },
    {
      "epoch": 4.400421003391417,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 18812
    },
    {
      "epoch": 4.4006548941644255,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 18813
    },
    {
      "epoch": 4.400888784937434,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 18814
    },
    {
      "epoch": 4.401122675710443,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 18815
    },
    {
      "epoch": 4.401356566483452,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 18816
    },
    {
      "epoch": 4.401590457256461,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 18817
    },
    {
      "epoch": 4.4018243480294705,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0122,
      "step": 18818
    },
    {
      "epoch": 4.402058238802479,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5616,
      "step": 18819
    },
    {
      "epoch": 4.402292129575488,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5756,
      "step": 18820
    },
    {
      "epoch": 4.402526020348497,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 18821
    },
    {
      "epoch": 4.402759911121506,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7244,
      "step": 18822
    },
    {
      "epoch": 4.402993801894516,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 18823
    },
    {
      "epoch": 4.4032276926675245,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 18824
    },
    {
      "epoch": 4.403461583440533,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 18825
    },
    {
      "epoch": 4.403695474213542,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 18826
    },
    {
      "epoch": 4.403929364986551,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 18827
    },
    {
      "epoch": 4.404163255759561,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 18828
    },
    {
      "epoch": 4.4043971465325695,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 18829
    },
    {
      "epoch": 4.404631037305578,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 18830
    },
    {
      "epoch": 4.404864928078587,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 18831
    },
    {
      "epoch": 4.405098818851596,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 18832
    },
    {
      "epoch": 4.405332709624606,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 18833
    },
    {
      "epoch": 4.405566600397615,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 18834
    },
    {
      "epoch": 4.4058004911706234,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8532,
      "step": 18835
    },
    {
      "epoch": 4.406034381943632,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7236,
      "step": 18836
    },
    {
      "epoch": 4.406268272716641,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6471,
      "step": 18837
    },
    {
      "epoch": 4.40650216348965,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2775,
      "step": 18838
    },
    {
      "epoch": 4.40673605426266,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 18839
    },
    {
      "epoch": 4.4069699450356685,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7872,
      "step": 18840
    },
    {
      "epoch": 4.407203835808677,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8382,
      "step": 18841
    },
    {
      "epoch": 4.407437726581686,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 18842
    },
    {
      "epoch": 4.407671617354695,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 18843
    },
    {
      "epoch": 4.407905508127705,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 18844
    },
    {
      "epoch": 4.408139398900714,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 18845
    },
    {
      "epoch": 4.408373289673722,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 18846
    },
    {
      "epoch": 4.408607180446731,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 18847
    },
    {
      "epoch": 4.40884107121974,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 18848
    },
    {
      "epoch": 4.409074961992749,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 18849
    },
    {
      "epoch": 4.409308852765759,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 18850
    },
    {
      "epoch": 4.4095427435387675,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 18851
    },
    {
      "epoch": 4.409776634311776,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 18852
    },
    {
      "epoch": 4.410010525084785,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 18853
    },
    {
      "epoch": 4.410244415857794,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.6035,
      "step": 18854
    },
    {
      "epoch": 4.410478306630804,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 18855
    },
    {
      "epoch": 4.410712197403813,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5263,
      "step": 18856
    },
    {
      "epoch": 4.410946088176821,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 18857
    },
    {
      "epoch": 4.41117997894983,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 18858
    },
    {
      "epoch": 4.411413869722839,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 18859
    },
    {
      "epoch": 4.411647760495849,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 18860
    },
    {
      "epoch": 4.411881651268858,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 18861
    },
    {
      "epoch": 4.4121155420418665,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 18862
    },
    {
      "epoch": 4.412349432814875,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 18863
    },
    {
      "epoch": 4.412583323587884,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 18864
    },
    {
      "epoch": 4.412817214360894,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 18865
    },
    {
      "epoch": 4.413051105133903,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6652,
      "step": 18866
    },
    {
      "epoch": 4.413284995906912,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 18867
    },
    {
      "epoch": 4.41351888667992,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 18868
    },
    {
      "epoch": 4.413752777452929,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 18869
    },
    {
      "epoch": 4.413986668225938,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 18870
    },
    {
      "epoch": 4.414220558998948,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 18871
    },
    {
      "epoch": 4.414454449771957,
      "grad_norm": 7.8125,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 18872
    },
    {
      "epoch": 4.4146883405449655,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.552,
      "step": 18873
    },
    {
      "epoch": 4.414922231317974,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 18874
    },
    {
      "epoch": 4.415156122090983,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5993,
      "step": 18875
    },
    {
      "epoch": 4.415390012863993,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6402,
      "step": 18876
    },
    {
      "epoch": 4.415623903637002,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 18877
    },
    {
      "epoch": 4.415857794410011,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 18878
    },
    {
      "epoch": 4.416091685183019,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 18879
    },
    {
      "epoch": 4.416325575956028,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 18880
    },
    {
      "epoch": 4.416559466729037,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 18881
    },
    {
      "epoch": 4.416793357502047,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 18882
    },
    {
      "epoch": 4.417027248275056,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 18883
    },
    {
      "epoch": 4.4172611390480645,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 18884
    },
    {
      "epoch": 4.417495029821073,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 18885
    },
    {
      "epoch": 4.417728920594082,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6672,
      "step": 18886
    },
    {
      "epoch": 4.417962811367092,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6908,
      "step": 18887
    },
    {
      "epoch": 4.418196702140101,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 18888
    },
    {
      "epoch": 4.41843059291311,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 18889
    },
    {
      "epoch": 4.418664483686118,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 18890
    },
    {
      "epoch": 4.418898374459127,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5262,
      "step": 18891
    },
    {
      "epoch": 4.419132265232137,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5992,
      "step": 18892
    },
    {
      "epoch": 4.419366156005146,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 18893
    },
    {
      "epoch": 4.419600046778155,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6901,
      "step": 18894
    },
    {
      "epoch": 4.4198339375511635,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8007,
      "step": 18895
    },
    {
      "epoch": 4.420067828324172,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 18896
    },
    {
      "epoch": 4.420301719097182,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.147,
      "step": 18897
    },
    {
      "epoch": 4.420535609870191,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 18898
    },
    {
      "epoch": 4.4207695006432,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9061,
      "step": 18899
    },
    {
      "epoch": 4.421003391416209,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 18900
    },
    {
      "epoch": 4.421003391416209,
      "eval_runtime": 4.5968,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 18900
    },
    {
      "epoch": 4.421237282189217,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7474,
      "step": 18901
    },
    {
      "epoch": 4.421471172962226,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 1.5665,
      "step": 18902
    },
    {
      "epoch": 4.421705063735236,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 18903
    },
    {
      "epoch": 4.421938954508245,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 18904
    },
    {
      "epoch": 4.422172845281254,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 18905
    },
    {
      "epoch": 4.4224067360542625,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 18906
    },
    {
      "epoch": 4.422640626827271,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7355,
      "step": 18907
    },
    {
      "epoch": 4.422874517600281,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6425,
      "step": 18908
    },
    {
      "epoch": 4.42310840837329,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 18909
    },
    {
      "epoch": 4.423342299146299,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 18910
    },
    {
      "epoch": 4.423576189919308,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 18911
    },
    {
      "epoch": 4.423810080692316,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7531,
      "step": 18912
    },
    {
      "epoch": 4.424043971465325,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.3838,
      "step": 18913
    },
    {
      "epoch": 4.424277862238335,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 18914
    },
    {
      "epoch": 4.424511753011344,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 18915
    },
    {
      "epoch": 4.424745643784353,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6819,
      "step": 18916
    },
    {
      "epoch": 4.4249795345573615,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 18917
    },
    {
      "epoch": 4.42521342533037,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 18918
    },
    {
      "epoch": 4.42544731610338,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 18919
    },
    {
      "epoch": 4.425681206876389,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 18920
    },
    {
      "epoch": 4.425915097649398,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 18921
    },
    {
      "epoch": 4.426148988422407,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 18922
    },
    {
      "epoch": 4.426382879195415,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 18923
    },
    {
      "epoch": 4.426616769968425,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 18924
    },
    {
      "epoch": 4.426850660741434,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 18925
    },
    {
      "epoch": 4.427084551514443,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 18926
    },
    {
      "epoch": 4.427318442287452,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6098,
      "step": 18927
    },
    {
      "epoch": 4.4275523330604605,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 18928
    },
    {
      "epoch": 4.42778622383347,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4734,
      "step": 18929
    },
    {
      "epoch": 4.428020114606479,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 18930
    },
    {
      "epoch": 4.428254005379488,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4406,
      "step": 18931
    },
    {
      "epoch": 4.428487896152497,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6957,
      "step": 18932
    },
    {
      "epoch": 4.4287217869255056,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.3906,
      "step": 18933
    },
    {
      "epoch": 4.428955677698514,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 18934
    },
    {
      "epoch": 4.429189568471524,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 18935
    },
    {
      "epoch": 4.429423459244533,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 18936
    },
    {
      "epoch": 4.429657350017542,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 18937
    },
    {
      "epoch": 4.429891240790551,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 18938
    },
    {
      "epoch": 4.4301251315635595,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 18939
    },
    {
      "epoch": 4.430359022336569,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6621,
      "step": 18940
    },
    {
      "epoch": 4.430592913109578,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6648,
      "step": 18941
    },
    {
      "epoch": 4.430826803882587,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 18942
    },
    {
      "epoch": 4.431060694655596,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 18943
    },
    {
      "epoch": 4.4312945854286045,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8183,
      "step": 18944
    },
    {
      "epoch": 4.431528476201613,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 18945
    },
    {
      "epoch": 4.431762366974623,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 18946
    },
    {
      "epoch": 4.431996257747632,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7576,
      "step": 18947
    },
    {
      "epoch": 4.432230148520641,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 18948
    },
    {
      "epoch": 4.43246403929365,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 18949
    },
    {
      "epoch": 4.4326979300666585,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7908,
      "step": 18950
    },
    {
      "epoch": 4.432931820839668,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 18951
    },
    {
      "epoch": 4.433165711612677,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 18952
    },
    {
      "epoch": 4.433399602385686,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 18953
    },
    {
      "epoch": 4.433633493158695,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 18954
    },
    {
      "epoch": 4.4338673839317035,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 18955
    },
    {
      "epoch": 4.434101274704713,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 18956
    },
    {
      "epoch": 4.434335165477722,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 18957
    },
    {
      "epoch": 4.434569056250731,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 18958
    },
    {
      "epoch": 4.43480294702374,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 18959
    },
    {
      "epoch": 4.435036837796749,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 18960
    },
    {
      "epoch": 4.435270728569758,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 18961
    },
    {
      "epoch": 4.435504619342767,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 18962
    },
    {
      "epoch": 4.435738510115776,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 18963
    },
    {
      "epoch": 4.435972400888785,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 18964
    },
    {
      "epoch": 4.436206291661794,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 18965
    },
    {
      "epoch": 4.4364401824348025,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 18966
    },
    {
      "epoch": 4.436674073207812,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 18967
    },
    {
      "epoch": 4.436907963980821,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6878,
      "step": 18968
    },
    {
      "epoch": 4.43714185475383,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 18969
    },
    {
      "epoch": 4.437375745526839,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 18970
    },
    {
      "epoch": 4.437609636299848,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 18971
    },
    {
      "epoch": 4.437843527072857,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 18972
    },
    {
      "epoch": 4.438077417845866,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 18973
    },
    {
      "epoch": 4.438311308618875,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 18974
    },
    {
      "epoch": 4.438545199391884,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 18975
    },
    {
      "epoch": 4.438779090164893,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 18976
    },
    {
      "epoch": 4.4390129809379015,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 18977
    },
    {
      "epoch": 4.439246871710911,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 18978
    },
    {
      "epoch": 4.43948076248392,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 18979
    },
    {
      "epoch": 4.439714653256929,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 18980
    },
    {
      "epoch": 4.439948544029938,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7866,
      "step": 18981
    },
    {
      "epoch": 4.440182434802947,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6088,
      "step": 18982
    },
    {
      "epoch": 4.440416325575956,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 18983
    },
    {
      "epoch": 4.440650216348965,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 18984
    },
    {
      "epoch": 4.440884107121974,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 18985
    },
    {
      "epoch": 4.441117997894983,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6364,
      "step": 18986
    },
    {
      "epoch": 4.441351888667992,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 18987
    },
    {
      "epoch": 4.441585779441001,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 18988
    },
    {
      "epoch": 4.44181967021401,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.424,
      "step": 18989
    },
    {
      "epoch": 4.442053560987019,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6129,
      "step": 18990
    },
    {
      "epoch": 4.442287451760028,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 18991
    },
    {
      "epoch": 4.442521342533037,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 18992
    },
    {
      "epoch": 4.4427552333060465,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5554,
      "step": 18993
    },
    {
      "epoch": 4.442989124079055,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 18994
    },
    {
      "epoch": 4.443223014852064,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 18995
    },
    {
      "epoch": 4.443456905625073,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 18996
    },
    {
      "epoch": 4.443690796398082,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 18997
    },
    {
      "epoch": 4.443924687171091,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 18998
    },
    {
      "epoch": 4.4441585779441,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5548,
      "step": 18999
    },
    {
      "epoch": 4.444392468717109,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 19000
    },
    {
      "epoch": 4.444392468717109,
      "eval_runtime": 4.6868,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 19000
    },
    {
      "epoch": 4.444626359490118,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 19001
    },
    {
      "epoch": 4.444860250263127,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7947,
      "step": 19002
    },
    {
      "epoch": 4.445094141036136,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6293,
      "step": 19003
    },
    {
      "epoch": 4.4453280318091455,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 19004
    },
    {
      "epoch": 4.445561922582154,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 19005
    },
    {
      "epoch": 4.445795813355163,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6159,
      "step": 19006
    },
    {
      "epoch": 4.446029704128172,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 19007
    },
    {
      "epoch": 4.446263594901181,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 19008
    },
    {
      "epoch": 4.446497485674191,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 19009
    },
    {
      "epoch": 4.446731376447199,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 19010
    },
    {
      "epoch": 4.446965267220208,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 19011
    },
    {
      "epoch": 4.447199157993217,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 19012
    },
    {
      "epoch": 4.447433048766226,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 19013
    },
    {
      "epoch": 4.447666939539236,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 19014
    },
    {
      "epoch": 4.4479008303122445,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 19015
    },
    {
      "epoch": 4.448134721085253,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 19016
    },
    {
      "epoch": 4.448368611858262,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 19017
    },
    {
      "epoch": 4.448602502631271,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5893,
      "step": 19018
    },
    {
      "epoch": 4.44883639340428,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7419,
      "step": 19019
    },
    {
      "epoch": 4.44907028417729,
      "grad_norm": 7.40625,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 19020
    },
    {
      "epoch": 4.449304174950298,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 19021
    },
    {
      "epoch": 4.449538065723307,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 19022
    },
    {
      "epoch": 4.449771956496316,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 19023
    },
    {
      "epoch": 4.450005847269325,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 19024
    },
    {
      "epoch": 4.450239738042335,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 19025
    },
    {
      "epoch": 4.4504736288153435,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 19026
    },
    {
      "epoch": 4.450707519588352,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5535,
      "step": 19027
    },
    {
      "epoch": 4.450941410361361,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 19028
    },
    {
      "epoch": 4.45117530113437,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 19029
    },
    {
      "epoch": 4.451409191907379,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 19030
    },
    {
      "epoch": 4.4516430826803886,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 19031
    },
    {
      "epoch": 4.451876973453397,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5852,
      "step": 19032
    },
    {
      "epoch": 4.452110864226406,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 19033
    },
    {
      "epoch": 4.452344754999415,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5,
      "step": 19034
    },
    {
      "epoch": 4.452578645772424,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 19035
    },
    {
      "epoch": 4.452812536545434,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 19036
    },
    {
      "epoch": 4.4530464273184425,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6034,
      "step": 19037
    },
    {
      "epoch": 4.453280318091451,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 19038
    },
    {
      "epoch": 4.45351420886446,
      "grad_norm": 11.5625,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 19039
    },
    {
      "epoch": 4.453748099637469,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 19040
    },
    {
      "epoch": 4.453981990410479,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 19041
    },
    {
      "epoch": 4.4542158811834875,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 19042
    },
    {
      "epoch": 4.454449771956496,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 19043
    },
    {
      "epoch": 4.454683662729505,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 19044
    },
    {
      "epoch": 4.454917553502514,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 19045
    },
    {
      "epoch": 4.455151444275524,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 19046
    },
    {
      "epoch": 4.455385335048533,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 19047
    },
    {
      "epoch": 4.4556192258215415,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 19048
    },
    {
      "epoch": 4.45585311659455,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 19049
    },
    {
      "epoch": 4.456087007367559,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5961,
      "step": 19050
    },
    {
      "epoch": 4.456320898140568,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 19051
    },
    {
      "epoch": 4.456554788913578,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6575,
      "step": 19052
    },
    {
      "epoch": 4.4567886796865865,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 19053
    },
    {
      "epoch": 4.457022570459595,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7229,
      "step": 19054
    },
    {
      "epoch": 4.457256461232604,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 19055
    },
    {
      "epoch": 4.457490352005613,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 19056
    },
    {
      "epoch": 4.457724242778623,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5985,
      "step": 19057
    },
    {
      "epoch": 4.457958133551632,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7221,
      "step": 19058
    },
    {
      "epoch": 4.4581920243246405,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 19059
    },
    {
      "epoch": 4.458425915097649,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5699,
      "step": 19060
    },
    {
      "epoch": 4.458659805870658,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 19061
    },
    {
      "epoch": 4.458893696643667,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 19062
    },
    {
      "epoch": 4.459127587416677,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 19063
    },
    {
      "epoch": 4.4593614781896855,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 19064
    },
    {
      "epoch": 4.459595368962694,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 19065
    },
    {
      "epoch": 4.459829259735703,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.28,
      "step": 19066
    },
    {
      "epoch": 4.460063150508712,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7359,
      "step": 19067
    },
    {
      "epoch": 4.460297041281722,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7066,
      "step": 19068
    },
    {
      "epoch": 4.460530932054731,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7515,
      "step": 19069
    },
    {
      "epoch": 4.4607648228277395,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 19070
    },
    {
      "epoch": 4.460998713600748,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9303,
      "step": 19071
    },
    {
      "epoch": 4.461232604373757,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.5545,
      "step": 19072
    },
    {
      "epoch": 4.461466495146767,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 19073
    },
    {
      "epoch": 4.461700385919776,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 19074
    },
    {
      "epoch": 4.4619342766927845,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 19075
    },
    {
      "epoch": 4.462168167465793,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 19076
    },
    {
      "epoch": 4.462402058238802,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 19077
    },
    {
      "epoch": 4.462635949011812,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7883,
      "step": 19078
    },
    {
      "epoch": 4.462869839784821,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6952,
      "step": 19079
    },
    {
      "epoch": 4.46310373055783,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 19080
    },
    {
      "epoch": 4.4633376213308384,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 19081
    },
    {
      "epoch": 4.463571512103847,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6201,
      "step": 19082
    },
    {
      "epoch": 4.463805402876856,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 19083
    },
    {
      "epoch": 4.464039293649866,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5095,
      "step": 19084
    },
    {
      "epoch": 4.464273184422875,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 19085
    },
    {
      "epoch": 4.4645070751958835,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.533,
      "step": 19086
    },
    {
      "epoch": 4.464740965968892,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 19087
    },
    {
      "epoch": 4.464974856741901,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6471,
      "step": 19088
    },
    {
      "epoch": 4.465208747514911,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 19089
    },
    {
      "epoch": 4.46544263828792,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 19090
    },
    {
      "epoch": 4.465676529060929,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 19091
    },
    {
      "epoch": 4.465910419833937,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 19092
    },
    {
      "epoch": 4.466144310606946,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 19093
    },
    {
      "epoch": 4.466378201379955,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 19094
    },
    {
      "epoch": 4.466612092152965,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 19095
    },
    {
      "epoch": 4.466845982925974,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 19096
    },
    {
      "epoch": 4.4670798736989825,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7628,
      "step": 19097
    },
    {
      "epoch": 4.467313764471991,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6803,
      "step": 19098
    },
    {
      "epoch": 4.467547655245,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6938,
      "step": 19099
    },
    {
      "epoch": 4.46778154601801,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7107,
      "step": 19100
    },
    {
      "epoch": 4.46778154601801,
      "eval_runtime": 4.6198,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 19100
    },
    {
      "epoch": 4.468015436791019,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 19101
    },
    {
      "epoch": 4.468249327564028,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 19102
    },
    {
      "epoch": 4.468483218337036,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6579,
      "step": 19103
    },
    {
      "epoch": 4.468717109110045,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 19104
    },
    {
      "epoch": 4.468950999883055,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 19105
    },
    {
      "epoch": 4.469184890656064,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 19106
    },
    {
      "epoch": 4.469418781429073,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 19107
    },
    {
      "epoch": 4.4696526722020815,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 19108
    },
    {
      "epoch": 4.46988656297509,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 19109
    },
    {
      "epoch": 4.4701204537481,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 19110
    },
    {
      "epoch": 4.470354344521109,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 19111
    },
    {
      "epoch": 4.470588235294118,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 19112
    },
    {
      "epoch": 4.470822126067127,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 19113
    },
    {
      "epoch": 4.471056016840135,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6195,
      "step": 19114
    },
    {
      "epoch": 4.471289907613144,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8688,
      "step": 19115
    },
    {
      "epoch": 4.471523798386154,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 19116
    },
    {
      "epoch": 4.471757689159163,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 19117
    },
    {
      "epoch": 4.471991579932172,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 19118
    },
    {
      "epoch": 4.4722254707051805,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5751,
      "step": 19119
    },
    {
      "epoch": 4.472459361478189,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 19120
    },
    {
      "epoch": 4.472693252251199,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 19121
    },
    {
      "epoch": 4.472927143024208,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 19122
    },
    {
      "epoch": 4.473161033797217,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 19123
    },
    {
      "epoch": 4.473394924570226,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7495,
      "step": 19124
    },
    {
      "epoch": 4.473628815343234,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 19125
    },
    {
      "epoch": 4.473862706116243,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 19126
    },
    {
      "epoch": 4.474096596889253,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4294,
      "step": 19127
    },
    {
      "epoch": 4.474330487662262,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 19128
    },
    {
      "epoch": 4.474564378435271,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 19129
    },
    {
      "epoch": 4.4747982692082795,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 19130
    },
    {
      "epoch": 4.475032159981288,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 19131
    },
    {
      "epoch": 4.475266050754298,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 19132
    },
    {
      "epoch": 4.475499941527307,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 19133
    },
    {
      "epoch": 4.475733832300316,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 19134
    },
    {
      "epoch": 4.475967723073325,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7671,
      "step": 19135
    },
    {
      "epoch": 4.476201613846333,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 19136
    },
    {
      "epoch": 4.476435504619343,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7082,
      "step": 19137
    },
    {
      "epoch": 4.476669395392352,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 19138
    },
    {
      "epoch": 4.476903286165361,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 19139
    },
    {
      "epoch": 4.47713717693837,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 19140
    },
    {
      "epoch": 4.4773710677113785,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 19141
    },
    {
      "epoch": 4.477604958484388,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 19142
    },
    {
      "epoch": 4.477838849257397,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 19143
    },
    {
      "epoch": 4.478072740030406,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 19144
    },
    {
      "epoch": 4.478306630803415,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 19145
    },
    {
      "epoch": 4.478540521576424,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6784,
      "step": 19146
    },
    {
      "epoch": 4.478774412349432,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 19147
    },
    {
      "epoch": 4.479008303122442,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 19148
    },
    {
      "epoch": 4.479242193895451,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 19149
    },
    {
      "epoch": 4.47947608466846,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 19150
    },
    {
      "epoch": 4.479709975441469,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5546,
      "step": 19151
    },
    {
      "epoch": 4.4799438662144775,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 19152
    },
    {
      "epoch": 4.480177756987487,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6468,
      "step": 19153
    },
    {
      "epoch": 4.480411647760496,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 19154
    },
    {
      "epoch": 4.480645538533505,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 19155
    },
    {
      "epoch": 4.480879429306514,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 19156
    },
    {
      "epoch": 4.481113320079523,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7814,
      "step": 19157
    },
    {
      "epoch": 4.481347210852531,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6347,
      "step": 19158
    },
    {
      "epoch": 4.481581101625541,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 19159
    },
    {
      "epoch": 4.48181499239855,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6489,
      "step": 19160
    },
    {
      "epoch": 4.482048883171559,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6825,
      "step": 19161
    },
    {
      "epoch": 4.482282773944568,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.736,
      "step": 19162
    },
    {
      "epoch": 4.4825166647175765,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6957,
      "step": 19163
    },
    {
      "epoch": 4.482750555490586,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 19164
    },
    {
      "epoch": 4.482984446263595,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 19165
    },
    {
      "epoch": 4.483218337036604,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 19166
    },
    {
      "epoch": 4.483452227809613,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 19167
    },
    {
      "epoch": 4.483686118582622,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 19168
    },
    {
      "epoch": 4.483920009355631,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 19169
    },
    {
      "epoch": 4.48415390012864,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6796,
      "step": 19170
    },
    {
      "epoch": 4.484387790901649,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 19171
    },
    {
      "epoch": 4.484621681674658,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5582,
      "step": 19172
    },
    {
      "epoch": 4.484855572447667,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 19173
    },
    {
      "epoch": 4.485089463220676,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 19174
    },
    {
      "epoch": 4.485323353993685,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 19175
    },
    {
      "epoch": 4.485557244766694,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 19176
    },
    {
      "epoch": 4.485791135539703,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 19177
    },
    {
      "epoch": 4.486025026312712,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 19178
    },
    {
      "epoch": 4.4862589170857206,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 19179
    },
    {
      "epoch": 4.48649280785873,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 19180
    },
    {
      "epoch": 4.486726698631739,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 19181
    },
    {
      "epoch": 4.486960589404748,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 19182
    },
    {
      "epoch": 4.487194480177757,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 19183
    },
    {
      "epoch": 4.487428370950766,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 19184
    },
    {
      "epoch": 4.487662261723775,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 19185
    },
    {
      "epoch": 4.487896152496784,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 19186
    },
    {
      "epoch": 4.488130043269793,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 19187
    },
    {
      "epoch": 4.488363934042802,
      "grad_norm": 2.5,
      "learning_rate": 3e-05,
      "loss": 1.4627,
      "step": 19188
    },
    {
      "epoch": 4.488597824815811,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 19189
    },
    {
      "epoch": 4.48883171558882,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7119,
      "step": 19190
    },
    {
      "epoch": 4.489065606361829,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 19191
    },
    {
      "epoch": 4.489299497134838,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 19192
    },
    {
      "epoch": 4.489533387907847,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 19193
    },
    {
      "epoch": 4.489767278680856,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5765,
      "step": 19194
    },
    {
      "epoch": 4.4900011694538655,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 19195
    },
    {
      "epoch": 4.490235060226874,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 19196
    },
    {
      "epoch": 4.490468950999883,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 19197
    },
    {
      "epoch": 4.490702841772892,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 19198
    },
    {
      "epoch": 4.490936732545901,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 19199
    },
    {
      "epoch": 4.49117062331891,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 19200
    },
    {
      "epoch": 4.49117062331891,
      "eval_runtime": 4.601,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 19200
    },
    {
      "epoch": 4.491404514091919,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 19201
    },
    {
      "epoch": 4.491638404864928,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7104,
      "step": 19202
    },
    {
      "epoch": 4.491872295637937,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6593,
      "step": 19203
    },
    {
      "epoch": 4.492106186410946,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 19204
    },
    {
      "epoch": 4.492340077183955,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 19205
    },
    {
      "epoch": 4.4925739679569645,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7212,
      "step": 19206
    },
    {
      "epoch": 4.492807858729973,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 19207
    },
    {
      "epoch": 4.493041749502982,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 19208
    },
    {
      "epoch": 4.493275640275991,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 19209
    },
    {
      "epoch": 4.493509531049,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7698,
      "step": 19210
    },
    {
      "epoch": 4.493743421822009,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8242,
      "step": 19211
    },
    {
      "epoch": 4.493977312595018,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6928,
      "step": 19212
    },
    {
      "epoch": 4.494211203368027,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 19213
    },
    {
      "epoch": 4.494445094141036,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 19214
    },
    {
      "epoch": 4.494678984914045,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6648,
      "step": 19215
    },
    {
      "epoch": 4.494912875687054,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 19216
    },
    {
      "epoch": 4.4951467664600635,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 19217
    },
    {
      "epoch": 4.495380657233072,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 19218
    },
    {
      "epoch": 4.495614548006081,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 19219
    },
    {
      "epoch": 4.49584843877909,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6797,
      "step": 19220
    },
    {
      "epoch": 4.496082329552099,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.713,
      "step": 19221
    },
    {
      "epoch": 4.496316220325109,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 19222
    },
    {
      "epoch": 4.496550111098117,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 19223
    },
    {
      "epoch": 4.496784001871126,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 19224
    },
    {
      "epoch": 4.497017892644135,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 19225
    },
    {
      "epoch": 4.497251783417144,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 19226
    },
    {
      "epoch": 4.497485674190154,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 19227
    },
    {
      "epoch": 4.4977195649631625,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5978,
      "step": 19228
    },
    {
      "epoch": 4.497953455736171,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 19229
    },
    {
      "epoch": 4.49818734650918,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6263,
      "step": 19230
    },
    {
      "epoch": 4.498421237282189,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7447,
      "step": 19231
    },
    {
      "epoch": 4.498655128055198,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6771,
      "step": 19232
    },
    {
      "epoch": 4.498889018828208,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 19233
    },
    {
      "epoch": 4.499122909601216,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 19234
    },
    {
      "epoch": 4.499356800374225,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 19235
    },
    {
      "epoch": 4.499590691147234,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 19236
    },
    {
      "epoch": 4.499824581920243,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 19237
    },
    {
      "epoch": 4.500058472693253,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 19238
    },
    {
      "epoch": 4.5002923634662615,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 19239
    },
    {
      "epoch": 4.50052625423927,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5906,
      "step": 19240
    },
    {
      "epoch": 4.500760145012279,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 19241
    },
    {
      "epoch": 4.500994035785288,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 19242
    },
    {
      "epoch": 4.501227926558297,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 19243
    },
    {
      "epoch": 4.501461817331307,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7115,
      "step": 19244
    },
    {
      "epoch": 4.501695708104315,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 19245
    },
    {
      "epoch": 4.501929598877324,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7147,
      "step": 19246
    },
    {
      "epoch": 4.502163489650333,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 19247
    },
    {
      "epoch": 4.502397380423342,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 19248
    },
    {
      "epoch": 4.502631271196352,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 19249
    },
    {
      "epoch": 4.5028651619693605,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7732,
      "step": 19250
    },
    {
      "epoch": 4.503099052742369,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 19251
    },
    {
      "epoch": 4.503332943515378,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 19252
    },
    {
      "epoch": 4.503566834288387,
      "grad_norm": 7.625,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 19253
    },
    {
      "epoch": 4.503800725061396,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5755,
      "step": 19254
    },
    {
      "epoch": 4.504034615834406,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7395,
      "step": 19255
    },
    {
      "epoch": 4.504268506607414,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 19256
    },
    {
      "epoch": 4.504502397380423,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.718,
      "step": 19257
    },
    {
      "epoch": 4.504736288153432,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 19258
    },
    {
      "epoch": 4.504970178926442,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 19259
    },
    {
      "epoch": 4.505204069699451,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 19260
    },
    {
      "epoch": 4.5054379604724595,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 19261
    },
    {
      "epoch": 4.505671851245468,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 19262
    },
    {
      "epoch": 4.505905742018477,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5143,
      "step": 19263
    },
    {
      "epoch": 4.506139632791486,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7622,
      "step": 19264
    },
    {
      "epoch": 4.506373523564496,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 19265
    },
    {
      "epoch": 4.506607414337505,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 19266
    },
    {
      "epoch": 4.506841305110513,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 19267
    },
    {
      "epoch": 4.507075195883522,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 19268
    },
    {
      "epoch": 4.507309086656531,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 19269
    },
    {
      "epoch": 4.507542977429541,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6752,
      "step": 19270
    },
    {
      "epoch": 4.50777686820255,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.743,
      "step": 19271
    },
    {
      "epoch": 4.5080107589755585,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 19272
    },
    {
      "epoch": 4.508244649748567,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 19273
    },
    {
      "epoch": 4.508478540521576,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 19274
    },
    {
      "epoch": 4.508712431294585,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 19275
    },
    {
      "epoch": 4.508946322067595,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 19276
    },
    {
      "epoch": 4.5091802128406036,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 19277
    },
    {
      "epoch": 4.509414103613612,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 19278
    },
    {
      "epoch": 4.509647994386621,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4775,
      "step": 19279
    },
    {
      "epoch": 4.50988188515963,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 19280
    },
    {
      "epoch": 4.51011577593264,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 19281
    },
    {
      "epoch": 4.510349666705649,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7335,
      "step": 19282
    },
    {
      "epoch": 4.5105835574786575,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 19283
    },
    {
      "epoch": 4.510817448251666,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 19284
    },
    {
      "epoch": 4.511051339024675,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 19285
    },
    {
      "epoch": 4.511285229797684,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4105,
      "step": 19286
    },
    {
      "epoch": 4.511519120570694,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 1.779,
      "step": 19287
    },
    {
      "epoch": 4.5117530113437025,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 19288
    },
    {
      "epoch": 4.511986902116711,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 19289
    },
    {
      "epoch": 4.51222079288972,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 19290
    },
    {
      "epoch": 4.51245468366273,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 19291
    },
    {
      "epoch": 4.512688574435739,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 19292
    },
    {
      "epoch": 4.512922465208748,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 19293
    },
    {
      "epoch": 4.5131563559817565,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4306,
      "step": 19294
    },
    {
      "epoch": 4.513390246754765,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 19295
    },
    {
      "epoch": 4.513624137527774,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 19296
    },
    {
      "epoch": 4.513858028300784,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6356,
      "step": 19297
    },
    {
      "epoch": 4.514091919073793,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 19298
    },
    {
      "epoch": 4.5143258098468015,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 19299
    },
    {
      "epoch": 4.51455970061981,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 19300
    },
    {
      "epoch": 4.51455970061981,
      "eval_runtime": 4.5993,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 19300
    },
    {
      "epoch": 4.514793591392819,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2051,
      "step": 19301
    },
    {
      "epoch": 4.515027482165829,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 19302
    },
    {
      "epoch": 4.515261372938838,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 19303
    },
    {
      "epoch": 4.515495263711847,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5712,
      "step": 19304
    },
    {
      "epoch": 4.5157291544848555,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 19305
    },
    {
      "epoch": 4.515963045257864,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6037,
      "step": 19306
    },
    {
      "epoch": 4.516196936030873,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 19307
    },
    {
      "epoch": 4.516430826803883,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 19308
    },
    {
      "epoch": 4.516664717576892,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 19309
    },
    {
      "epoch": 4.5168986083499005,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 19310
    },
    {
      "epoch": 4.517132499122909,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 19311
    },
    {
      "epoch": 4.517366389895919,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5325,
      "step": 19312
    },
    {
      "epoch": 4.517600280668928,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 19313
    },
    {
      "epoch": 4.517834171441937,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 19314
    },
    {
      "epoch": 4.518068062214946,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 19315
    },
    {
      "epoch": 4.5183019529879545,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7208,
      "step": 19316
    },
    {
      "epoch": 4.518535843760963,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 19317
    },
    {
      "epoch": 4.518769734533973,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 19318
    },
    {
      "epoch": 4.519003625306982,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 19319
    },
    {
      "epoch": 4.519237516079991,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 19320
    },
    {
      "epoch": 4.5194714068529995,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 19321
    },
    {
      "epoch": 4.519705297626008,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 19322
    },
    {
      "epoch": 4.519939188399018,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7172,
      "step": 19323
    },
    {
      "epoch": 4.520173079172027,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 19324
    },
    {
      "epoch": 4.520406969945036,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6698,
      "step": 19325
    },
    {
      "epoch": 4.520640860718045,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 19326
    },
    {
      "epoch": 4.5208747514910534,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 19327
    },
    {
      "epoch": 4.521108642264062,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 19328
    },
    {
      "epoch": 4.521342533037072,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 19329
    },
    {
      "epoch": 4.521576423810081,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 19330
    },
    {
      "epoch": 4.52181031458309,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4994,
      "step": 19331
    },
    {
      "epoch": 4.5220442053560985,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4563,
      "step": 19332
    },
    {
      "epoch": 4.522278096129107,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 19333
    },
    {
      "epoch": 4.522511986902117,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 19334
    },
    {
      "epoch": 4.522745877675126,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 19335
    },
    {
      "epoch": 4.522979768448135,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 19336
    },
    {
      "epoch": 4.523213659221144,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7329,
      "step": 19337
    },
    {
      "epoch": 4.523447549994152,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 19338
    },
    {
      "epoch": 4.523681440767161,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6506,
      "step": 19339
    },
    {
      "epoch": 4.523915331540171,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7728,
      "step": 19340
    },
    {
      "epoch": 4.52414922231318,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 19341
    },
    {
      "epoch": 4.524383113086189,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 19342
    },
    {
      "epoch": 4.5246170038591975,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 19343
    },
    {
      "epoch": 4.524850894632207,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6312,
      "step": 19344
    },
    {
      "epoch": 4.525084785405216,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 19345
    },
    {
      "epoch": 4.525318676178225,
      "grad_norm": 9.5625,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 19346
    },
    {
      "epoch": 4.525552566951234,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7406,
      "step": 19347
    },
    {
      "epoch": 4.525786457724243,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 19348
    },
    {
      "epoch": 4.526020348497251,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6531,
      "step": 19349
    },
    {
      "epoch": 4.526254239270261,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 19350
    },
    {
      "epoch": 4.52648813004327,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 19351
    },
    {
      "epoch": 4.526722020816279,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6787,
      "step": 19352
    },
    {
      "epoch": 4.526955911589288,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.755,
      "step": 19353
    },
    {
      "epoch": 4.5271898023622965,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 19354
    },
    {
      "epoch": 4.527423693135306,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7118,
      "step": 19355
    },
    {
      "epoch": 4.527657583908315,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 19356
    },
    {
      "epoch": 4.527891474681324,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 19357
    },
    {
      "epoch": 4.528125365454333,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 19358
    },
    {
      "epoch": 4.528359256227342,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7984,
      "step": 19359
    },
    {
      "epoch": 4.52859314700035,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 19360
    },
    {
      "epoch": 4.52882703777336,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 19361
    },
    {
      "epoch": 4.529060928546369,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 19362
    },
    {
      "epoch": 4.529294819319378,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 19363
    },
    {
      "epoch": 4.529528710092387,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 19364
    },
    {
      "epoch": 4.5297626008653955,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 19365
    },
    {
      "epoch": 4.529996491638405,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 19366
    },
    {
      "epoch": 4.530230382411414,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5882,
      "step": 19367
    },
    {
      "epoch": 4.530464273184423,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 19368
    },
    {
      "epoch": 4.530698163957432,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 19369
    },
    {
      "epoch": 4.530932054730441,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 19370
    },
    {
      "epoch": 4.531165945503449,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 19371
    },
    {
      "epoch": 4.531399836276459,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 19372
    },
    {
      "epoch": 4.531633727049468,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 19373
    },
    {
      "epoch": 4.531867617822477,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.7489,
      "step": 19374
    },
    {
      "epoch": 4.532101508595486,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 19375
    },
    {
      "epoch": 4.532335399368495,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 19376
    },
    {
      "epoch": 4.532569290141504,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 19377
    },
    {
      "epoch": 4.532803180914513,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.6127,
      "step": 19378
    },
    {
      "epoch": 4.533037071687522,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5789,
      "step": 19379
    },
    {
      "epoch": 4.533270962460531,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 19380
    },
    {
      "epoch": 4.53350485323354,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 19381
    },
    {
      "epoch": 4.533738744006549,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 19382
    },
    {
      "epoch": 4.533972634779558,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 19383
    },
    {
      "epoch": 4.534206525552567,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 19384
    },
    {
      "epoch": 4.534440416325576,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.951,
      "step": 19385
    },
    {
      "epoch": 4.534674307098585,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.063,
      "step": 19386
    },
    {
      "epoch": 4.534908197871594,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.755,
      "step": 19387
    },
    {
      "epoch": 4.535142088644603,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 19388
    },
    {
      "epoch": 4.535375979417612,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 19389
    },
    {
      "epoch": 4.535609870190621,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 19390
    },
    {
      "epoch": 4.53584376096363,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 19391
    },
    {
      "epoch": 4.536077651736639,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 19392
    },
    {
      "epoch": 4.536311542509648,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5688,
      "step": 19393
    },
    {
      "epoch": 4.536545433282657,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 19394
    },
    {
      "epoch": 4.536779324055666,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7708,
      "step": 19395
    },
    {
      "epoch": 4.537013214828675,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 19396
    },
    {
      "epoch": 4.537247105601684,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6833,
      "step": 19397
    },
    {
      "epoch": 4.537480996374693,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 19398
    },
    {
      "epoch": 4.537714887147702,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7166,
      "step": 19399
    },
    {
      "epoch": 4.537948777920711,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 19400
    },
    {
      "epoch": 4.537948777920711,
      "eval_runtime": 4.6218,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 19400
    },
    {
      "epoch": 4.53818266869372,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 19401
    },
    {
      "epoch": 4.538416559466729,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 19402
    },
    {
      "epoch": 4.538650450239738,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5285,
      "step": 19403
    },
    {
      "epoch": 4.538884341012747,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 19404
    },
    {
      "epoch": 4.539118231785756,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 19405
    },
    {
      "epoch": 4.539352122558765,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 19406
    },
    {
      "epoch": 4.539586013331774,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6349,
      "step": 19407
    },
    {
      "epoch": 4.5398199041047835,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4484,
      "step": 19408
    },
    {
      "epoch": 4.540053794877792,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 19409
    },
    {
      "epoch": 4.540287685650801,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 19410
    },
    {
      "epoch": 4.54052157642381,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 19411
    },
    {
      "epoch": 4.540755467196819,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 19412
    },
    {
      "epoch": 4.540989357969828,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5395,
      "step": 19413
    },
    {
      "epoch": 4.5412232487428374,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7532,
      "step": 19414
    },
    {
      "epoch": 4.541457139515846,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 19415
    },
    {
      "epoch": 4.541691030288855,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 19416
    },
    {
      "epoch": 4.541924921061864,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6159,
      "step": 19417
    },
    {
      "epoch": 4.542158811834873,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 19418
    },
    {
      "epoch": 4.5423927026078825,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6281,
      "step": 19419
    },
    {
      "epoch": 4.542626593380891,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 19420
    },
    {
      "epoch": 4.5428604841539,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 19421
    },
    {
      "epoch": 4.543094374926909,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.45,
      "step": 19422
    },
    {
      "epoch": 4.543328265699918,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 19423
    },
    {
      "epoch": 4.543562156472927,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 19424
    },
    {
      "epoch": 4.543796047245936,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 19425
    },
    {
      "epoch": 4.544029938018945,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 19426
    },
    {
      "epoch": 4.544263828791954,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 19427
    },
    {
      "epoch": 4.544497719564963,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 19428
    },
    {
      "epoch": 4.544731610337972,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2155,
      "step": 19429
    },
    {
      "epoch": 4.5449655011109815,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 19430
    },
    {
      "epoch": 4.54519939188399,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 19431
    },
    {
      "epoch": 4.545433282656999,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 19432
    },
    {
      "epoch": 4.545667173430008,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 19433
    },
    {
      "epoch": 4.545901064203017,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 19434
    },
    {
      "epoch": 4.546134954976026,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 19435
    },
    {
      "epoch": 4.546368845749035,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.2991,
      "step": 19436
    },
    {
      "epoch": 4.546602736522044,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 19437
    },
    {
      "epoch": 4.546836627295053,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 19438
    },
    {
      "epoch": 4.547070518068062,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 19439
    },
    {
      "epoch": 4.547304408841072,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6023,
      "step": 19440
    },
    {
      "epoch": 4.5475382996140805,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 19441
    },
    {
      "epoch": 4.547772190387089,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 19442
    },
    {
      "epoch": 4.548006081160098,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 19443
    },
    {
      "epoch": 4.548239971933107,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 19444
    },
    {
      "epoch": 4.548473862706116,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 19445
    },
    {
      "epoch": 4.548707753479126,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 19446
    },
    {
      "epoch": 4.548941644252134,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 19447
    },
    {
      "epoch": 4.549175535025143,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 19448
    },
    {
      "epoch": 4.549409425798152,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5995,
      "step": 19449
    },
    {
      "epoch": 4.549643316571161,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 19450
    },
    {
      "epoch": 4.549877207344171,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 19451
    },
    {
      "epoch": 4.5501110981171795,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 19452
    },
    {
      "epoch": 4.550344988890188,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7589,
      "step": 19453
    },
    {
      "epoch": 4.550578879663197,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 19454
    },
    {
      "epoch": 4.550812770436206,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 19455
    },
    {
      "epoch": 4.551046661209215,
      "grad_norm": 2.484375,
      "learning_rate": 3e-05,
      "loss": 1.3932,
      "step": 19456
    },
    {
      "epoch": 4.551280551982225,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 19457
    },
    {
      "epoch": 4.551514442755233,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 19458
    },
    {
      "epoch": 4.551748333528242,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 19459
    },
    {
      "epoch": 4.551982224301251,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 19460
    },
    {
      "epoch": 4.55221611507426,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 19461
    },
    {
      "epoch": 4.55245000584727,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 19462
    },
    {
      "epoch": 4.5526838966202785,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 19463
    },
    {
      "epoch": 4.552917787393287,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 19464
    },
    {
      "epoch": 4.553151678166296,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 19465
    },
    {
      "epoch": 4.553385568939305,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 19466
    },
    {
      "epoch": 4.553619459712314,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 19467
    },
    {
      "epoch": 4.553853350485324,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 19468
    },
    {
      "epoch": 4.554087241258332,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 19469
    },
    {
      "epoch": 4.554321132031341,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 19470
    },
    {
      "epoch": 4.55455502280435,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 19471
    },
    {
      "epoch": 4.55478891357736,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 19472
    },
    {
      "epoch": 4.555022804350369,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 19473
    },
    {
      "epoch": 4.5552566951233775,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 19474
    },
    {
      "epoch": 4.555490585896386,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 19475
    },
    {
      "epoch": 4.555724476669395,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 19476
    },
    {
      "epoch": 4.555958367442404,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4977,
      "step": 19477
    },
    {
      "epoch": 4.556192258215414,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 19478
    },
    {
      "epoch": 4.556426148988423,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 19479
    },
    {
      "epoch": 4.556660039761431,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 19480
    },
    {
      "epoch": 4.55689393053444,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 19481
    },
    {
      "epoch": 4.557127821307449,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 19482
    },
    {
      "epoch": 4.557361712080459,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 19483
    },
    {
      "epoch": 4.557595602853468,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 19484
    },
    {
      "epoch": 4.5578294936264765,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 19485
    },
    {
      "epoch": 4.558063384399485,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 19486
    },
    {
      "epoch": 4.558297275172494,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 19487
    },
    {
      "epoch": 4.558531165945503,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 19488
    },
    {
      "epoch": 4.558765056718513,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 19489
    },
    {
      "epoch": 4.558998947491522,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 19490
    },
    {
      "epoch": 4.55923283826453,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 19491
    },
    {
      "epoch": 4.559466729037539,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5525,
      "step": 19492
    },
    {
      "epoch": 4.559700619810549,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5618,
      "step": 19493
    },
    {
      "epoch": 4.559934510583558,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 19494
    },
    {
      "epoch": 4.560168401356567,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 19495
    },
    {
      "epoch": 4.5604022921295755,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 19496
    },
    {
      "epoch": 4.560636182902584,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2097,
      "step": 19497
    },
    {
      "epoch": 4.560870073675593,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.15,
      "step": 19498
    },
    {
      "epoch": 4.561103964448603,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 19499
    },
    {
      "epoch": 4.561337855221612,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.3586,
      "step": 19500
    },
    {
      "epoch": 4.561337855221612,
      "eval_runtime": 4.6979,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 19500
    },
    {
      "epoch": 4.561571745994621,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 19501
    },
    {
      "epoch": 4.561805636767629,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 19502
    },
    {
      "epoch": 4.562039527540638,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 19503
    },
    {
      "epoch": 4.562273418313648,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6748,
      "step": 19504
    },
    {
      "epoch": 4.562507309086657,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 19505
    },
    {
      "epoch": 4.562741199859666,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 19506
    },
    {
      "epoch": 4.5629750906326745,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.5581,
      "step": 19507
    },
    {
      "epoch": 4.563208981405683,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7166,
      "step": 19508
    },
    {
      "epoch": 4.563442872178692,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 19509
    },
    {
      "epoch": 4.563676762951702,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0348,
      "step": 19510
    },
    {
      "epoch": 4.563910653724711,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 19511
    },
    {
      "epoch": 4.56414454449772,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 19512
    },
    {
      "epoch": 4.564378435270728,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 19513
    },
    {
      "epoch": 4.564612326043737,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 19514
    },
    {
      "epoch": 4.564846216816747,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7175,
      "step": 19515
    },
    {
      "epoch": 4.565080107589756,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 19516
    },
    {
      "epoch": 4.565313998362765,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 19517
    },
    {
      "epoch": 4.5655478891357735,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6398,
      "step": 19518
    },
    {
      "epoch": 4.565781779908782,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 19519
    },
    {
      "epoch": 4.566015670681791,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7501,
      "step": 19520
    },
    {
      "epoch": 4.566249561454801,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 19521
    },
    {
      "epoch": 4.56648345222781,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 19522
    },
    {
      "epoch": 4.5667173430008186,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5551,
      "step": 19523
    },
    {
      "epoch": 4.566951233773827,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 19524
    },
    {
      "epoch": 4.567185124546837,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.4796,
      "step": 19525
    },
    {
      "epoch": 4.567419015319846,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 19526
    },
    {
      "epoch": 4.567652906092855,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 19527
    },
    {
      "epoch": 4.567886796865864,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 19528
    },
    {
      "epoch": 4.5681206876388725,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.3423,
      "step": 19529
    },
    {
      "epoch": 4.568354578411881,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2344,
      "step": 19530
    },
    {
      "epoch": 4.568588469184891,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4897,
      "step": 19531
    },
    {
      "epoch": 4.5688223599579,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 19532
    },
    {
      "epoch": 4.569056250730909,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 19533
    },
    {
      "epoch": 4.5692901415039175,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7206,
      "step": 19534
    },
    {
      "epoch": 4.569524032276926,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8047,
      "step": 19535
    },
    {
      "epoch": 4.569757923049936,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 19536
    },
    {
      "epoch": 4.569991813822945,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 19537
    },
    {
      "epoch": 4.570225704595954,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 19538
    },
    {
      "epoch": 4.570459595368963,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 19539
    },
    {
      "epoch": 4.5706934861419715,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 19540
    },
    {
      "epoch": 4.57092737691498,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 19541
    },
    {
      "epoch": 4.57116126768799,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 19542
    },
    {
      "epoch": 4.571395158460999,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 19543
    },
    {
      "epoch": 4.571629049234008,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.712,
      "step": 19544
    },
    {
      "epoch": 4.5718629400070165,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6897,
      "step": 19545
    },
    {
      "epoch": 4.572096830780025,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 19546
    },
    {
      "epoch": 4.572330721553035,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 19547
    },
    {
      "epoch": 4.572564612326044,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 19548
    },
    {
      "epoch": 4.572798503099053,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 19549
    },
    {
      "epoch": 4.573032393872062,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 19550
    },
    {
      "epoch": 4.5732662846450705,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 19551
    },
    {
      "epoch": 4.573500175418079,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 19552
    },
    {
      "epoch": 4.573734066191089,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 19553
    },
    {
      "epoch": 4.573967956964098,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 19554
    },
    {
      "epoch": 4.574201847737107,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 19555
    },
    {
      "epoch": 4.5744357385101155,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 19556
    },
    {
      "epoch": 4.574669629283125,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 19557
    },
    {
      "epoch": 4.574903520056134,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 19558
    },
    {
      "epoch": 4.575137410829143,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 19559
    },
    {
      "epoch": 4.575371301602152,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7282,
      "step": 19560
    },
    {
      "epoch": 4.575605192375161,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 19561
    },
    {
      "epoch": 4.5758390831481695,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.7107,
      "step": 19562
    },
    {
      "epoch": 4.576072973921179,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 19563
    },
    {
      "epoch": 4.576306864694188,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6697,
      "step": 19564
    },
    {
      "epoch": 4.576540755467197,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 19565
    },
    {
      "epoch": 4.576774646240206,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 19566
    },
    {
      "epoch": 4.5770085370132145,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 19567
    },
    {
      "epoch": 4.577242427786224,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 19568
    },
    {
      "epoch": 4.577476318559233,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 19569
    },
    {
      "epoch": 4.577710209332242,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 19570
    },
    {
      "epoch": 4.577944100105251,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 19571
    },
    {
      "epoch": 4.57817799087826,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 19572
    },
    {
      "epoch": 4.5784118816512684,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7559,
      "step": 19573
    },
    {
      "epoch": 4.578645772424278,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 19574
    },
    {
      "epoch": 4.578879663197287,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5768,
      "step": 19575
    },
    {
      "epoch": 4.579113553970296,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 19576
    },
    {
      "epoch": 4.579347444743305,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6022,
      "step": 19577
    },
    {
      "epoch": 4.5795813355163135,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 19578
    },
    {
      "epoch": 4.579815226289323,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 19579
    },
    {
      "epoch": 4.580049117062332,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 19580
    },
    {
      "epoch": 4.580283007835341,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 19581
    },
    {
      "epoch": 4.58051689860835,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 19582
    },
    {
      "epoch": 4.580750789381359,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 19583
    },
    {
      "epoch": 4.580984680154367,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9937,
      "step": 19584
    },
    {
      "epoch": 4.581218570927377,
      "grad_norm": 9.75,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 19585
    },
    {
      "epoch": 4.581452461700386,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 19586
    },
    {
      "epoch": 4.581686352473395,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 19587
    },
    {
      "epoch": 4.581920243246404,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 19588
    },
    {
      "epoch": 4.582154134019413,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 19589
    },
    {
      "epoch": 4.582388024792422,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 19590
    },
    {
      "epoch": 4.582621915565431,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 19591
    },
    {
      "epoch": 4.58285580633844,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6016,
      "step": 19592
    },
    {
      "epoch": 4.583089697111449,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 19593
    },
    {
      "epoch": 4.583323587884458,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5294,
      "step": 19594
    },
    {
      "epoch": 4.583557478657467,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 19595
    },
    {
      "epoch": 4.583791369430476,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 19596
    },
    {
      "epoch": 4.584025260203485,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 19597
    },
    {
      "epoch": 4.584259150976494,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.8205,
      "step": 19598
    },
    {
      "epoch": 4.584493041749503,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.886,
      "step": 19599
    },
    {
      "epoch": 4.584726932522512,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 19600
    },
    {
      "epoch": 4.584726932522512,
      "eval_runtime": 4.6213,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 19600
    },
    {
      "epoch": 4.584960823295521,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6759,
      "step": 19601
    },
    {
      "epoch": 4.58519471406853,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2595,
      "step": 19602
    },
    {
      "epoch": 4.585428604841539,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.4452,
      "step": 19603
    },
    {
      "epoch": 4.585662495614548,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 19604
    },
    {
      "epoch": 4.585896386387557,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.54,
      "step": 19605
    },
    {
      "epoch": 4.586130277160566,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 19606
    },
    {
      "epoch": 4.586364167933575,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 19607
    },
    {
      "epoch": 4.586598058706584,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 19608
    },
    {
      "epoch": 4.586831949479593,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 19609
    },
    {
      "epoch": 4.587065840252602,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 19610
    },
    {
      "epoch": 4.587299731025611,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 19611
    },
    {
      "epoch": 4.58753362179862,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 19612
    },
    {
      "epoch": 4.587767512571629,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 19613
    },
    {
      "epoch": 4.588001403344638,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.578,
      "step": 19614
    },
    {
      "epoch": 4.588235294117647,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 19615
    },
    {
      "epoch": 4.588469184890656,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8034,
      "step": 19616
    },
    {
      "epoch": 4.588703075663665,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 19617
    },
    {
      "epoch": 4.588936966436674,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.681,
      "step": 19618
    },
    {
      "epoch": 4.589170857209683,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 19619
    },
    {
      "epoch": 4.589404747982692,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 19620
    },
    {
      "epoch": 4.5896386387557015,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 19621
    },
    {
      "epoch": 4.58987252952871,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 19622
    },
    {
      "epoch": 4.590106420301719,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 19623
    },
    {
      "epoch": 4.590340311074728,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 19624
    },
    {
      "epoch": 4.590574201847737,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.584,
      "step": 19625
    },
    {
      "epoch": 4.590808092620746,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 19626
    },
    {
      "epoch": 4.5910419833937555,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 19627
    },
    {
      "epoch": 4.591275874166764,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 19628
    },
    {
      "epoch": 4.591509764939773,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 19629
    },
    {
      "epoch": 4.591743655712782,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 19630
    },
    {
      "epoch": 4.591977546485791,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 19631
    },
    {
      "epoch": 4.5922114372588005,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6456,
      "step": 19632
    },
    {
      "epoch": 4.592445328031809,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 19633
    },
    {
      "epoch": 4.592679218804818,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 19634
    },
    {
      "epoch": 4.592913109577827,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 19635
    },
    {
      "epoch": 4.593147000350836,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 19636
    },
    {
      "epoch": 4.593380891123845,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 19637
    },
    {
      "epoch": 4.5936147818968545,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 19638
    },
    {
      "epoch": 4.593848672669863,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 19639
    },
    {
      "epoch": 4.594082563442872,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 19640
    },
    {
      "epoch": 4.594316454215881,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7075,
      "step": 19641
    },
    {
      "epoch": 4.59455034498889,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 19642
    },
    {
      "epoch": 4.5947842357618995,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 19643
    },
    {
      "epoch": 4.595018126534908,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6206,
      "step": 19644
    },
    {
      "epoch": 4.595252017307917,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 19645
    },
    {
      "epoch": 4.595485908080926,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 19646
    },
    {
      "epoch": 4.595719798853935,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 19647
    },
    {
      "epoch": 4.595953689626944,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7632,
      "step": 19648
    },
    {
      "epoch": 4.5961875803999535,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 19649
    },
    {
      "epoch": 4.596421471172962,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5111,
      "step": 19650
    },
    {
      "epoch": 4.596655361945971,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 19651
    },
    {
      "epoch": 4.59688925271898,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 19652
    },
    {
      "epoch": 4.59712314349199,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 19653
    },
    {
      "epoch": 4.5973570342649985,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 19654
    },
    {
      "epoch": 4.597590925038007,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8252,
      "step": 19655
    },
    {
      "epoch": 4.597824815811016,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 19656
    },
    {
      "epoch": 4.598058706584025,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 19657
    },
    {
      "epoch": 4.598292597357034,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.5563,
      "step": 19658
    },
    {
      "epoch": 4.598526488130044,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5936,
      "step": 19659
    },
    {
      "epoch": 4.5987603789030524,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 19660
    },
    {
      "epoch": 4.598994269676061,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 19661
    },
    {
      "epoch": 4.59922816044907,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 19662
    },
    {
      "epoch": 4.599462051222079,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 19663
    },
    {
      "epoch": 4.599695941995089,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 19664
    },
    {
      "epoch": 4.5999298327680975,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 19665
    },
    {
      "epoch": 4.600163723541106,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 19666
    },
    {
      "epoch": 4.600397614314115,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 19667
    },
    {
      "epoch": 4.600631505087124,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 19668
    },
    {
      "epoch": 4.600865395860133,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.613,
      "step": 19669
    },
    {
      "epoch": 4.601099286633143,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 19670
    },
    {
      "epoch": 4.601333177406151,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 19671
    },
    {
      "epoch": 4.60156706817916,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 19672
    },
    {
      "epoch": 4.601800958952169,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 19673
    },
    {
      "epoch": 4.602034849725178,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5504,
      "step": 19674
    },
    {
      "epoch": 4.602268740498188,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 19675
    },
    {
      "epoch": 4.6025026312711965,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7021,
      "step": 19676
    },
    {
      "epoch": 4.602736522044205,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 19677
    },
    {
      "epoch": 4.602970412817214,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6384,
      "step": 19678
    },
    {
      "epoch": 4.603204303590223,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 19679
    },
    {
      "epoch": 4.603438194363232,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 19680
    },
    {
      "epoch": 4.603672085136242,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 19681
    },
    {
      "epoch": 4.60390597590925,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 19682
    },
    {
      "epoch": 4.604139866682259,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.69,
      "step": 19683
    },
    {
      "epoch": 4.604373757455268,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 19684
    },
    {
      "epoch": 4.604607648228278,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 19685
    },
    {
      "epoch": 4.604841539001287,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 19686
    },
    {
      "epoch": 4.6050754297742955,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 19687
    },
    {
      "epoch": 4.605309320547304,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 19688
    },
    {
      "epoch": 4.605543211320313,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4963,
      "step": 19689
    },
    {
      "epoch": 4.605777102093322,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6574,
      "step": 19690
    },
    {
      "epoch": 4.606010992866332,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6516,
      "step": 19691
    },
    {
      "epoch": 4.606244883639341,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 19692
    },
    {
      "epoch": 4.606478774412349,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 19693
    },
    {
      "epoch": 4.606712665185358,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 19694
    },
    {
      "epoch": 4.606946555958367,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 19695
    },
    {
      "epoch": 4.607180446731377,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 19696
    },
    {
      "epoch": 4.607414337504386,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7731,
      "step": 19697
    },
    {
      "epoch": 4.6076482282773945,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 19698
    },
    {
      "epoch": 4.607882119050403,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.393,
      "step": 19699
    },
    {
      "epoch": 4.608116009823412,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6449,
      "step": 19700
    },
    {
      "epoch": 4.608116009823412,
      "eval_runtime": 4.5956,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 19700
    },
    {
      "epoch": 4.608349900596421,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 19701
    },
    {
      "epoch": 4.608583791369431,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 19702
    },
    {
      "epoch": 4.60881768214244,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7615,
      "step": 19703
    },
    {
      "epoch": 4.609051572915448,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7229,
      "step": 19704
    },
    {
      "epoch": 4.609285463688457,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 19705
    },
    {
      "epoch": 4.609519354461467,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6562,
      "step": 19706
    },
    {
      "epoch": 4.609753245234476,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 19707
    },
    {
      "epoch": 4.609987136007485,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 19708
    },
    {
      "epoch": 4.6102210267804935,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 19709
    },
    {
      "epoch": 4.610454917553502,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 19710
    },
    {
      "epoch": 4.610688808326511,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 19711
    },
    {
      "epoch": 4.610922699099521,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8576,
      "step": 19712
    },
    {
      "epoch": 4.61115658987253,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6417,
      "step": 19713
    },
    {
      "epoch": 4.611390480645539,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 19714
    },
    {
      "epoch": 4.611624371418547,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 19715
    },
    {
      "epoch": 4.611858262191556,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 19716
    },
    {
      "epoch": 4.612092152964566,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 19717
    },
    {
      "epoch": 4.612326043737575,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6857,
      "step": 19718
    },
    {
      "epoch": 4.612559934510584,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 19719
    },
    {
      "epoch": 4.6127938252835925,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 19720
    },
    {
      "epoch": 4.613027716056601,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 19721
    },
    {
      "epoch": 4.61326160682961,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 19722
    },
    {
      "epoch": 4.61349549760262,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7101,
      "step": 19723
    },
    {
      "epoch": 4.613729388375629,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5894,
      "step": 19724
    },
    {
      "epoch": 4.613963279148638,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8026,
      "step": 19725
    },
    {
      "epoch": 4.614197169921646,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 19726
    },
    {
      "epoch": 4.614431060694655,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 19727
    },
    {
      "epoch": 4.614664951467665,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 19728
    },
    {
      "epoch": 4.614898842240674,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 19729
    },
    {
      "epoch": 4.615132733013683,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 19730
    },
    {
      "epoch": 4.6153666237866915,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 19731
    },
    {
      "epoch": 4.6156005145597,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 19732
    },
    {
      "epoch": 4.615834405332709,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 19733
    },
    {
      "epoch": 4.616068296105719,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 19734
    },
    {
      "epoch": 4.616302186878728,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 19735
    },
    {
      "epoch": 4.616536077651737,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 19736
    },
    {
      "epoch": 4.616769968424745,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 19737
    },
    {
      "epoch": 4.617003859197755,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 19738
    },
    {
      "epoch": 4.617237749970764,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 19739
    },
    {
      "epoch": 4.617471640743773,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 19740
    },
    {
      "epoch": 4.617705531516782,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 19741
    },
    {
      "epoch": 4.6179394222897905,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 19742
    },
    {
      "epoch": 4.618173313062799,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 19743
    },
    {
      "epoch": 4.618407203835809,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 19744
    },
    {
      "epoch": 4.618641094608818,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 19745
    },
    {
      "epoch": 4.618874985381827,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 19746
    },
    {
      "epoch": 4.619108876154836,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6834,
      "step": 19747
    },
    {
      "epoch": 4.619342766927844,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 19748
    },
    {
      "epoch": 4.619576657700854,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 19749
    },
    {
      "epoch": 4.619810548473863,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7792,
      "step": 19750
    },
    {
      "epoch": 4.620044439246872,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8553,
      "step": 19751
    },
    {
      "epoch": 4.620278330019881,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 19752
    },
    {
      "epoch": 4.6205122207928895,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 19753
    },
    {
      "epoch": 4.620746111565898,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 19754
    },
    {
      "epoch": 4.620980002338908,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5254,
      "step": 19755
    },
    {
      "epoch": 4.621213893111917,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 19756
    },
    {
      "epoch": 4.621447783884926,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 19757
    },
    {
      "epoch": 4.621681674657935,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 19758
    },
    {
      "epoch": 4.621915565430943,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 19759
    },
    {
      "epoch": 4.622149456203953,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8634,
      "step": 19760
    },
    {
      "epoch": 4.622383346976962,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 19761
    },
    {
      "epoch": 4.622617237749971,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 19762
    },
    {
      "epoch": 4.62285112852298,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7528,
      "step": 19763
    },
    {
      "epoch": 4.6230850192959885,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9811,
      "step": 19764
    },
    {
      "epoch": 4.623318910068997,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6867,
      "step": 19765
    },
    {
      "epoch": 4.623552800842007,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 19766
    },
    {
      "epoch": 4.623786691615016,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7413,
      "step": 19767
    },
    {
      "epoch": 4.624020582388025,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.533,
      "step": 19768
    },
    {
      "epoch": 4.6242544731610336,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.675,
      "step": 19769
    },
    {
      "epoch": 4.624488363934043,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 19770
    },
    {
      "epoch": 4.624722254707052,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 19771
    },
    {
      "epoch": 4.624956145480061,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 19772
    },
    {
      "epoch": 4.62519003625307,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 19773
    },
    {
      "epoch": 4.625423927026079,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 19774
    },
    {
      "epoch": 4.6256578177990875,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6738,
      "step": 19775
    },
    {
      "epoch": 4.625891708572097,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 19776
    },
    {
      "epoch": 4.626125599345106,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 19777
    },
    {
      "epoch": 4.626359490118115,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5686,
      "step": 19778
    },
    {
      "epoch": 4.626593380891124,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 19779
    },
    {
      "epoch": 4.6268272716641325,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 19780
    },
    {
      "epoch": 4.627061162437142,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 19781
    },
    {
      "epoch": 4.627295053210151,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.76,
      "step": 19782
    },
    {
      "epoch": 4.62752894398316,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 19783
    },
    {
      "epoch": 4.627762834756169,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 19784
    },
    {
      "epoch": 4.627996725529178,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 19785
    },
    {
      "epoch": 4.6282306163021865,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 19786
    },
    {
      "epoch": 4.628464507075196,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.617,
      "step": 19787
    },
    {
      "epoch": 4.628698397848205,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7186,
      "step": 19788
    },
    {
      "epoch": 4.628932288621214,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 19789
    },
    {
      "epoch": 4.629166179394223,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6694,
      "step": 19790
    },
    {
      "epoch": 4.6294000701672315,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 19791
    },
    {
      "epoch": 4.629633960940241,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 19792
    },
    {
      "epoch": 4.62986785171325,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 19793
    },
    {
      "epoch": 4.630101742486259,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5968,
      "step": 19794
    },
    {
      "epoch": 4.630335633259268,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 19795
    },
    {
      "epoch": 4.630569524032277,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 19796
    },
    {
      "epoch": 4.6308034148052855,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 19797
    },
    {
      "epoch": 4.631037305578295,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5545,
      "step": 19798
    },
    {
      "epoch": 4.631271196351304,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 19799
    },
    {
      "epoch": 4.631505087124313,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 19800
    },
    {
      "epoch": 4.631505087124313,
      "eval_runtime": 4.5965,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 19800
    },
    {
      "epoch": 4.631738977897322,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6586,
      "step": 19801
    },
    {
      "epoch": 4.631972868670331,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 19802
    },
    {
      "epoch": 4.63220675944334,
      "grad_norm": 14.5625,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 19803
    },
    {
      "epoch": 4.632440650216349,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6853,
      "step": 19804
    },
    {
      "epoch": 4.632674540989358,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 19805
    },
    {
      "epoch": 4.632908431762367,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7362,
      "step": 19806
    },
    {
      "epoch": 4.633142322535376,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 19807
    },
    {
      "epoch": 4.633376213308385,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5879,
      "step": 19808
    },
    {
      "epoch": 4.633610104081394,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 19809
    },
    {
      "epoch": 4.633843994854403,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4699,
      "step": 19810
    },
    {
      "epoch": 4.634077885627412,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 19811
    },
    {
      "epoch": 4.634311776400421,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 19812
    },
    {
      "epoch": 4.63454566717343,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6943,
      "step": 19813
    },
    {
      "epoch": 4.634779557946439,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9209,
      "step": 19814
    },
    {
      "epoch": 4.635013448719448,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.793,
      "step": 19815
    },
    {
      "epoch": 4.635247339492457,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 19816
    },
    {
      "epoch": 4.635481230265466,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5998,
      "step": 19817
    },
    {
      "epoch": 4.635715121038475,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4136,
      "step": 19818
    },
    {
      "epoch": 4.635949011811484,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 19819
    },
    {
      "epoch": 4.636182902584493,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 19820
    },
    {
      "epoch": 4.636416793357502,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 19821
    },
    {
      "epoch": 4.636650684130511,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 19822
    },
    {
      "epoch": 4.63688457490352,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 19823
    },
    {
      "epoch": 4.637118465676529,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 19824
    },
    {
      "epoch": 4.637352356449538,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 19825
    },
    {
      "epoch": 4.637586247222547,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6925,
      "step": 19826
    },
    {
      "epoch": 4.637820137995556,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 19827
    },
    {
      "epoch": 4.638054028768565,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 19828
    },
    {
      "epoch": 4.638287919541574,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 19829
    },
    {
      "epoch": 4.638521810314583,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5876,
      "step": 19830
    },
    {
      "epoch": 4.638755701087592,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 19831
    },
    {
      "epoch": 4.638989591860601,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.64,
      "step": 19832
    },
    {
      "epoch": 4.63922348263361,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 19833
    },
    {
      "epoch": 4.63945737340662,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 19834
    },
    {
      "epoch": 4.639691264179628,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.4423,
      "step": 19835
    },
    {
      "epoch": 4.639925154952637,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8674,
      "step": 19836
    },
    {
      "epoch": 4.640159045725646,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 19837
    },
    {
      "epoch": 4.640392936498655,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3082,
      "step": 19838
    },
    {
      "epoch": 4.640626827271664,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 19839
    },
    {
      "epoch": 4.6408607180446735,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 19840
    },
    {
      "epoch": 4.641094608817682,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 19841
    },
    {
      "epoch": 4.641328499590691,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 19842
    },
    {
      "epoch": 4.6415623903637,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 19843
    },
    {
      "epoch": 4.641796281136709,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.5965,
      "step": 19844
    },
    {
      "epoch": 4.642030171909719,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 19845
    },
    {
      "epoch": 4.642264062682727,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 19846
    },
    {
      "epoch": 4.642497953455736,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 19847
    },
    {
      "epoch": 4.642731844228745,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 19848
    },
    {
      "epoch": 4.642965735001754,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 19849
    },
    {
      "epoch": 4.643199625774763,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 19850
    },
    {
      "epoch": 4.6434335165477725,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6576,
      "step": 19851
    },
    {
      "epoch": 4.643667407320781,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 19852
    },
    {
      "epoch": 4.64390129809379,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 19853
    },
    {
      "epoch": 4.644135188866799,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 19854
    },
    {
      "epoch": 4.644369079639808,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 19855
    },
    {
      "epoch": 4.6446029704128176,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 19856
    },
    {
      "epoch": 4.644836861185826,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 19857
    },
    {
      "epoch": 4.645070751958835,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7532,
      "step": 19858
    },
    {
      "epoch": 4.645304642731844,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 19859
    },
    {
      "epoch": 4.645538533504853,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6311,
      "step": 19860
    },
    {
      "epoch": 4.645772424277862,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 19861
    },
    {
      "epoch": 4.6460063150508715,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 19862
    },
    {
      "epoch": 4.64624020582388,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 19863
    },
    {
      "epoch": 4.646474096596889,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 19864
    },
    {
      "epoch": 4.646707987369898,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7265,
      "step": 19865
    },
    {
      "epoch": 4.646941878142908,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7203,
      "step": 19866
    },
    {
      "epoch": 4.6471757689159165,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 19867
    },
    {
      "epoch": 4.647409659688925,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 19868
    },
    {
      "epoch": 4.647643550461934,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 19869
    },
    {
      "epoch": 4.647877441234943,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 19870
    },
    {
      "epoch": 4.648111332007952,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 19871
    },
    {
      "epoch": 4.648345222780962,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 19872
    },
    {
      "epoch": 4.6485791135539705,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2852,
      "step": 19873
    },
    {
      "epoch": 4.648813004326979,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 19874
    },
    {
      "epoch": 4.649046895099988,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 19875
    },
    {
      "epoch": 4.649280785872997,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 19876
    },
    {
      "epoch": 4.649514676646007,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 19877
    },
    {
      "epoch": 4.6497485674190155,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 19878
    },
    {
      "epoch": 4.649982458192024,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 19879
    },
    {
      "epoch": 4.650216348965033,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 19880
    },
    {
      "epoch": 4.650450239738042,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7772,
      "step": 19881
    },
    {
      "epoch": 4.650684130511051,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 19882
    },
    {
      "epoch": 4.650918021284061,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 19883
    },
    {
      "epoch": 4.6511519120570695,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 19884
    },
    {
      "epoch": 4.651385802830078,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 19885
    },
    {
      "epoch": 4.651619693603087,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 19886
    },
    {
      "epoch": 4.651853584376097,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7215,
      "step": 19887
    },
    {
      "epoch": 4.652087475149106,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 19888
    },
    {
      "epoch": 4.6523213659221145,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 19889
    },
    {
      "epoch": 4.652555256695123,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 19890
    },
    {
      "epoch": 4.652789147468132,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8528,
      "step": 19891
    },
    {
      "epoch": 4.653023038241141,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 19892
    },
    {
      "epoch": 4.653256929014151,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 19893
    },
    {
      "epoch": 4.65349081978716,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 19894
    },
    {
      "epoch": 4.6537247105601685,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 19895
    },
    {
      "epoch": 4.653958601333177,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6124,
      "step": 19896
    },
    {
      "epoch": 4.654192492106186,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 19897
    },
    {
      "epoch": 4.654426382879196,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1772,
      "step": 19898
    },
    {
      "epoch": 4.654660273652205,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 19899
    },
    {
      "epoch": 4.6548941644252135,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 19900
    },
    {
      "epoch": 4.6548941644252135,
      "eval_runtime": 4.6279,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 19900
    },
    {
      "epoch": 4.655128055198222,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7916,
      "step": 19901
    },
    {
      "epoch": 4.655361945971231,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7208,
      "step": 19902
    },
    {
      "epoch": 4.65559583674424,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 19903
    },
    {
      "epoch": 4.65582972751725,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 19904
    },
    {
      "epoch": 4.656063618290259,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 19905
    },
    {
      "epoch": 4.6562975090632674,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 19906
    },
    {
      "epoch": 4.656531399836276,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 19907
    },
    {
      "epoch": 4.656765290609285,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6912,
      "step": 19908
    },
    {
      "epoch": 4.656999181382295,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6668,
      "step": 19909
    },
    {
      "epoch": 4.657233072155304,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 19910
    },
    {
      "epoch": 4.6574669629283125,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 19911
    },
    {
      "epoch": 4.657700853701321,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 19912
    },
    {
      "epoch": 4.65793474447433,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6927,
      "step": 19913
    },
    {
      "epoch": 4.658168635247339,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6562,
      "step": 19914
    },
    {
      "epoch": 4.658402526020349,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 19915
    },
    {
      "epoch": 4.658636416793358,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9001,
      "step": 19916
    },
    {
      "epoch": 4.658870307566366,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 19917
    },
    {
      "epoch": 4.659104198339375,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9221,
      "step": 19918
    },
    {
      "epoch": 4.659338089112385,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 19919
    },
    {
      "epoch": 4.659571979885394,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 19920
    },
    {
      "epoch": 4.659805870658403,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 19921
    },
    {
      "epoch": 4.6600397614314115,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 19922
    },
    {
      "epoch": 4.66027365220442,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 19923
    },
    {
      "epoch": 4.660507542977429,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 19924
    },
    {
      "epoch": 4.660741433750439,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 19925
    },
    {
      "epoch": 4.660975324523448,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 19926
    },
    {
      "epoch": 4.661209215296457,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8108,
      "step": 19927
    },
    {
      "epoch": 4.661443106069465,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 19928
    },
    {
      "epoch": 4.661676996842474,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7349,
      "step": 19929
    },
    {
      "epoch": 4.661910887615484,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 19930
    },
    {
      "epoch": 4.662144778388493,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 19931
    },
    {
      "epoch": 4.662378669161502,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 19932
    },
    {
      "epoch": 4.6626125599345105,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 19933
    },
    {
      "epoch": 4.662846450707519,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 19934
    },
    {
      "epoch": 4.663080341480528,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 19935
    },
    {
      "epoch": 4.663314232253538,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 19936
    },
    {
      "epoch": 4.663548123026547,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 19937
    },
    {
      "epoch": 4.663782013799556,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5852,
      "step": 19938
    },
    {
      "epoch": 4.664015904572564,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 19939
    },
    {
      "epoch": 4.664249795345573,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9061,
      "step": 19940
    },
    {
      "epoch": 4.664483686118583,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0112,
      "step": 19941
    },
    {
      "epoch": 4.664717576891592,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 19942
    },
    {
      "epoch": 4.664951467664601,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 19943
    },
    {
      "epoch": 4.6651853584376095,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 19944
    },
    {
      "epoch": 4.665419249210618,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1327,
      "step": 19945
    },
    {
      "epoch": 4.665653139983627,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6481,
      "step": 19946
    },
    {
      "epoch": 4.665887030756637,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 19947
    },
    {
      "epoch": 4.666120921529646,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 19948
    },
    {
      "epoch": 4.666354812302655,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7592,
      "step": 19949
    },
    {
      "epoch": 4.666588703075663,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7054,
      "step": 19950
    },
    {
      "epoch": 4.666822593848673,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 19951
    },
    {
      "epoch": 4.667056484621682,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 19952
    },
    {
      "epoch": 4.667290375394691,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 19953
    },
    {
      "epoch": 4.6675242661677,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4594,
      "step": 19954
    },
    {
      "epoch": 4.6677581569407085,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 19955
    },
    {
      "epoch": 4.667992047713717,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 19956
    },
    {
      "epoch": 4.668225938486727,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 19957
    },
    {
      "epoch": 4.668459829259736,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8107,
      "step": 19958
    },
    {
      "epoch": 4.668693720032745,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 19959
    },
    {
      "epoch": 4.668927610805754,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 19960
    },
    {
      "epoch": 4.669161501578762,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6909,
      "step": 19961
    },
    {
      "epoch": 4.669395392351772,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.5993,
      "step": 19962
    },
    {
      "epoch": 4.669629283124781,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 19963
    },
    {
      "epoch": 4.66986317389779,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 19964
    },
    {
      "epoch": 4.670097064670799,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 19965
    },
    {
      "epoch": 4.6703309554438075,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 19966
    },
    {
      "epoch": 4.670564846216816,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 19967
    },
    {
      "epoch": 4.670798736989826,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0928,
      "step": 19968
    },
    {
      "epoch": 4.671032627762835,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 19969
    },
    {
      "epoch": 4.671266518535844,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 19970
    },
    {
      "epoch": 4.671500409308853,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 19971
    },
    {
      "epoch": 4.671734300081861,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 19972
    },
    {
      "epoch": 4.671968190854871,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.61,
      "step": 19973
    },
    {
      "epoch": 4.67220208162788,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.3986,
      "step": 19974
    },
    {
      "epoch": 4.672435972400889,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 19975
    },
    {
      "epoch": 4.672669863173898,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 19976
    },
    {
      "epoch": 4.6729037539469065,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7525,
      "step": 19977
    },
    {
      "epoch": 4.673137644719915,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 19978
    },
    {
      "epoch": 4.673371535492925,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5735,
      "step": 19979
    },
    {
      "epoch": 4.673605426265934,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 19980
    },
    {
      "epoch": 4.673839317038943,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 19981
    },
    {
      "epoch": 4.674073207811952,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 19982
    },
    {
      "epoch": 4.674307098584961,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 19983
    },
    {
      "epoch": 4.67454098935797,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1518,
      "step": 19984
    },
    {
      "epoch": 4.674774880130979,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 19985
    },
    {
      "epoch": 4.675008770903988,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 19986
    },
    {
      "epoch": 4.675242661676997,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 19987
    },
    {
      "epoch": 4.6754765524500055,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5625,
      "step": 19988
    },
    {
      "epoch": 4.675710443223015,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 19989
    },
    {
      "epoch": 4.675944333996024,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 19990
    },
    {
      "epoch": 4.676178224769033,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 19991
    },
    {
      "epoch": 4.676412115542042,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8337,
      "step": 19992
    },
    {
      "epoch": 4.676646006315051,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 19993
    },
    {
      "epoch": 4.67687989708806,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 19994
    },
    {
      "epoch": 4.677113787861069,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 19995
    },
    {
      "epoch": 4.677347678634078,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 19996
    },
    {
      "epoch": 4.677581569407087,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8349,
      "step": 19997
    },
    {
      "epoch": 4.677815460180096,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 19998
    },
    {
      "epoch": 4.6780493509531045,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.8601,
      "step": 19999
    },
    {
      "epoch": 4.678283241726114,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 20000
    },
    {
      "epoch": 4.678283241726114,
      "eval_runtime": 4.577,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 20000
    },
    {
      "epoch": 4.678517132499123,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.699,
      "step": 20001
    },
    {
      "epoch": 4.678751023272132,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8001,
      "step": 20002
    },
    {
      "epoch": 4.678984914045141,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 20003
    },
    {
      "epoch": 4.67921880481815,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2678,
      "step": 20004
    },
    {
      "epoch": 4.679452695591159,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 20005
    },
    {
      "epoch": 4.679686586364168,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7221,
      "step": 20006
    },
    {
      "epoch": 4.679920477137177,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 20007
    },
    {
      "epoch": 4.680154367910186,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5648,
      "step": 20008
    },
    {
      "epoch": 4.680388258683195,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 20009
    },
    {
      "epoch": 4.6806221494562035,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 20010
    },
    {
      "epoch": 4.680856040229213,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.505,
      "step": 20011
    },
    {
      "epoch": 4.681089931002222,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 20012
    },
    {
      "epoch": 4.681323821775231,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.7478,
      "step": 20013
    },
    {
      "epoch": 4.68155771254824,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 20014
    },
    {
      "epoch": 4.681791603321249,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8468,
      "step": 20015
    },
    {
      "epoch": 4.682025494094258,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9241,
      "step": 20016
    },
    {
      "epoch": 4.682259384867267,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 20017
    },
    {
      "epoch": 4.682493275640276,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6034,
      "step": 20018
    },
    {
      "epoch": 4.682727166413285,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 20019
    },
    {
      "epoch": 4.682961057186294,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 20020
    },
    {
      "epoch": 4.683194947959303,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 20021
    },
    {
      "epoch": 4.683428838732312,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 20022
    },
    {
      "epoch": 4.683662729505321,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 20023
    },
    {
      "epoch": 4.68389662027833,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 20024
    },
    {
      "epoch": 4.684130511051339,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 20025
    },
    {
      "epoch": 4.684364401824348,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 20026
    },
    {
      "epoch": 4.684598292597357,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 20027
    },
    {
      "epoch": 4.684832183370366,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0601,
      "step": 20028
    },
    {
      "epoch": 4.685066074143375,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7558,
      "step": 20029
    },
    {
      "epoch": 4.685299964916384,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 20030
    },
    {
      "epoch": 4.685533855689393,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 20031
    },
    {
      "epoch": 4.685767746462402,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 20032
    },
    {
      "epoch": 4.686001637235411,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 20033
    },
    {
      "epoch": 4.68623552800842,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 20034
    },
    {
      "epoch": 4.686469418781429,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 20035
    },
    {
      "epoch": 4.686703309554438,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 20036
    },
    {
      "epoch": 4.686937200327447,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 20037
    },
    {
      "epoch": 4.687171091100456,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 20038
    },
    {
      "epoch": 4.687404981873465,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6205,
      "step": 20039
    },
    {
      "epoch": 4.687638872646474,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6476,
      "step": 20040
    },
    {
      "epoch": 4.687872763419483,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 20041
    },
    {
      "epoch": 4.688106654192492,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 20042
    },
    {
      "epoch": 4.688340544965501,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 20043
    },
    {
      "epoch": 4.68857443573851,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 20044
    },
    {
      "epoch": 4.688808326511519,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6855,
      "step": 20045
    },
    {
      "epoch": 4.689042217284528,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 20046
    },
    {
      "epoch": 4.689276108057538,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3266,
      "step": 20047
    },
    {
      "epoch": 4.689509998830546,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 20048
    },
    {
      "epoch": 4.689743889603555,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4812,
      "step": 20049
    },
    {
      "epoch": 4.689977780376564,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 20050
    },
    {
      "epoch": 4.690211671149573,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.6359,
      "step": 20051
    },
    {
      "epoch": 4.690445561922582,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 20052
    },
    {
      "epoch": 4.6906794526955915,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 20053
    },
    {
      "epoch": 4.6909133434686,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 20054
    },
    {
      "epoch": 4.691147234241609,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 20055
    },
    {
      "epoch": 4.691381125014618,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6606,
      "step": 20056
    },
    {
      "epoch": 4.691615015787627,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 20057
    },
    {
      "epoch": 4.691848906560637,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 20058
    },
    {
      "epoch": 4.692082797333645,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 20059
    },
    {
      "epoch": 4.692316688106654,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 20060
    },
    {
      "epoch": 4.692550578879663,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 20061
    },
    {
      "epoch": 4.692784469652672,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 20062
    },
    {
      "epoch": 4.693018360425681,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 20063
    },
    {
      "epoch": 4.6932522511986905,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5946,
      "step": 20064
    },
    {
      "epoch": 4.693486141971699,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 20065
    },
    {
      "epoch": 4.693720032744708,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 20066
    },
    {
      "epoch": 4.693953923517717,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0265,
      "step": 20067
    },
    {
      "epoch": 4.694187814290726,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6812,
      "step": 20068
    },
    {
      "epoch": 4.694421705063736,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 20069
    },
    {
      "epoch": 4.694655595836744,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 20070
    },
    {
      "epoch": 4.694889486609753,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 20071
    },
    {
      "epoch": 4.695123377382762,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8553,
      "step": 20072
    },
    {
      "epoch": 4.695357268155771,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 20073
    },
    {
      "epoch": 4.69559115892878,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 20074
    },
    {
      "epoch": 4.6958250497017895,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6475,
      "step": 20075
    },
    {
      "epoch": 4.696058940474798,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5477,
      "step": 20076
    },
    {
      "epoch": 4.696292831247807,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 20077
    },
    {
      "epoch": 4.696526722020816,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 20078
    },
    {
      "epoch": 4.696760612793826,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 20079
    },
    {
      "epoch": 4.696994503566835,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 20080
    },
    {
      "epoch": 4.697228394339843,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 20081
    },
    {
      "epoch": 4.697462285112852,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7651,
      "step": 20082
    },
    {
      "epoch": 4.697696175885861,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 20083
    },
    {
      "epoch": 4.69793006665887,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 20084
    },
    {
      "epoch": 4.69816395743188,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 20085
    },
    {
      "epoch": 4.6983978482048885,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 20086
    },
    {
      "epoch": 4.698631738977897,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 20087
    },
    {
      "epoch": 4.698865629750906,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 20088
    },
    {
      "epoch": 4.699099520523915,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.2381,
      "step": 20089
    },
    {
      "epoch": 4.699333411296925,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.3896,
      "step": 20090
    },
    {
      "epoch": 4.699567302069934,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 20091
    },
    {
      "epoch": 4.699801192842942,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 20092
    },
    {
      "epoch": 4.700035083615951,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 20093
    },
    {
      "epoch": 4.70026897438896,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 20094
    },
    {
      "epoch": 4.700502865161969,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 20095
    },
    {
      "epoch": 4.700736755934979,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.62,
      "step": 20096
    },
    {
      "epoch": 4.7009706467079875,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 20097
    },
    {
      "epoch": 4.701204537480996,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 20098
    },
    {
      "epoch": 4.701438428254005,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7463,
      "step": 20099
    },
    {
      "epoch": 4.701672319027015,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 20100
    },
    {
      "epoch": 4.701672319027015,
      "eval_runtime": 4.7042,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 20100
    },
    {
      "epoch": 4.701906209800024,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 20101
    },
    {
      "epoch": 4.7021401005730326,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.603,
      "step": 20102
    },
    {
      "epoch": 4.702373991346041,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.697,
      "step": 20103
    },
    {
      "epoch": 4.70260788211905,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 20104
    },
    {
      "epoch": 4.702841772892059,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6713,
      "step": 20105
    },
    {
      "epoch": 4.703075663665069,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 20106
    },
    {
      "epoch": 4.703309554438078,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 20107
    },
    {
      "epoch": 4.7035434452110865,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 20108
    },
    {
      "epoch": 4.703777335984095,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 20109
    },
    {
      "epoch": 4.704011226757104,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 20110
    },
    {
      "epoch": 4.704245117530114,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 20111
    },
    {
      "epoch": 4.704479008303123,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7407,
      "step": 20112
    },
    {
      "epoch": 4.7047128990761315,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6181,
      "step": 20113
    },
    {
      "epoch": 4.70494678984914,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 20114
    },
    {
      "epoch": 4.705180680622149,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 20115
    },
    {
      "epoch": 4.705414571395158,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 20116
    },
    {
      "epoch": 4.705648462168168,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 20117
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 20118
    },
    {
      "epoch": 4.7061162437141855,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 20119
    },
    {
      "epoch": 4.706350134487194,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6163,
      "step": 20120
    },
    {
      "epoch": 4.706584025260203,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7547,
      "step": 20121
    },
    {
      "epoch": 4.706817916033213,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 20122
    },
    {
      "epoch": 4.707051806806222,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 20123
    },
    {
      "epoch": 4.7072856975792305,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 20124
    },
    {
      "epoch": 4.707519588352239,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 20125
    },
    {
      "epoch": 4.707753479125248,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 20126
    },
    {
      "epoch": 4.707987369898257,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 20127
    },
    {
      "epoch": 4.708221260671267,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 20128
    },
    {
      "epoch": 4.708455151444276,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7307,
      "step": 20129
    },
    {
      "epoch": 4.7086890422172845,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 20130
    },
    {
      "epoch": 4.708922932990293,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 20131
    },
    {
      "epoch": 4.709156823763303,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 20132
    },
    {
      "epoch": 4.709390714536312,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 20133
    },
    {
      "epoch": 4.709624605309321,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 20134
    },
    {
      "epoch": 4.7098584960823295,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 20135
    },
    {
      "epoch": 4.710092386855338,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 20136
    },
    {
      "epoch": 4.710326277628347,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 20137
    },
    {
      "epoch": 4.710560168401357,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 20138
    },
    {
      "epoch": 4.710794059174366,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 20139
    },
    {
      "epoch": 4.711027949947375,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6792,
      "step": 20140
    },
    {
      "epoch": 4.7112618407203835,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7625,
      "step": 20141
    },
    {
      "epoch": 4.711495731493392,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9538,
      "step": 20142
    },
    {
      "epoch": 4.711729622266402,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6306,
      "step": 20143
    },
    {
      "epoch": 4.711963513039411,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 20144
    },
    {
      "epoch": 4.71219740381242,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 20145
    },
    {
      "epoch": 4.7124312945854285,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 20146
    },
    {
      "epoch": 4.712665185358437,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 20147
    },
    {
      "epoch": 4.712899076131446,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 20148
    },
    {
      "epoch": 4.713132966904456,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8424,
      "step": 20149
    },
    {
      "epoch": 4.713366857677465,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 20150
    },
    {
      "epoch": 4.713600748450474,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 20151
    },
    {
      "epoch": 4.7138346392234824,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 20152
    },
    {
      "epoch": 4.714068529996491,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 20153
    },
    {
      "epoch": 4.714302420769501,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 20154
    },
    {
      "epoch": 4.71453631154251,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 20155
    },
    {
      "epoch": 4.714770202315519,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 20156
    },
    {
      "epoch": 4.7150040930885275,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 20157
    },
    {
      "epoch": 4.715237983861536,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 20158
    },
    {
      "epoch": 4.715471874634545,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 20159
    },
    {
      "epoch": 4.715705765407555,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5874,
      "step": 20160
    },
    {
      "epoch": 4.715939656180564,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 20161
    },
    {
      "epoch": 4.716173546953573,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 20162
    },
    {
      "epoch": 4.716407437726581,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5286,
      "step": 20163
    },
    {
      "epoch": 4.716641328499591,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 20164
    },
    {
      "epoch": 4.7168752192726,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 20165
    },
    {
      "epoch": 4.717109110045609,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 20166
    },
    {
      "epoch": 4.717343000818618,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6647,
      "step": 20167
    },
    {
      "epoch": 4.7175768915916265,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 20168
    },
    {
      "epoch": 4.717810782364635,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7356,
      "step": 20169
    },
    {
      "epoch": 4.718044673137645,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 20170
    },
    {
      "epoch": 4.718278563910654,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 20171
    },
    {
      "epoch": 4.718512454683663,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5867,
      "step": 20172
    },
    {
      "epoch": 4.718746345456672,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 20173
    },
    {
      "epoch": 4.71898023622968,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 20174
    },
    {
      "epoch": 4.71921412700269,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6174,
      "step": 20175
    },
    {
      "epoch": 4.719448017775699,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 20176
    },
    {
      "epoch": 4.719681908548708,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 20177
    },
    {
      "epoch": 4.719915799321717,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 20178
    },
    {
      "epoch": 4.7201496900947255,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7018,
      "step": 20179
    },
    {
      "epoch": 4.720383580867734,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 20180
    },
    {
      "epoch": 4.720617471640744,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 20181
    },
    {
      "epoch": 4.720851362413753,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 20182
    },
    {
      "epoch": 4.721085253186762,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 20183
    },
    {
      "epoch": 4.721319143959771,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5656,
      "step": 20184
    },
    {
      "epoch": 4.721553034732779,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 20185
    },
    {
      "epoch": 4.721786925505789,
      "grad_norm": 7.71875,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 20186
    },
    {
      "epoch": 4.722020816278798,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6095,
      "step": 20187
    },
    {
      "epoch": 4.722254707051807,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 20188
    },
    {
      "epoch": 4.722488597824816,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 20189
    },
    {
      "epoch": 4.7227224885978245,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 20190
    },
    {
      "epoch": 4.722956379370833,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 20191
    },
    {
      "epoch": 4.723190270143843,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 20192
    },
    {
      "epoch": 4.723424160916852,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 20193
    },
    {
      "epoch": 4.723658051689861,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.531,
      "step": 20194
    },
    {
      "epoch": 4.72389194246287,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7103,
      "step": 20195
    },
    {
      "epoch": 4.724125833235879,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 20196
    },
    {
      "epoch": 4.724359724008888,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6996,
      "step": 20197
    },
    {
      "epoch": 4.724593614781897,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 20198
    },
    {
      "epoch": 4.724827505554906,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 20199
    },
    {
      "epoch": 4.725061396327915,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 20200
    },
    {
      "epoch": 4.725061396327915,
      "eval_runtime": 4.6478,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 20200
    },
    {
      "epoch": 4.7252952871009235,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 20201
    },
    {
      "epoch": 4.725529177873933,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6836,
      "step": 20202
    },
    {
      "epoch": 4.725763068646942,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.528,
      "step": 20203
    },
    {
      "epoch": 4.725996959419951,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6319,
      "step": 20204
    },
    {
      "epoch": 4.72623085019296,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 20205
    },
    {
      "epoch": 4.726464740965969,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 20206
    },
    {
      "epoch": 4.726698631738978,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 20207
    },
    {
      "epoch": 4.726932522511987,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7047,
      "step": 20208
    },
    {
      "epoch": 4.727166413284996,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8653,
      "step": 20209
    },
    {
      "epoch": 4.727400304058005,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 20210
    },
    {
      "epoch": 4.727634194831014,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8891,
      "step": 20211
    },
    {
      "epoch": 4.7278680856040225,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 20212
    },
    {
      "epoch": 4.728101976377032,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 20213
    },
    {
      "epoch": 4.728335867150041,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 20214
    },
    {
      "epoch": 4.72856975792305,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.715,
      "step": 20215
    },
    {
      "epoch": 4.728803648696059,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 20216
    },
    {
      "epoch": 4.729037539469068,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 20217
    },
    {
      "epoch": 4.729271430242077,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 20218
    },
    {
      "epoch": 4.729505321015086,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 20219
    },
    {
      "epoch": 4.729739211788095,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 20220
    },
    {
      "epoch": 4.729973102561104,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 20221
    },
    {
      "epoch": 4.730206993334113,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 20222
    },
    {
      "epoch": 4.7304408841071215,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 20223
    },
    {
      "epoch": 4.730674774880131,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 20224
    },
    {
      "epoch": 4.73090866565314,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7277,
      "step": 20225
    },
    {
      "epoch": 4.731142556426149,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 20226
    },
    {
      "epoch": 4.731376447199158,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 20227
    },
    {
      "epoch": 4.7316103379721675,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 20228
    },
    {
      "epoch": 4.731844228745176,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 20229
    },
    {
      "epoch": 4.732078119518185,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 20230
    },
    {
      "epoch": 4.732312010291194,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 20231
    },
    {
      "epoch": 4.732545901064203,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 20232
    },
    {
      "epoch": 4.732779791837212,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 20233
    },
    {
      "epoch": 4.733013682610221,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 20234
    },
    {
      "epoch": 4.73324757338323,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 20235
    },
    {
      "epoch": 4.733481464156239,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 20236
    },
    {
      "epoch": 4.733715354929248,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8456,
      "step": 20237
    },
    {
      "epoch": 4.733949245702257,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6271,
      "step": 20238
    },
    {
      "epoch": 4.7341831364752665,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 20239
    },
    {
      "epoch": 4.734417027248275,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 20240
    },
    {
      "epoch": 4.734650918021284,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 20241
    },
    {
      "epoch": 4.734884808794293,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 20242
    },
    {
      "epoch": 4.735118699567302,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 20243
    },
    {
      "epoch": 4.735352590340311,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2828,
      "step": 20244
    },
    {
      "epoch": 4.73558648111332,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 20245
    },
    {
      "epoch": 4.735820371886329,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 20246
    },
    {
      "epoch": 4.736054262659338,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 20247
    },
    {
      "epoch": 4.736288153432347,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 20248
    },
    {
      "epoch": 4.736522044205356,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 20249
    },
    {
      "epoch": 4.7367559349783654,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 20250
    },
    {
      "epoch": 4.736989825751374,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6415,
      "step": 20251
    },
    {
      "epoch": 4.737223716524383,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 20252
    },
    {
      "epoch": 4.737457607297392,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 20253
    },
    {
      "epoch": 4.737691498070401,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 20254
    },
    {
      "epoch": 4.73792538884341,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 20255
    },
    {
      "epoch": 4.738159279616419,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 20256
    },
    {
      "epoch": 4.738393170389428,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 20257
    },
    {
      "epoch": 4.738627061162437,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 20258
    },
    {
      "epoch": 4.738860951935446,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6798,
      "step": 20259
    },
    {
      "epoch": 4.739094842708456,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 20260
    },
    {
      "epoch": 4.739328733481464,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 20261
    },
    {
      "epoch": 4.739562624254473,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.624,
      "step": 20262
    },
    {
      "epoch": 4.739796515027482,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 20263
    },
    {
      "epoch": 4.740030405800491,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 20264
    },
    {
      "epoch": 4.7402642965735,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 20265
    },
    {
      "epoch": 4.7404981873465095,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 20266
    },
    {
      "epoch": 4.740732078119518,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 20267
    },
    {
      "epoch": 4.740965968892527,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 20268
    },
    {
      "epoch": 4.741199859665536,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 20269
    },
    {
      "epoch": 4.741433750438545,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.611,
      "step": 20270
    },
    {
      "epoch": 4.741667641211555,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.4987,
      "step": 20271
    },
    {
      "epoch": 4.741901531984563,
      "grad_norm": 14.25,
      "learning_rate": 3e-05,
      "loss": 2.1615,
      "step": 20272
    },
    {
      "epoch": 4.742135422757572,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2704,
      "step": 20273
    },
    {
      "epoch": 4.742369313530581,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6576,
      "step": 20274
    },
    {
      "epoch": 4.74260320430359,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 20275
    },
    {
      "epoch": 4.742837095076599,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 20276
    },
    {
      "epoch": 4.7430709858496085,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7185,
      "step": 20277
    },
    {
      "epoch": 4.743304876622617,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 20278
    },
    {
      "epoch": 4.743538767395626,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 20279
    },
    {
      "epoch": 4.743772658168635,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6973,
      "step": 20280
    },
    {
      "epoch": 4.744006548941644,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7396,
      "step": 20281
    },
    {
      "epoch": 4.744240439714654,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 20282
    },
    {
      "epoch": 4.744474330487662,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 20283
    },
    {
      "epoch": 4.744708221260671,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7419,
      "step": 20284
    },
    {
      "epoch": 4.74494211203368,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 20285
    },
    {
      "epoch": 4.745176002806689,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 20286
    },
    {
      "epoch": 4.745409893579699,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9506,
      "step": 20287
    },
    {
      "epoch": 4.7456437843527075,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7694,
      "step": 20288
    },
    {
      "epoch": 4.745877675125716,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 20289
    },
    {
      "epoch": 4.746111565898725,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 20290
    },
    {
      "epoch": 4.746345456671734,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8568,
      "step": 20291
    },
    {
      "epoch": 4.746579347444744,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 20292
    },
    {
      "epoch": 4.746813238217753,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 20293
    },
    {
      "epoch": 4.747047128990761,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 20294
    },
    {
      "epoch": 4.74728101976377,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5232,
      "step": 20295
    },
    {
      "epoch": 4.747514910536779,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6164,
      "step": 20296
    },
    {
      "epoch": 4.747748801309788,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6601,
      "step": 20297
    },
    {
      "epoch": 4.747982692082798,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6383,
      "step": 20298
    },
    {
      "epoch": 4.7482165828558065,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 20299
    },
    {
      "epoch": 4.748450473628815,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 20300
    },
    {
      "epoch": 4.748450473628815,
      "eval_runtime": 4.6401,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 20300
    },
    {
      "epoch": 4.748684364401824,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 20301
    },
    {
      "epoch": 4.748918255174833,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 20302
    },
    {
      "epoch": 4.749152145947843,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 20303
    },
    {
      "epoch": 4.749386036720852,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 20304
    },
    {
      "epoch": 4.74961992749386,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 20305
    },
    {
      "epoch": 4.749853818266869,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.6385,
      "step": 20306
    },
    {
      "epoch": 4.750087709039878,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 20307
    },
    {
      "epoch": 4.750321599812887,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 2.3158,
      "step": 20308
    },
    {
      "epoch": 4.750555490585897,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 20309
    },
    {
      "epoch": 4.7507893813589055,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6334,
      "step": 20310
    },
    {
      "epoch": 4.751023272131914,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6375,
      "step": 20311
    },
    {
      "epoch": 4.751257162904923,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 20312
    },
    {
      "epoch": 4.751491053677933,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 20313
    },
    {
      "epoch": 4.751724944450942,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8837,
      "step": 20314
    },
    {
      "epoch": 4.751958835223951,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.762,
      "step": 20315
    },
    {
      "epoch": 4.752192725996959,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6542,
      "step": 20316
    },
    {
      "epoch": 4.752426616769968,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7152,
      "step": 20317
    },
    {
      "epoch": 4.752660507542977,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 20318
    },
    {
      "epoch": 4.752894398315987,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7554,
      "step": 20319
    },
    {
      "epoch": 4.753128289088996,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 20320
    },
    {
      "epoch": 4.7533621798620045,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 20321
    },
    {
      "epoch": 4.753596070635013,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7482,
      "step": 20322
    },
    {
      "epoch": 4.753829961408022,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 20323
    },
    {
      "epoch": 4.754063852181032,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6473,
      "step": 20324
    },
    {
      "epoch": 4.754297742954041,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 20325
    },
    {
      "epoch": 4.75453163372705,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6075,
      "step": 20326
    },
    {
      "epoch": 4.754765524500058,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5861,
      "step": 20327
    },
    {
      "epoch": 4.754999415273067,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2637,
      "step": 20328
    },
    {
      "epoch": 4.755233306046076,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 20329
    },
    {
      "epoch": 4.755467196819086,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3676,
      "step": 20330
    },
    {
      "epoch": 4.755701087592095,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 20331
    },
    {
      "epoch": 4.7559349783651035,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 20332
    },
    {
      "epoch": 4.756168869138112,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7694,
      "step": 20333
    },
    {
      "epoch": 4.756402759911121,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 20334
    },
    {
      "epoch": 4.756636650684131,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 20335
    },
    {
      "epoch": 4.75687054145714,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 20336
    },
    {
      "epoch": 4.757104432230149,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 20337
    },
    {
      "epoch": 4.757338323003157,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 20338
    },
    {
      "epoch": 4.757572213776166,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 20339
    },
    {
      "epoch": 4.757806104549175,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 20340
    },
    {
      "epoch": 4.758039995322185,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 20341
    },
    {
      "epoch": 4.758273886095194,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 20342
    },
    {
      "epoch": 4.7585077768682025,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6893,
      "step": 20343
    },
    {
      "epoch": 4.758741667641211,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 20344
    },
    {
      "epoch": 4.758975558414221,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 20345
    },
    {
      "epoch": 4.75920944918723,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 20346
    },
    {
      "epoch": 4.759443339960239,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 20347
    },
    {
      "epoch": 4.7596772307332476,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 20348
    },
    {
      "epoch": 4.759911121506256,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 20349
    },
    {
      "epoch": 4.760145012279265,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 20350
    },
    {
      "epoch": 4.760378903052275,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7212,
      "step": 20351
    },
    {
      "epoch": 4.760612793825284,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 20352
    },
    {
      "epoch": 4.760846684598293,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 20353
    },
    {
      "epoch": 4.7610805753713015,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 20354
    },
    {
      "epoch": 4.76131446614431,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 20355
    },
    {
      "epoch": 4.76154835691732,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 20356
    },
    {
      "epoch": 4.761782247690329,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 20357
    },
    {
      "epoch": 4.762016138463338,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 20358
    },
    {
      "epoch": 4.7622500292363465,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 20359
    },
    {
      "epoch": 4.762483920009355,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 20360
    },
    {
      "epoch": 4.762717810782364,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0138,
      "step": 20361
    },
    {
      "epoch": 4.762951701555374,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 20362
    },
    {
      "epoch": 4.763185592328383,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 20363
    },
    {
      "epoch": 4.763419483101392,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8177,
      "step": 20364
    },
    {
      "epoch": 4.7636533738744005,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7335,
      "step": 20365
    },
    {
      "epoch": 4.763887264647409,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 20366
    },
    {
      "epoch": 4.764121155420419,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 20367
    },
    {
      "epoch": 4.764355046193428,
      "grad_norm": 11.0625,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 20368
    },
    {
      "epoch": 4.764588936966437,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 20369
    },
    {
      "epoch": 4.7648228277394455,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 20370
    },
    {
      "epoch": 4.765056718512454,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 20371
    },
    {
      "epoch": 4.765290609285463,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 20372
    },
    {
      "epoch": 4.765524500058473,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 20373
    },
    {
      "epoch": 4.765758390831482,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.5776,
      "step": 20374
    },
    {
      "epoch": 4.765992281604491,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 20375
    },
    {
      "epoch": 4.7662261723774995,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 20376
    },
    {
      "epoch": 4.766460063150509,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 20377
    },
    {
      "epoch": 4.766693953923518,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 20378
    },
    {
      "epoch": 4.766927844696527,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 20379
    },
    {
      "epoch": 4.767161735469536,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 20380
    },
    {
      "epoch": 4.7673956262425445,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 20381
    },
    {
      "epoch": 4.767629517015553,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 20382
    },
    {
      "epoch": 4.767863407788563,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 20383
    },
    {
      "epoch": 4.768097298561572,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4191,
      "step": 20384
    },
    {
      "epoch": 4.768331189334581,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6568,
      "step": 20385
    },
    {
      "epoch": 4.76856508010759,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 20386
    },
    {
      "epoch": 4.7687989708805985,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 20387
    },
    {
      "epoch": 4.769032861653608,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8902,
      "step": 20388
    },
    {
      "epoch": 4.769266752426617,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 20389
    },
    {
      "epoch": 4.769500643199626,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9756,
      "step": 20390
    },
    {
      "epoch": 4.769734533972635,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 20391
    },
    {
      "epoch": 4.7699684247456435,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 20392
    },
    {
      "epoch": 4.770202315518652,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 20393
    },
    {
      "epoch": 4.770436206291662,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5982,
      "step": 20394
    },
    {
      "epoch": 4.770670097064671,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5205,
      "step": 20395
    },
    {
      "epoch": 4.77090398783768,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 20396
    },
    {
      "epoch": 4.771137878610689,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 20397
    },
    {
      "epoch": 4.7713717693836974,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 20398
    },
    {
      "epoch": 4.771605660156707,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 20399
    },
    {
      "epoch": 4.771839550929716,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 20400
    },
    {
      "epoch": 4.771839550929716,
      "eval_runtime": 4.5807,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 20400
    },
    {
      "epoch": 4.772073441702725,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 20401
    },
    {
      "epoch": 4.772307332475734,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5146,
      "step": 20402
    },
    {
      "epoch": 4.7725412232487425,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 20403
    },
    {
      "epoch": 4.772775114021751,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 20404
    },
    {
      "epoch": 4.773009004794761,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 20405
    },
    {
      "epoch": 4.77324289556777,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 20406
    },
    {
      "epoch": 4.773476786340779,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.685,
      "step": 20407
    },
    {
      "epoch": 4.773710677113788,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 20408
    },
    {
      "epoch": 4.773944567886797,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 20409
    },
    {
      "epoch": 4.774178458659806,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 20410
    },
    {
      "epoch": 4.774412349432815,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 20411
    },
    {
      "epoch": 4.774646240205824,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 20412
    },
    {
      "epoch": 4.774880130978833,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 20413
    },
    {
      "epoch": 4.7751140217518415,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 20414
    },
    {
      "epoch": 4.775347912524851,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5515,
      "step": 20415
    },
    {
      "epoch": 4.77558180329786,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 20416
    },
    {
      "epoch": 4.775815694070869,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 20417
    },
    {
      "epoch": 4.776049584843878,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 20418
    },
    {
      "epoch": 4.776283475616887,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 20419
    },
    {
      "epoch": 4.776517366389896,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6283,
      "step": 20420
    },
    {
      "epoch": 4.776751257162905,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 20421
    },
    {
      "epoch": 4.776985147935914,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 20422
    },
    {
      "epoch": 4.777219038708923,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 20423
    },
    {
      "epoch": 4.777452929481932,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 20424
    },
    {
      "epoch": 4.7776868202549405,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.7627,
      "step": 20425
    },
    {
      "epoch": 4.77792071102795,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7893,
      "step": 20426
    },
    {
      "epoch": 4.778154601800959,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 20427
    },
    {
      "epoch": 4.778388492573968,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.678,
      "step": 20428
    },
    {
      "epoch": 4.778622383346977,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 20429
    },
    {
      "epoch": 4.778856274119986,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 20430
    },
    {
      "epoch": 4.779090164892995,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 20431
    },
    {
      "epoch": 4.779324055666004,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 20432
    },
    {
      "epoch": 4.779557946439013,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 20433
    },
    {
      "epoch": 4.779791837212022,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 20434
    },
    {
      "epoch": 4.780025727985031,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 20435
    },
    {
      "epoch": 4.7802596187580395,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 20436
    },
    {
      "epoch": 4.780493509531049,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 20437
    },
    {
      "epoch": 4.780727400304058,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 20438
    },
    {
      "epoch": 4.780961291077067,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 20439
    },
    {
      "epoch": 4.781195181850076,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 20440
    },
    {
      "epoch": 4.7814290726230855,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 20441
    },
    {
      "epoch": 4.781662963396094,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 20442
    },
    {
      "epoch": 4.781896854169103,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 20443
    },
    {
      "epoch": 4.782130744942112,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 20444
    },
    {
      "epoch": 4.782364635715121,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5915,
      "step": 20445
    },
    {
      "epoch": 4.78259852648813,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 20446
    },
    {
      "epoch": 4.782832417261139,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 20447
    },
    {
      "epoch": 4.783066308034148,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 20448
    },
    {
      "epoch": 4.783300198807157,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 20449
    },
    {
      "epoch": 4.783534089580166,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 20450
    },
    {
      "epoch": 4.783767980353175,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6313,
      "step": 20451
    },
    {
      "epoch": 4.7840018711261845,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 20452
    },
    {
      "epoch": 4.784235761899193,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 20453
    },
    {
      "epoch": 4.784469652672202,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8001,
      "step": 20454
    },
    {
      "epoch": 4.784703543445211,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 20455
    },
    {
      "epoch": 4.78493743421822,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 20456
    },
    {
      "epoch": 4.785171324991229,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 20457
    },
    {
      "epoch": 4.785405215764238,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 20458
    },
    {
      "epoch": 4.785639106537247,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 20459
    },
    {
      "epoch": 4.785872997310256,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 20460
    },
    {
      "epoch": 4.786106888083265,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 20461
    },
    {
      "epoch": 4.786340778856274,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 20462
    },
    {
      "epoch": 4.7865746696292835,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 20463
    },
    {
      "epoch": 4.786808560402292,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 20464
    },
    {
      "epoch": 4.787042451175301,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 20465
    },
    {
      "epoch": 4.78727634194831,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 20466
    },
    {
      "epoch": 4.787510232721319,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 20467
    },
    {
      "epoch": 4.787744123494328,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 20468
    },
    {
      "epoch": 4.787978014267337,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 20469
    },
    {
      "epoch": 4.788211905040346,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6559,
      "step": 20470
    },
    {
      "epoch": 4.788445795813355,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 20471
    },
    {
      "epoch": 4.788679686586364,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 20472
    },
    {
      "epoch": 4.788913577359374,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6794,
      "step": 20473
    },
    {
      "epoch": 4.7891474681323825,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 20474
    },
    {
      "epoch": 4.789381358905391,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 20475
    },
    {
      "epoch": 4.7896152496784,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 20476
    },
    {
      "epoch": 4.789849140451409,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7883,
      "step": 20477
    },
    {
      "epoch": 4.790083031224418,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 20478
    },
    {
      "epoch": 4.7903169219974275,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4098,
      "step": 20479
    },
    {
      "epoch": 4.790550812770436,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 20480
    },
    {
      "epoch": 4.790784703543445,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 20481
    },
    {
      "epoch": 4.791018594316454,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 20482
    },
    {
      "epoch": 4.791252485089463,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 20483
    },
    {
      "epoch": 4.791486375862473,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 20484
    },
    {
      "epoch": 4.7917202666354815,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 20485
    },
    {
      "epoch": 4.79195415740849,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2233,
      "step": 20486
    },
    {
      "epoch": 4.792188048181499,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 20487
    },
    {
      "epoch": 4.792421938954508,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 20488
    },
    {
      "epoch": 4.792655829727517,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 20489
    },
    {
      "epoch": 4.7928897205005265,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 20490
    },
    {
      "epoch": 4.793123611273535,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 20491
    },
    {
      "epoch": 4.793357502046544,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6496,
      "step": 20492
    },
    {
      "epoch": 4.793591392819553,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 20493
    },
    {
      "epoch": 4.793825283592563,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 20494
    },
    {
      "epoch": 4.794059174365572,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 20495
    },
    {
      "epoch": 4.7942930651385804,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 20496
    },
    {
      "epoch": 4.794526955911589,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 20497
    },
    {
      "epoch": 4.794760846684598,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.579,
      "step": 20498
    },
    {
      "epoch": 4.794994737457607,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5595,
      "step": 20499
    },
    {
      "epoch": 4.795228628230617,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5606,
      "step": 20500
    },
    {
      "epoch": 4.795228628230617,
      "eval_runtime": 4.5817,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 20500
    },
    {
      "epoch": 4.7954625190036255,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 20501
    },
    {
      "epoch": 4.795696409776634,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 20502
    },
    {
      "epoch": 4.795930300549643,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 20503
    },
    {
      "epoch": 4.796164191322652,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 20504
    },
    {
      "epoch": 4.796398082095662,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 20505
    },
    {
      "epoch": 4.796631972868671,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 20506
    },
    {
      "epoch": 4.796865863641679,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 20507
    },
    {
      "epoch": 4.797099754414688,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 20508
    },
    {
      "epoch": 4.797333645187697,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 20509
    },
    {
      "epoch": 4.797567535960706,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 20510
    },
    {
      "epoch": 4.797801426733716,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6432,
      "step": 20511
    },
    {
      "epoch": 4.7980353175067245,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 20512
    },
    {
      "epoch": 4.798269208279733,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5219,
      "step": 20513
    },
    {
      "epoch": 4.798503099052742,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 20514
    },
    {
      "epoch": 4.798736989825751,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 20515
    },
    {
      "epoch": 4.798970880598761,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 20516
    },
    {
      "epoch": 4.79920477137177,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 20517
    },
    {
      "epoch": 4.799438662144778,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7669,
      "step": 20518
    },
    {
      "epoch": 4.799672552917787,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 20519
    },
    {
      "epoch": 4.799906443690796,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 20520
    },
    {
      "epoch": 4.800140334463805,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.727,
      "step": 20521
    },
    {
      "epoch": 4.800374225236815,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7254,
      "step": 20522
    },
    {
      "epoch": 4.8006081160098235,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 20523
    },
    {
      "epoch": 4.800842006782832,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 20524
    },
    {
      "epoch": 4.801075897555841,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 20525
    },
    {
      "epoch": 4.801309788328851,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.762,
      "step": 20526
    },
    {
      "epoch": 4.80154367910186,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 20527
    },
    {
      "epoch": 4.801777569874869,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 20528
    },
    {
      "epoch": 4.802011460647877,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 20529
    },
    {
      "epoch": 4.802245351420886,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 20530
    },
    {
      "epoch": 4.802479242193895,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 20531
    },
    {
      "epoch": 4.802713132966905,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 20532
    },
    {
      "epoch": 4.802947023739914,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 20533
    },
    {
      "epoch": 4.8031809145129225,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 20534
    },
    {
      "epoch": 4.803414805285931,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6592,
      "step": 20535
    },
    {
      "epoch": 4.80364869605894,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6961,
      "step": 20536
    },
    {
      "epoch": 4.80388258683195,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8038,
      "step": 20537
    },
    {
      "epoch": 4.804116477604959,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4886,
      "step": 20538
    },
    {
      "epoch": 4.804350368377968,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 20539
    },
    {
      "epoch": 4.804584259150976,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 20540
    },
    {
      "epoch": 4.804818149923985,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 20541
    },
    {
      "epoch": 4.805052040696994,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6736,
      "step": 20542
    },
    {
      "epoch": 4.805285931470004,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0967,
      "step": 20543
    },
    {
      "epoch": 4.805519822243013,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 20544
    },
    {
      "epoch": 4.8057537130160215,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 20545
    },
    {
      "epoch": 4.80598760378903,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 20546
    },
    {
      "epoch": 4.806221494562039,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 20547
    },
    {
      "epoch": 4.806455385335049,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 20548
    },
    {
      "epoch": 4.806689276108058,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 20549
    },
    {
      "epoch": 4.806923166881067,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 20550
    },
    {
      "epoch": 4.807157057654075,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5502,
      "step": 20551
    },
    {
      "epoch": 4.807390948427084,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.676,
      "step": 20552
    },
    {
      "epoch": 4.807624839200093,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 20553
    },
    {
      "epoch": 4.807858729973103,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8607,
      "step": 20554
    },
    {
      "epoch": 4.808092620746112,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 20555
    },
    {
      "epoch": 4.8083265115191205,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 20556
    },
    {
      "epoch": 4.808560402292129,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 20557
    },
    {
      "epoch": 4.808794293065139,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 20558
    },
    {
      "epoch": 4.809028183838148,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 20559
    },
    {
      "epoch": 4.809262074611157,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 20560
    },
    {
      "epoch": 4.809495965384166,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 20561
    },
    {
      "epoch": 4.809729856157174,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 20562
    },
    {
      "epoch": 4.809963746930183,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 20563
    },
    {
      "epoch": 4.810197637703193,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 20564
    },
    {
      "epoch": 4.810431528476202,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 20565
    },
    {
      "epoch": 4.810665419249211,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 20566
    },
    {
      "epoch": 4.8108993100222195,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5492,
      "step": 20567
    },
    {
      "epoch": 4.811133200795228,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4948,
      "step": 20568
    },
    {
      "epoch": 4.811367091568238,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7906,
      "step": 20569
    },
    {
      "epoch": 4.811600982341247,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 20570
    },
    {
      "epoch": 4.811834873114256,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 20571
    },
    {
      "epoch": 4.812068763887265,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 20572
    },
    {
      "epoch": 4.812302654660273,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.676,
      "step": 20573
    },
    {
      "epoch": 4.812536545433282,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5024,
      "step": 20574
    },
    {
      "epoch": 4.812770436206292,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 20575
    },
    {
      "epoch": 4.813004326979301,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6309,
      "step": 20576
    },
    {
      "epoch": 4.81323821775231,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 20577
    },
    {
      "epoch": 4.8134721085253185,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 20578
    },
    {
      "epoch": 4.813705999298327,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7947,
      "step": 20579
    },
    {
      "epoch": 4.813939890071337,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.688,
      "step": 20580
    },
    {
      "epoch": 4.814173780844346,
      "grad_norm": 7.71875,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 20581
    },
    {
      "epoch": 4.814407671617355,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7175,
      "step": 20582
    },
    {
      "epoch": 4.814641562390364,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6667,
      "step": 20583
    },
    {
      "epoch": 4.814875453163372,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 20584
    },
    {
      "epoch": 4.815109343936381,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8503,
      "step": 20585
    },
    {
      "epoch": 4.815343234709391,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 20586
    },
    {
      "epoch": 4.8155771254824,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 20587
    },
    {
      "epoch": 4.815811016255409,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7308,
      "step": 20588
    },
    {
      "epoch": 4.8160449070284175,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 20589
    },
    {
      "epoch": 4.816278797801427,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3382,
      "step": 20590
    },
    {
      "epoch": 4.816512688574436,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 20591
    },
    {
      "epoch": 4.816746579347445,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 20592
    },
    {
      "epoch": 4.816980470120454,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8576,
      "step": 20593
    },
    {
      "epoch": 4.8172143608934626,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5146,
      "step": 20594
    },
    {
      "epoch": 4.817448251666471,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 20595
    },
    {
      "epoch": 4.817682142439481,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 20596
    },
    {
      "epoch": 4.81791603321249,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 20597
    },
    {
      "epoch": 4.818149923985499,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 20598
    },
    {
      "epoch": 4.818383814758508,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4088,
      "step": 20599
    },
    {
      "epoch": 4.8186177055315165,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 20600
    },
    {
      "epoch": 4.8186177055315165,
      "eval_runtime": 4.6534,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 20600
    },
    {
      "epoch": 4.818851596304526,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 20601
    },
    {
      "epoch": 4.819085487077535,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 20602
    },
    {
      "epoch": 4.819319377850544,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9513,
      "step": 20603
    },
    {
      "epoch": 4.819553268623553,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 20604
    },
    {
      "epoch": 4.8197871593965615,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 20605
    },
    {
      "epoch": 4.82002105016957,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 20606
    },
    {
      "epoch": 4.82025494094258,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7282,
      "step": 20607
    },
    {
      "epoch": 4.820488831715589,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 20608
    },
    {
      "epoch": 4.820722722488598,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.566,
      "step": 20609
    },
    {
      "epoch": 4.820956613261607,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.593,
      "step": 20610
    },
    {
      "epoch": 4.8211905040346155,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 20611
    },
    {
      "epoch": 4.821424394807625,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 20612
    },
    {
      "epoch": 4.821658285580634,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 20613
    },
    {
      "epoch": 4.821892176353643,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 20614
    },
    {
      "epoch": 4.822126067126652,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 20615
    },
    {
      "epoch": 4.8223599578996605,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.4311,
      "step": 20616
    },
    {
      "epoch": 4.822593848672669,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 20617
    },
    {
      "epoch": 4.822827739445679,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 20618
    },
    {
      "epoch": 4.823061630218688,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 20619
    },
    {
      "epoch": 4.823295520991697,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 20620
    },
    {
      "epoch": 4.823529411764706,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 20621
    },
    {
      "epoch": 4.823763302537715,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 20622
    },
    {
      "epoch": 4.823997193310724,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 20623
    },
    {
      "epoch": 4.824231084083733,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 20624
    },
    {
      "epoch": 4.824464974856742,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 20625
    },
    {
      "epoch": 4.824698865629751,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 20626
    },
    {
      "epoch": 4.8249327564027595,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 20627
    },
    {
      "epoch": 4.825166647175769,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 20628
    },
    {
      "epoch": 4.825400537948778,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 20629
    },
    {
      "epoch": 4.825634428721787,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 20630
    },
    {
      "epoch": 4.825868319494796,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 20631
    },
    {
      "epoch": 4.826102210267805,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 20632
    },
    {
      "epoch": 4.826336101040814,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.6051,
      "step": 20633
    },
    {
      "epoch": 4.826569991813823,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 20634
    },
    {
      "epoch": 4.826803882586832,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9133,
      "step": 20635
    },
    {
      "epoch": 4.827037773359841,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7221,
      "step": 20636
    },
    {
      "epoch": 4.82727166413285,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6026,
      "step": 20637
    },
    {
      "epoch": 4.8275055549058585,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 20638
    },
    {
      "epoch": 4.827739445678868,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.677,
      "step": 20639
    },
    {
      "epoch": 4.827973336451877,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 20640
    },
    {
      "epoch": 4.828207227224886,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7954,
      "step": 20641
    },
    {
      "epoch": 4.828441117997895,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 20642
    },
    {
      "epoch": 4.828675008770904,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 20643
    },
    {
      "epoch": 4.828908899543913,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 20644
    },
    {
      "epoch": 4.829142790316922,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1573,
      "step": 20645
    },
    {
      "epoch": 4.829376681089931,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.637,
      "step": 20646
    },
    {
      "epoch": 4.82961057186294,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 20647
    },
    {
      "epoch": 4.829844462635949,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 20648
    },
    {
      "epoch": 4.8300783534089575,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 20649
    },
    {
      "epoch": 4.830312244181967,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 20650
    },
    {
      "epoch": 4.830546134954976,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 20651
    },
    {
      "epoch": 4.830780025727985,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 20652
    },
    {
      "epoch": 4.831013916500994,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 20653
    },
    {
      "epoch": 4.8312478072740035,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5795,
      "step": 20654
    },
    {
      "epoch": 4.831481698047012,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 20655
    },
    {
      "epoch": 4.831715588820021,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7108,
      "step": 20656
    },
    {
      "epoch": 4.83194947959303,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 20657
    },
    {
      "epoch": 4.832183370366039,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 20658
    },
    {
      "epoch": 4.832417261139048,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 20659
    },
    {
      "epoch": 4.832651151912057,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 20660
    },
    {
      "epoch": 4.832885042685066,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 20661
    },
    {
      "epoch": 4.833118933458075,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5099,
      "step": 20662
    },
    {
      "epoch": 4.833352824231084,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 20663
    },
    {
      "epoch": 4.833586715004093,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 20664
    },
    {
      "epoch": 4.8338206057771025,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0299,
      "step": 20665
    },
    {
      "epoch": 4.834054496550111,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 20666
    },
    {
      "epoch": 4.83428838732312,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5818,
      "step": 20667
    },
    {
      "epoch": 4.834522278096129,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7846,
      "step": 20668
    },
    {
      "epoch": 4.834756168869138,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 20669
    },
    {
      "epoch": 4.834990059642147,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7549,
      "step": 20670
    },
    {
      "epoch": 4.835223950415156,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 20671
    },
    {
      "epoch": 4.835457841188165,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 20672
    },
    {
      "epoch": 4.835691731961174,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 20673
    },
    {
      "epoch": 4.835925622734183,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 20674
    },
    {
      "epoch": 4.836159513507192,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5708,
      "step": 20675
    },
    {
      "epoch": 4.8363934042802015,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6633,
      "step": 20676
    },
    {
      "epoch": 4.83662729505321,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 20677
    },
    {
      "epoch": 4.836861185826219,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 20678
    },
    {
      "epoch": 4.837095076599228,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 20679
    },
    {
      "epoch": 4.837328967372237,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 20680
    },
    {
      "epoch": 4.837562858145247,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 20681
    },
    {
      "epoch": 4.837796748918255,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 20682
    },
    {
      "epoch": 4.838030639691264,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 20683
    },
    {
      "epoch": 4.838264530464273,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 20684
    },
    {
      "epoch": 4.838498421237282,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 20685
    },
    {
      "epoch": 4.838732312010292,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 20686
    },
    {
      "epoch": 4.8389662027833005,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5609,
      "step": 20687
    },
    {
      "epoch": 4.839200093556309,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 20688
    },
    {
      "epoch": 4.839433984329318,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 20689
    },
    {
      "epoch": 4.839667875102327,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9204,
      "step": 20690
    },
    {
      "epoch": 4.839901765875336,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 20691
    },
    {
      "epoch": 4.8401356566483456,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.64,
      "step": 20692
    },
    {
      "epoch": 4.840369547421354,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 20693
    },
    {
      "epoch": 4.840603438194363,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 20694
    },
    {
      "epoch": 4.840837328967372,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 20695
    },
    {
      "epoch": 4.841071219740381,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 20696
    },
    {
      "epoch": 4.841305110513391,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 20697
    },
    {
      "epoch": 4.8415390012863995,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 20698
    },
    {
      "epoch": 4.841772892059408,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 20699
    },
    {
      "epoch": 4.842006782832417,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 20700
    },
    {
      "epoch": 4.842006782832417,
      "eval_runtime": 4.6485,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 20700
    },
    {
      "epoch": 4.842240673605426,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 20701
    },
    {
      "epoch": 4.842474564378435,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 20702
    },
    {
      "epoch": 4.8427084551514445,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.4296,
      "step": 20703
    },
    {
      "epoch": 4.842942345924453,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 20704
    },
    {
      "epoch": 4.843176236697462,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 20705
    },
    {
      "epoch": 4.843410127470471,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 20706
    },
    {
      "epoch": 4.843644018243481,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7209,
      "step": 20707
    },
    {
      "epoch": 4.84387790901649,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 20708
    },
    {
      "epoch": 4.8441117997894985,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 20709
    },
    {
      "epoch": 4.844345690562507,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 20710
    },
    {
      "epoch": 4.844579581335516,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 20711
    },
    {
      "epoch": 4.844813472108525,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 20712
    },
    {
      "epoch": 4.845047362881535,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 20713
    },
    {
      "epoch": 4.8452812536545435,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 20714
    },
    {
      "epoch": 4.845515144427552,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 20715
    },
    {
      "epoch": 4.845749035200561,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4566,
      "step": 20716
    },
    {
      "epoch": 4.84598292597357,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 20717
    },
    {
      "epoch": 4.84621681674658,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 20718
    },
    {
      "epoch": 4.846450707519589,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 20719
    },
    {
      "epoch": 4.8466845982925975,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 20720
    },
    {
      "epoch": 4.846918489065606,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 20721
    },
    {
      "epoch": 4.847152379838615,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 20722
    },
    {
      "epoch": 4.847386270611624,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8535,
      "step": 20723
    },
    {
      "epoch": 4.847620161384634,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 20724
    },
    {
      "epoch": 4.8478540521576425,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7245,
      "step": 20725
    },
    {
      "epoch": 4.848087942930651,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7305,
      "step": 20726
    },
    {
      "epoch": 4.84832183370366,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 20727
    },
    {
      "epoch": 4.848555724476669,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 20728
    },
    {
      "epoch": 4.848789615249679,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 20729
    },
    {
      "epoch": 4.849023506022688,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 20730
    },
    {
      "epoch": 4.8492573967956965,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 20731
    },
    {
      "epoch": 4.849491287568705,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 20732
    },
    {
      "epoch": 4.849725178341714,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 20733
    },
    {
      "epoch": 4.849959069114723,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 20734
    },
    {
      "epoch": 4.850192959887733,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 20735
    },
    {
      "epoch": 4.8504268506607415,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6642,
      "step": 20736
    },
    {
      "epoch": 4.85066074143375,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 20737
    },
    {
      "epoch": 4.850894632206759,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 20738
    },
    {
      "epoch": 4.851128522979769,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 20739
    },
    {
      "epoch": 4.851362413752778,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 20740
    },
    {
      "epoch": 4.851596304525787,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 20741
    },
    {
      "epoch": 4.8518301952987954,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6812,
      "step": 20742
    },
    {
      "epoch": 4.852064086071804,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6243,
      "step": 20743
    },
    {
      "epoch": 4.852297976844813,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 20744
    },
    {
      "epoch": 4.852531867617823,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 20745
    },
    {
      "epoch": 4.852765758390832,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 20746
    },
    {
      "epoch": 4.8529996491638405,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 20747
    },
    {
      "epoch": 4.853233539936849,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 20748
    },
    {
      "epoch": 4.853467430709858,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 20749
    },
    {
      "epoch": 4.853701321482868,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 20750
    },
    {
      "epoch": 4.853935212255877,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6436,
      "step": 20751
    },
    {
      "epoch": 4.854169103028886,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5361,
      "step": 20752
    },
    {
      "epoch": 4.854402993801894,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 20753
    },
    {
      "epoch": 4.854636884574903,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 20754
    },
    {
      "epoch": 4.854870775347912,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 20755
    },
    {
      "epoch": 4.855104666120922,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7662,
      "step": 20756
    },
    {
      "epoch": 4.855338556893931,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 20757
    },
    {
      "epoch": 4.8555724476669395,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 20758
    },
    {
      "epoch": 4.855806338439948,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 20759
    },
    {
      "epoch": 4.856040229212957,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 20760
    },
    {
      "epoch": 4.856274119985967,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 20761
    },
    {
      "epoch": 4.856508010758976,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.4508,
      "step": 20762
    },
    {
      "epoch": 4.856741901531985,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 20763
    },
    {
      "epoch": 4.856975792304993,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 20764
    },
    {
      "epoch": 4.857209683078002,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 20765
    },
    {
      "epoch": 4.857443573851011,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 20766
    },
    {
      "epoch": 4.857677464624021,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 20767
    },
    {
      "epoch": 4.85791135539703,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 20768
    },
    {
      "epoch": 4.8581452461700385,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 20769
    },
    {
      "epoch": 4.858379136943047,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 20770
    },
    {
      "epoch": 4.858613027716057,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4696,
      "step": 20771
    },
    {
      "epoch": 4.858846918489066,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6889,
      "step": 20772
    },
    {
      "epoch": 4.859080809262075,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 20773
    },
    {
      "epoch": 4.859314700035084,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 20774
    },
    {
      "epoch": 4.859548590808092,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9363,
      "step": 20775
    },
    {
      "epoch": 4.859782481581101,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6235,
      "step": 20776
    },
    {
      "epoch": 4.860016372354111,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 20777
    },
    {
      "epoch": 4.86025026312712,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 20778
    },
    {
      "epoch": 4.860484153900129,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 20779
    },
    {
      "epoch": 4.8607180446731375,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 20780
    },
    {
      "epoch": 4.860951935446146,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 20781
    },
    {
      "epoch": 4.861185826219156,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.555,
      "step": 20782
    },
    {
      "epoch": 4.861419716992165,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6229,
      "step": 20783
    },
    {
      "epoch": 4.861653607765174,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 20784
    },
    {
      "epoch": 4.861887498538183,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8242,
      "step": 20785
    },
    {
      "epoch": 4.862121389311191,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 20786
    },
    {
      "epoch": 4.8623552800842,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 20787
    },
    {
      "epoch": 4.86258917085721,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6237,
      "step": 20788
    },
    {
      "epoch": 4.862823061630219,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3963,
      "step": 20789
    },
    {
      "epoch": 4.863056952403228,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 20790
    },
    {
      "epoch": 4.8632908431762365,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 20791
    },
    {
      "epoch": 4.863524733949245,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 20792
    },
    {
      "epoch": 4.863758624722255,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 20793
    },
    {
      "epoch": 4.863992515495264,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 20794
    },
    {
      "epoch": 4.864226406268273,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 20795
    },
    {
      "epoch": 4.864460297041282,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 20796
    },
    {
      "epoch": 4.86469418781429,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 20797
    },
    {
      "epoch": 4.864928078587299,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.758,
      "step": 20798
    },
    {
      "epoch": 4.865161969360309,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5781,
      "step": 20799
    },
    {
      "epoch": 4.865395860133318,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 20800
    },
    {
      "epoch": 4.865395860133318,
      "eval_runtime": 4.6401,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 20800
    },
    {
      "epoch": 4.865629750906327,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 20801
    },
    {
      "epoch": 4.8658636416793355,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5026,
      "step": 20802
    },
    {
      "epoch": 4.866097532452345,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 20803
    },
    {
      "epoch": 4.866331423225354,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 20804
    },
    {
      "epoch": 4.866565313998363,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 20805
    },
    {
      "epoch": 4.866799204771372,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 20806
    },
    {
      "epoch": 4.867033095544381,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6962,
      "step": 20807
    },
    {
      "epoch": 4.867266986317389,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 20808
    },
    {
      "epoch": 4.867500877090399,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6912,
      "step": 20809
    },
    {
      "epoch": 4.867734767863408,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 20810
    },
    {
      "epoch": 4.867968658636417,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8824,
      "step": 20811
    },
    {
      "epoch": 4.868202549409426,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 20812
    },
    {
      "epoch": 4.8684364401824345,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 20813
    },
    {
      "epoch": 4.868670330955444,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 20814
    },
    {
      "epoch": 4.868904221728453,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 20815
    },
    {
      "epoch": 4.869138112501462,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 20816
    },
    {
      "epoch": 4.869372003274471,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 20817
    },
    {
      "epoch": 4.86960589404748,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5813,
      "step": 20818
    },
    {
      "epoch": 4.869839784820488,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5926,
      "step": 20819
    },
    {
      "epoch": 4.870073675593498,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 20820
    },
    {
      "epoch": 4.870307566366507,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 20821
    },
    {
      "epoch": 4.870541457139516,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 20822
    },
    {
      "epoch": 4.870775347912525,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 20823
    },
    {
      "epoch": 4.8710092386855335,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 20824
    },
    {
      "epoch": 4.871243129458543,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.368,
      "step": 20825
    },
    {
      "epoch": 4.871477020231552,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 20826
    },
    {
      "epoch": 4.871710911004561,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.41,
      "step": 20827
    },
    {
      "epoch": 4.87194480177757,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 20828
    },
    {
      "epoch": 4.872178692550579,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 20829
    },
    {
      "epoch": 4.872412583323587,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6179,
      "step": 20830
    },
    {
      "epoch": 4.872646474096597,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 20831
    },
    {
      "epoch": 4.872880364869606,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 20832
    },
    {
      "epoch": 4.873114255642615,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 20833
    },
    {
      "epoch": 4.873348146415624,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 20834
    },
    {
      "epoch": 4.873582037188633,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 20835
    },
    {
      "epoch": 4.873815927961642,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 20836
    },
    {
      "epoch": 4.874049818734651,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 20837
    },
    {
      "epoch": 4.87428370950766,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 20838
    },
    {
      "epoch": 4.874517600280669,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 20839
    },
    {
      "epoch": 4.8747514910536776,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 20840
    },
    {
      "epoch": 4.874985381826687,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6727,
      "step": 20841
    },
    {
      "epoch": 4.875219272599696,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.5913,
      "step": 20842
    },
    {
      "epoch": 4.875453163372705,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 20843
    },
    {
      "epoch": 4.875687054145714,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 20844
    },
    {
      "epoch": 4.875920944918723,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 20845
    },
    {
      "epoch": 4.876154835691732,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 20846
    },
    {
      "epoch": 4.876388726464741,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 20847
    },
    {
      "epoch": 4.87662261723775,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9155,
      "step": 20848
    },
    {
      "epoch": 4.876856508010759,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 20849
    },
    {
      "epoch": 4.877090398783768,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 20850
    },
    {
      "epoch": 4.8773242895567765,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 20851
    },
    {
      "epoch": 4.877558180329786,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7153,
      "step": 20852
    },
    {
      "epoch": 4.877792071102795,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 20853
    },
    {
      "epoch": 4.878025961875804,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 20854
    },
    {
      "epoch": 4.878259852648813,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 20855
    },
    {
      "epoch": 4.878493743421822,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5812,
      "step": 20856
    },
    {
      "epoch": 4.878727634194831,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 20857
    },
    {
      "epoch": 4.87896152496784,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 20858
    },
    {
      "epoch": 4.879195415740849,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 20859
    },
    {
      "epoch": 4.879429306513858,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7023,
      "step": 20860
    },
    {
      "epoch": 4.879663197286867,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6186,
      "step": 20861
    },
    {
      "epoch": 4.8798970880598755,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 20862
    },
    {
      "epoch": 4.880130978832885,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 20863
    },
    {
      "epoch": 4.880364869605894,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 20864
    },
    {
      "epoch": 4.880598760378903,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 20865
    },
    {
      "epoch": 4.880832651151912,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6444,
      "step": 20866
    },
    {
      "epoch": 4.8810665419249215,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 20867
    },
    {
      "epoch": 4.88130043269793,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6826,
      "step": 20868
    },
    {
      "epoch": 4.881534323470939,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 20869
    },
    {
      "epoch": 4.881768214243948,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 20870
    },
    {
      "epoch": 4.882002105016957,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 20871
    },
    {
      "epoch": 4.882235995789966,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 20872
    },
    {
      "epoch": 4.882469886562975,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3137,
      "step": 20873
    },
    {
      "epoch": 4.882703777335984,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 20874
    },
    {
      "epoch": 4.882937668108993,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 20875
    },
    {
      "epoch": 4.883171558882002,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 20876
    },
    {
      "epoch": 4.883405449655011,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 20877
    },
    {
      "epoch": 4.8836393404280205,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3229,
      "step": 20878
    },
    {
      "epoch": 4.883873231201029,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 20879
    },
    {
      "epoch": 4.884107121974038,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 20880
    },
    {
      "epoch": 4.884341012747047,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 20881
    },
    {
      "epoch": 4.884574903520056,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7946,
      "step": 20882
    },
    {
      "epoch": 4.884808794293065,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 20883
    },
    {
      "epoch": 4.885042685066074,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.504,
      "step": 20884
    },
    {
      "epoch": 4.885276575839083,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 20885
    },
    {
      "epoch": 4.885510466612092,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 20886
    },
    {
      "epoch": 4.885744357385101,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6231,
      "step": 20887
    },
    {
      "epoch": 4.885978248158111,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6999,
      "step": 20888
    },
    {
      "epoch": 4.8862121389311195,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 20889
    },
    {
      "epoch": 4.886446029704128,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 20890
    },
    {
      "epoch": 4.886679920477137,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 20891
    },
    {
      "epoch": 4.886913811250146,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 20892
    },
    {
      "epoch": 4.887147702023155,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 20893
    },
    {
      "epoch": 4.887381592796165,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.784,
      "step": 20894
    },
    {
      "epoch": 4.887615483569173,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6178,
      "step": 20895
    },
    {
      "epoch": 4.887849374342182,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6586,
      "step": 20896
    },
    {
      "epoch": 4.888083265115191,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 20897
    },
    {
      "epoch": 4.8883171558882,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.76,
      "step": 20898
    },
    {
      "epoch": 4.88855104666121,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7504,
      "step": 20899
    },
    {
      "epoch": 4.8887849374342185,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 20900
    },
    {
      "epoch": 4.8887849374342185,
      "eval_runtime": 4.6207,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 20900
    },
    {
      "epoch": 4.889018828207227,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 20901
    },
    {
      "epoch": 4.889252718980236,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 20902
    },
    {
      "epoch": 4.889486609753245,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6042,
      "step": 20903
    },
    {
      "epoch": 4.889720500526254,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 20904
    },
    {
      "epoch": 4.889954391299264,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6626,
      "step": 20905
    },
    {
      "epoch": 4.890188282072272,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 20906
    },
    {
      "epoch": 4.890422172845281,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 20907
    },
    {
      "epoch": 4.89065606361829,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 20908
    },
    {
      "epoch": 4.890889954391299,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 20909
    },
    {
      "epoch": 4.891123845164309,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 20910
    },
    {
      "epoch": 4.8913577359373175,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 20911
    },
    {
      "epoch": 4.891591626710326,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 20912
    },
    {
      "epoch": 4.891825517483335,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 20913
    },
    {
      "epoch": 4.892059408256344,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 20914
    },
    {
      "epoch": 4.892293299029353,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 20915
    },
    {
      "epoch": 4.892527189802363,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 20916
    },
    {
      "epoch": 4.892761080575371,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 20917
    },
    {
      "epoch": 4.89299497134838,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 20918
    },
    {
      "epoch": 4.893228862121389,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 20919
    },
    {
      "epoch": 4.893462752894399,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 20920
    },
    {
      "epoch": 4.893696643667408,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 20921
    },
    {
      "epoch": 4.8939305344404165,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 20922
    },
    {
      "epoch": 4.894164425213425,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 20923
    },
    {
      "epoch": 4.894398315986434,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 20924
    },
    {
      "epoch": 4.894632206759443,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 20925
    },
    {
      "epoch": 4.894866097532453,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 20926
    },
    {
      "epoch": 4.895099988305462,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4526,
      "step": 20927
    },
    {
      "epoch": 4.89533387907847,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 20928
    },
    {
      "epoch": 4.895567769851479,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 20929
    },
    {
      "epoch": 4.895801660624488,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 20930
    },
    {
      "epoch": 4.896035551397498,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 20931
    },
    {
      "epoch": 4.896269442170507,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 20932
    },
    {
      "epoch": 4.8965033329435155,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 20933
    },
    {
      "epoch": 4.896737223716524,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 20934
    },
    {
      "epoch": 4.896971114489533,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 20935
    },
    {
      "epoch": 4.897205005262542,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6609,
      "step": 20936
    },
    {
      "epoch": 4.897438896035552,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 20937
    },
    {
      "epoch": 4.8976727868085606,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 20938
    },
    {
      "epoch": 4.897906677581569,
      "grad_norm": 10.9375,
      "learning_rate": 3e-05,
      "loss": 2.3121,
      "step": 20939
    },
    {
      "epoch": 4.898140568354578,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 20940
    },
    {
      "epoch": 4.898374459127587,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 20941
    },
    {
      "epoch": 4.898608349900597,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.656,
      "step": 20942
    },
    {
      "epoch": 4.898842240673606,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 20943
    },
    {
      "epoch": 4.8990761314466145,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6486,
      "step": 20944
    },
    {
      "epoch": 4.899310022219623,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 20945
    },
    {
      "epoch": 4.899543912992632,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8891,
      "step": 20946
    },
    {
      "epoch": 4.899777803765641,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 20947
    },
    {
      "epoch": 4.900011694538651,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 20948
    },
    {
      "epoch": 4.9002455853116595,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5577,
      "step": 20949
    },
    {
      "epoch": 4.900479476084668,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.748,
      "step": 20950
    },
    {
      "epoch": 4.900713366857677,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 20951
    },
    {
      "epoch": 4.900947257630687,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6356,
      "step": 20952
    },
    {
      "epoch": 4.901181148403696,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 20953
    },
    {
      "epoch": 4.901415039176705,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 20954
    },
    {
      "epoch": 4.9016489299497135,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 20955
    },
    {
      "epoch": 4.901882820722722,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 20956
    },
    {
      "epoch": 4.902116711495731,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5751,
      "step": 20957
    },
    {
      "epoch": 4.902350602268741,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 20958
    },
    {
      "epoch": 4.90258449304175,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7656,
      "step": 20959
    },
    {
      "epoch": 4.9028183838147585,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 20960
    },
    {
      "epoch": 4.903052274587767,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 20961
    },
    {
      "epoch": 4.903286165360776,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 20962
    },
    {
      "epoch": 4.903520056133786,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 20963
    },
    {
      "epoch": 4.903753946906795,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 20964
    },
    {
      "epoch": 4.903987837679804,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 20965
    },
    {
      "epoch": 4.9042217284528125,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 20966
    },
    {
      "epoch": 4.904455619225821,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8322,
      "step": 20967
    },
    {
      "epoch": 4.90468950999883,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 20968
    },
    {
      "epoch": 4.90492340077184,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4625,
      "step": 20969
    },
    {
      "epoch": 4.905157291544849,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7317,
      "step": 20970
    },
    {
      "epoch": 4.9053911823178575,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 20971
    },
    {
      "epoch": 4.905625073090866,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6347,
      "step": 20972
    },
    {
      "epoch": 4.905858963863875,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9152,
      "step": 20973
    },
    {
      "epoch": 4.906092854636885,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 20974
    },
    {
      "epoch": 4.906326745409894,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 20975
    },
    {
      "epoch": 4.906560636182903,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 20976
    },
    {
      "epoch": 4.9067945269559115,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 20977
    },
    {
      "epoch": 4.90702841772892,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 20978
    },
    {
      "epoch": 4.907262308501929,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7345,
      "step": 20979
    },
    {
      "epoch": 4.907496199274939,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 20980
    },
    {
      "epoch": 4.907730090047948,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6736,
      "step": 20981
    },
    {
      "epoch": 4.9079639808209565,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 20982
    },
    {
      "epoch": 4.908197871593965,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 20983
    },
    {
      "epoch": 4.908431762366975,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 20984
    },
    {
      "epoch": 4.908665653139984,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7525,
      "step": 20985
    },
    {
      "epoch": 4.908899543912993,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 20986
    },
    {
      "epoch": 4.909133434686002,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 20987
    },
    {
      "epoch": 4.9093673254590104,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.61,
      "step": 20988
    },
    {
      "epoch": 4.909601216232019,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5722,
      "step": 20989
    },
    {
      "epoch": 4.909835107005029,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6003,
      "step": 20990
    },
    {
      "epoch": 4.910068997778038,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 20991
    },
    {
      "epoch": 4.910302888551047,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 20992
    },
    {
      "epoch": 4.9105367793240555,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 20993
    },
    {
      "epoch": 4.910770670097064,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2589,
      "step": 20994
    },
    {
      "epoch": 4.911004560870074,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.251,
      "step": 20995
    },
    {
      "epoch": 4.911238451643083,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 20996
    },
    {
      "epoch": 4.911472342416092,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 20997
    },
    {
      "epoch": 4.911706233189101,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 20998
    },
    {
      "epoch": 4.911940123962109,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 20999
    },
    {
      "epoch": 4.912174014735118,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 21000
    },
    {
      "epoch": 4.912174014735118,
      "eval_runtime": 4.5786,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 21000
    },
    {
      "epoch": 4.912407905508128,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 21001
    },
    {
      "epoch": 4.912641796281137,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 21002
    },
    {
      "epoch": 4.912875687054146,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 21003
    },
    {
      "epoch": 4.9131095778271545,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 21004
    },
    {
      "epoch": 4.913343468600163,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 21005
    },
    {
      "epoch": 4.913577359373173,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 21006
    },
    {
      "epoch": 4.913811250146182,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 21007
    },
    {
      "epoch": 4.914045140919191,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 21008
    },
    {
      "epoch": 4.9142790316922,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7058,
      "step": 21009
    },
    {
      "epoch": 4.914512922465208,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 21010
    },
    {
      "epoch": 4.914746813238217,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 21011
    },
    {
      "epoch": 4.914980704011227,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 21012
    },
    {
      "epoch": 4.915214594784236,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8304,
      "step": 21013
    },
    {
      "epoch": 4.915448485557245,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 21014
    },
    {
      "epoch": 4.9156823763302535,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.5936,
      "step": 21015
    },
    {
      "epoch": 4.915916267103263,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 21016
    },
    {
      "epoch": 4.916150157876272,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7456,
      "step": 21017
    },
    {
      "epoch": 4.916384048649281,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 21018
    },
    {
      "epoch": 4.91661793942229,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 21019
    },
    {
      "epoch": 4.916851830195299,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6162,
      "step": 21020
    },
    {
      "epoch": 4.917085720968307,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 21021
    },
    {
      "epoch": 4.917319611741317,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8708,
      "step": 21022
    },
    {
      "epoch": 4.917553502514326,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7039,
      "step": 21023
    },
    {
      "epoch": 4.917787393287335,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 21024
    },
    {
      "epoch": 4.918021284060344,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 21025
    },
    {
      "epoch": 4.9182551748333525,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 21026
    },
    {
      "epoch": 4.918489065606362,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 21027
    },
    {
      "epoch": 4.918722956379371,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 21028
    },
    {
      "epoch": 4.91895684715238,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 21029
    },
    {
      "epoch": 4.919190737925389,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 21030
    },
    {
      "epoch": 4.919424628698398,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 21031
    },
    {
      "epoch": 4.919658519471406,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2759,
      "step": 21032
    },
    {
      "epoch": 4.919892410244416,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 21033
    },
    {
      "epoch": 4.920126301017425,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 21034
    },
    {
      "epoch": 4.920360191790434,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 21035
    },
    {
      "epoch": 4.920594082563443,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 21036
    },
    {
      "epoch": 4.9208279733364515,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 21037
    },
    {
      "epoch": 4.921061864109461,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.5167,
      "step": 21038
    },
    {
      "epoch": 4.92129575488247,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 21039
    },
    {
      "epoch": 4.921529645655479,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 21040
    },
    {
      "epoch": 4.921763536428488,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 21041
    },
    {
      "epoch": 4.921997427201497,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7116,
      "step": 21042
    },
    {
      "epoch": 4.922231317974505,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 21043
    },
    {
      "epoch": 4.922465208747515,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 21044
    },
    {
      "epoch": 4.922699099520524,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7573,
      "step": 21045
    },
    {
      "epoch": 4.922932990293533,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.3214,
      "step": 21046
    },
    {
      "epoch": 4.923166881066542,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6389,
      "step": 21047
    },
    {
      "epoch": 4.923400771839551,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 21048
    },
    {
      "epoch": 4.92363466261256,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8011,
      "step": 21049
    },
    {
      "epoch": 4.923868553385569,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6681,
      "step": 21050
    },
    {
      "epoch": 4.924102444158578,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7568,
      "step": 21051
    },
    {
      "epoch": 4.924336334931587,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 21052
    },
    {
      "epoch": 4.924570225704596,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 21053
    },
    {
      "epoch": 4.924804116477605,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7082,
      "step": 21054
    },
    {
      "epoch": 4.925038007250614,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 21055
    },
    {
      "epoch": 4.925271898023623,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 21056
    },
    {
      "epoch": 4.925505788796632,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 21057
    },
    {
      "epoch": 4.925739679569641,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5264,
      "step": 21058
    },
    {
      "epoch": 4.92597357034265,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 21059
    },
    {
      "epoch": 4.926207461115659,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5635,
      "step": 21060
    },
    {
      "epoch": 4.926441351888668,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5412,
      "step": 21061
    },
    {
      "epoch": 4.926675242661677,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 21062
    },
    {
      "epoch": 4.926909133434686,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 21063
    },
    {
      "epoch": 4.927143024207695,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 21064
    },
    {
      "epoch": 4.927376914980704,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5738,
      "step": 21065
    },
    {
      "epoch": 4.927610805753713,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 21066
    },
    {
      "epoch": 4.927844696526722,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5897,
      "step": 21067
    },
    {
      "epoch": 4.928078587299731,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0509,
      "step": 21068
    },
    {
      "epoch": 4.92831247807274,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 21069
    },
    {
      "epoch": 4.928546368845749,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 21070
    },
    {
      "epoch": 4.928780259618758,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5823,
      "step": 21071
    },
    {
      "epoch": 4.929014150391767,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 21072
    },
    {
      "epoch": 4.929248041164776,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 21073
    },
    {
      "epoch": 4.929481931937785,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 21074
    },
    {
      "epoch": 4.929715822710794,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.589,
      "step": 21075
    },
    {
      "epoch": 4.929949713483803,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 21076
    },
    {
      "epoch": 4.930183604256812,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 21077
    },
    {
      "epoch": 4.930417495029821,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.76,
      "step": 21078
    },
    {
      "epoch": 4.93065138580283,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 21079
    },
    {
      "epoch": 4.9308852765758395,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 21080
    },
    {
      "epoch": 4.931119167348848,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8415,
      "step": 21081
    },
    {
      "epoch": 4.931353058121857,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6561,
      "step": 21082
    },
    {
      "epoch": 4.931586948894866,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 21083
    },
    {
      "epoch": 4.931820839667875,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.4667,
      "step": 21084
    },
    {
      "epoch": 4.932054730440884,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 21085
    },
    {
      "epoch": 4.932288621213893,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 21086
    },
    {
      "epoch": 4.932522511986902,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 21087
    },
    {
      "epoch": 4.932756402759911,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6203,
      "step": 21088
    },
    {
      "epoch": 4.93299029353292,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 21089
    },
    {
      "epoch": 4.933224184305929,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 21090
    },
    {
      "epoch": 4.9334580750789385,
      "grad_norm": 8.625,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 21091
    },
    {
      "epoch": 4.933691965851947,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8532,
      "step": 21092
    },
    {
      "epoch": 4.933925856624956,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 21093
    },
    {
      "epoch": 4.934159747397965,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 21094
    },
    {
      "epoch": 4.934393638170974,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.3171,
      "step": 21095
    },
    {
      "epoch": 4.934627528943983,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 21096
    },
    {
      "epoch": 4.934861419716992,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 21097
    },
    {
      "epoch": 4.935095310490001,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 21098
    },
    {
      "epoch": 4.93532920126301,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 21099
    },
    {
      "epoch": 4.935563092036019,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6384,
      "step": 21100
    },
    {
      "epoch": 4.935563092036019,
      "eval_runtime": 4.6239,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 21100
    },
    {
      "epoch": 4.935796982809029,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 21101
    },
    {
      "epoch": 4.9360308735820375,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 21102
    },
    {
      "epoch": 4.936264764355046,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.708,
      "step": 21103
    },
    {
      "epoch": 4.936498655128055,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 21104
    },
    {
      "epoch": 4.936732545901064,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 21105
    },
    {
      "epoch": 4.936966436674073,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 21106
    },
    {
      "epoch": 4.937200327447083,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 21107
    },
    {
      "epoch": 4.937434218220091,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.4995,
      "step": 21108
    },
    {
      "epoch": 4.9376681089931,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7088,
      "step": 21109
    },
    {
      "epoch": 4.937901999766109,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6889,
      "step": 21110
    },
    {
      "epoch": 4.938135890539118,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 21111
    },
    {
      "epoch": 4.938369781312128,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5646,
      "step": 21112
    },
    {
      "epoch": 4.9386036720851365,
      "grad_norm": 10.3125,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 21113
    },
    {
      "epoch": 4.938837562858145,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6713,
      "step": 21114
    },
    {
      "epoch": 4.939071453631154,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 21115
    },
    {
      "epoch": 4.939305344404163,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6572,
      "step": 21116
    },
    {
      "epoch": 4.939539235177172,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 21117
    },
    {
      "epoch": 4.939773125950182,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 21118
    },
    {
      "epoch": 4.94000701672319,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 21119
    },
    {
      "epoch": 4.940240907496199,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 21120
    },
    {
      "epoch": 4.940474798269208,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6415,
      "step": 21121
    },
    {
      "epoch": 4.940708689042217,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 21122
    },
    {
      "epoch": 4.940942579815227,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 21123
    },
    {
      "epoch": 4.9411764705882355,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 21124
    },
    {
      "epoch": 4.941410361361244,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 21125
    },
    {
      "epoch": 4.941644252134253,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 21126
    },
    {
      "epoch": 4.941878142907262,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 21127
    },
    {
      "epoch": 4.942112033680271,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 21128
    },
    {
      "epoch": 4.942345924453281,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5928,
      "step": 21129
    },
    {
      "epoch": 4.942579815226289,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 21130
    },
    {
      "epoch": 4.942813705999298,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 21131
    },
    {
      "epoch": 4.943047596772307,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8701,
      "step": 21132
    },
    {
      "epoch": 4.943281487545317,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 21133
    },
    {
      "epoch": 4.943515378318326,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 21134
    },
    {
      "epoch": 4.9437492690913345,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9912,
      "step": 21135
    },
    {
      "epoch": 4.943983159864343,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 21136
    },
    {
      "epoch": 4.944217050637352,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 21137
    },
    {
      "epoch": 4.944450941410361,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 21138
    },
    {
      "epoch": 4.944684832183371,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6733,
      "step": 21139
    },
    {
      "epoch": 4.94491872295638,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 21140
    },
    {
      "epoch": 4.945152613729388,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 21141
    },
    {
      "epoch": 4.945386504502397,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6902,
      "step": 21142
    },
    {
      "epoch": 4.945620395275406,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 21143
    },
    {
      "epoch": 4.945854286048416,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.775,
      "step": 21144
    },
    {
      "epoch": 4.946088176821425,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 21145
    },
    {
      "epoch": 4.9463220675944335,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 21146
    },
    {
      "epoch": 4.946555958367442,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 21147
    },
    {
      "epoch": 4.946789849140451,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 21148
    },
    {
      "epoch": 4.94702373991346,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5355,
      "step": 21149
    },
    {
      "epoch": 4.94725763068647,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 21150
    },
    {
      "epoch": 4.947491521459479,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 21151
    },
    {
      "epoch": 4.947725412232487,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 21152
    },
    {
      "epoch": 4.947959303005496,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2623,
      "step": 21153
    },
    {
      "epoch": 4.948193193778505,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 21154
    },
    {
      "epoch": 4.948427084551515,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 21155
    },
    {
      "epoch": 4.948660975324524,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 21156
    },
    {
      "epoch": 4.9488948660975325,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 21157
    },
    {
      "epoch": 4.949128756870541,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6643,
      "step": 21158
    },
    {
      "epoch": 4.94936264764355,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6621,
      "step": 21159
    },
    {
      "epoch": 4.949596538416559,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.037,
      "step": 21160
    },
    {
      "epoch": 4.949830429189569,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 21161
    },
    {
      "epoch": 4.950064319962578,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 21162
    },
    {
      "epoch": 4.950298210735586,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 21163
    },
    {
      "epoch": 4.950532101508595,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 21164
    },
    {
      "epoch": 4.950765992281605,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6335,
      "step": 21165
    },
    {
      "epoch": 4.950999883054614,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 21166
    },
    {
      "epoch": 4.951233773827623,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 21167
    },
    {
      "epoch": 4.9514676646006315,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 21168
    },
    {
      "epoch": 4.95170155537364,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 21169
    },
    {
      "epoch": 4.951935446146649,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 21170
    },
    {
      "epoch": 4.952169336919659,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 21171
    },
    {
      "epoch": 4.952403227692668,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 21172
    },
    {
      "epoch": 4.952637118465677,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6089,
      "step": 21173
    },
    {
      "epoch": 4.952871009238685,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6909,
      "step": 21174
    },
    {
      "epoch": 4.953104900011694,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.617,
      "step": 21175
    },
    {
      "epoch": 4.953338790784704,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 21176
    },
    {
      "epoch": 4.953572681557713,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 21177
    },
    {
      "epoch": 4.953806572330722,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 21178
    },
    {
      "epoch": 4.9540404631037305,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 21179
    },
    {
      "epoch": 4.954274353876739,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 21180
    },
    {
      "epoch": 4.954508244649748,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 21181
    },
    {
      "epoch": 4.954742135422758,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 21182
    },
    {
      "epoch": 4.954976026195767,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7467,
      "step": 21183
    },
    {
      "epoch": 4.9552099169687756,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 21184
    },
    {
      "epoch": 4.955443807741784,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7154,
      "step": 21185
    },
    {
      "epoch": 4.955677698514793,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 21186
    },
    {
      "epoch": 4.955911589287803,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 21187
    },
    {
      "epoch": 4.956145480060812,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6597,
      "step": 21188
    },
    {
      "epoch": 4.956379370833821,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 21189
    },
    {
      "epoch": 4.9566132616068295,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 21190
    },
    {
      "epoch": 4.956847152379838,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5755,
      "step": 21191
    },
    {
      "epoch": 4.957081043152847,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 21192
    },
    {
      "epoch": 4.957314933925857,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 21193
    },
    {
      "epoch": 4.957548824698866,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 21194
    },
    {
      "epoch": 4.9577827154718745,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 21195
    },
    {
      "epoch": 4.958016606244883,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 21196
    },
    {
      "epoch": 4.958250497017893,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 21197
    },
    {
      "epoch": 4.958484387790902,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 21198
    },
    {
      "epoch": 4.958718278563911,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 21199
    },
    {
      "epoch": 4.95895216933692,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8366,
      "step": 21200
    },
    {
      "epoch": 4.95895216933692,
      "eval_runtime": 4.6275,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 21200
    },
    {
      "epoch": 4.9591860601099285,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.832,
      "step": 21201
    },
    {
      "epoch": 4.959419950882937,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 21202
    },
    {
      "epoch": 4.959653841655947,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 21203
    },
    {
      "epoch": 4.959887732428956,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 21204
    },
    {
      "epoch": 4.960121623201965,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 21205
    },
    {
      "epoch": 4.9603555139749735,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 21206
    },
    {
      "epoch": 4.960589404747982,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 21207
    },
    {
      "epoch": 4.960823295520992,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 21208
    },
    {
      "epoch": 4.961057186294001,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 21209
    },
    {
      "epoch": 4.96129107706701,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7013,
      "step": 21210
    },
    {
      "epoch": 4.961524967840019,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5405,
      "step": 21211
    },
    {
      "epoch": 4.9617588586130275,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 21212
    },
    {
      "epoch": 4.961992749386036,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 21213
    },
    {
      "epoch": 4.962226640159046,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 21214
    },
    {
      "epoch": 4.962460530932055,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 21215
    },
    {
      "epoch": 4.962694421705064,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 21216
    },
    {
      "epoch": 4.9629283124780725,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 21217
    },
    {
      "epoch": 4.963162203251081,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 21218
    },
    {
      "epoch": 4.963396094024091,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 21219
    },
    {
      "epoch": 4.9636299847971,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 21220
    },
    {
      "epoch": 4.963863875570109,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 21221
    },
    {
      "epoch": 4.964097766343118,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6498,
      "step": 21222
    },
    {
      "epoch": 4.9643316571161265,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 21223
    },
    {
      "epoch": 4.964565547889135,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 21224
    },
    {
      "epoch": 4.964799438662145,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 21225
    },
    {
      "epoch": 4.965033329435154,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 21226
    },
    {
      "epoch": 4.965267220208163,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 21227
    },
    {
      "epoch": 4.9655011109811715,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 21228
    },
    {
      "epoch": 4.965735001754181,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 21229
    },
    {
      "epoch": 4.96596889252719,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 21230
    },
    {
      "epoch": 4.966202783300199,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7527,
      "step": 21231
    },
    {
      "epoch": 4.966436674073208,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 21232
    },
    {
      "epoch": 4.966670564846217,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 21233
    },
    {
      "epoch": 4.9669044556192254,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 21234
    },
    {
      "epoch": 4.967138346392235,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 21235
    },
    {
      "epoch": 4.967372237165244,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 21236
    },
    {
      "epoch": 4.967606127938253,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8153,
      "step": 21237
    },
    {
      "epoch": 4.967840018711262,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5324,
      "step": 21238
    },
    {
      "epoch": 4.9680739094842705,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 21239
    },
    {
      "epoch": 4.96830780025728,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 21240
    },
    {
      "epoch": 4.968541691030289,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 21241
    },
    {
      "epoch": 4.968775581803298,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 21242
    },
    {
      "epoch": 4.969009472576307,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.668,
      "step": 21243
    },
    {
      "epoch": 4.969243363349316,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7432,
      "step": 21244
    },
    {
      "epoch": 4.969477254122324,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 21245
    },
    {
      "epoch": 4.969711144895334,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6225,
      "step": 21246
    },
    {
      "epoch": 4.969945035668343,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 21247
    },
    {
      "epoch": 4.970178926441352,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.748,
      "step": 21248
    },
    {
      "epoch": 4.970412817214361,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 21249
    },
    {
      "epoch": 4.9706467079873695,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 21250
    },
    {
      "epoch": 4.970880598760379,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 21251
    },
    {
      "epoch": 4.971114489533388,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 21252
    },
    {
      "epoch": 4.971348380306397,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 21253
    },
    {
      "epoch": 4.971582271079406,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 21254
    },
    {
      "epoch": 4.971816161852415,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6107,
      "step": 21255
    },
    {
      "epoch": 4.972050052625423,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 21256
    },
    {
      "epoch": 4.972283943398433,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7664,
      "step": 21257
    },
    {
      "epoch": 4.972517834171442,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 21258
    },
    {
      "epoch": 4.972751724944451,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 21259
    },
    {
      "epoch": 4.97298561571746,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 21260
    },
    {
      "epoch": 4.973219506490469,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9432,
      "step": 21261
    },
    {
      "epoch": 4.973453397263478,
      "grad_norm": 2.484375,
      "learning_rate": 3e-05,
      "loss": 1.4598,
      "step": 21262
    },
    {
      "epoch": 4.973687288036487,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 21263
    },
    {
      "epoch": 4.973921178809496,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 21264
    },
    {
      "epoch": 4.974155069582505,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 21265
    },
    {
      "epoch": 4.974388960355514,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 21266
    },
    {
      "epoch": 4.974622851128523,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 21267
    },
    {
      "epoch": 4.974856741901532,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 21268
    },
    {
      "epoch": 4.975090632674541,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 21269
    },
    {
      "epoch": 4.97532452344755,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 21270
    },
    {
      "epoch": 4.975558414220559,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 21271
    },
    {
      "epoch": 4.975792304993568,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 21272
    },
    {
      "epoch": 4.976026195766577,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 21273
    },
    {
      "epoch": 4.976260086539586,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.784,
      "step": 21274
    },
    {
      "epoch": 4.976493977312595,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 21275
    },
    {
      "epoch": 4.976727868085604,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5713,
      "step": 21276
    },
    {
      "epoch": 4.976961758858613,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3525,
      "step": 21277
    },
    {
      "epoch": 4.977195649631622,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 21278
    },
    {
      "epoch": 4.977429540404631,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 21279
    },
    {
      "epoch": 4.97766343117764,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 21280
    },
    {
      "epoch": 4.977897321950649,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 21281
    },
    {
      "epoch": 4.9781312127236585,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7275,
      "step": 21282
    },
    {
      "epoch": 4.978365103496667,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 21283
    },
    {
      "epoch": 4.978598994269676,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 21284
    },
    {
      "epoch": 4.978832885042685,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6457,
      "step": 21285
    },
    {
      "epoch": 4.979066775815694,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 21286
    },
    {
      "epoch": 4.979300666588703,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 21287
    },
    {
      "epoch": 4.9795345573617125,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 21288
    },
    {
      "epoch": 4.979768448134721,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 21289
    },
    {
      "epoch": 4.98000233890773,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 21290
    },
    {
      "epoch": 4.980236229680739,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 21291
    },
    {
      "epoch": 4.980470120453748,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8047,
      "step": 21292
    },
    {
      "epoch": 4.9807040112267575,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 21293
    },
    {
      "epoch": 4.980937901999766,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 21294
    },
    {
      "epoch": 4.981171792772775,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7499,
      "step": 21295
    },
    {
      "epoch": 4.981405683545784,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 21296
    },
    {
      "epoch": 4.981639574318793,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 21297
    },
    {
      "epoch": 4.981873465091802,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5811,
      "step": 21298
    },
    {
      "epoch": 4.9821073558648115,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6897,
      "step": 21299
    },
    {
      "epoch": 4.98234124663782,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 21300
    },
    {
      "epoch": 4.98234124663782,
      "eval_runtime": 4.6097,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 21300
    },
    {
      "epoch": 4.982575137410829,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 21301
    },
    {
      "epoch": 4.982809028183838,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 21302
    },
    {
      "epoch": 4.983042918956847,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 21303
    },
    {
      "epoch": 4.9832768097298565,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 21304
    },
    {
      "epoch": 4.983510700502865,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 21305
    },
    {
      "epoch": 4.983744591275874,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 21306
    },
    {
      "epoch": 4.983978482048883,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 21307
    },
    {
      "epoch": 4.984212372821892,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6411,
      "step": 21308
    },
    {
      "epoch": 4.984446263594901,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3204,
      "step": 21309
    },
    {
      "epoch": 4.9846801543679105,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 21310
    },
    {
      "epoch": 4.984914045140919,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 21311
    },
    {
      "epoch": 4.985147935913928,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5236,
      "step": 21312
    },
    {
      "epoch": 4.985381826686937,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 21313
    },
    {
      "epoch": 4.985615717459947,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 21314
    },
    {
      "epoch": 4.9858496082329555,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 21315
    },
    {
      "epoch": 4.986083499005964,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5809,
      "step": 21316
    },
    {
      "epoch": 4.986317389778973,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2277,
      "step": 21317
    },
    {
      "epoch": 4.986551280551982,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 21318
    },
    {
      "epoch": 4.986785171324991,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7082,
      "step": 21319
    },
    {
      "epoch": 4.987019062098001,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6941,
      "step": 21320
    },
    {
      "epoch": 4.9872529528710094,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 21321
    },
    {
      "epoch": 4.987486843644018,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 21322
    },
    {
      "epoch": 4.987720734417027,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 21323
    },
    {
      "epoch": 4.987954625190036,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 21324
    },
    {
      "epoch": 4.988188515963046,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 21325
    },
    {
      "epoch": 4.9884224067360545,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 21326
    },
    {
      "epoch": 4.988656297509063,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 21327
    },
    {
      "epoch": 4.988890188282072,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 21328
    },
    {
      "epoch": 4.989124079055081,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6256,
      "step": 21329
    },
    {
      "epoch": 4.98935796982809,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6844,
      "step": 21330
    },
    {
      "epoch": 4.9895918606011,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 21331
    },
    {
      "epoch": 4.989825751374108,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 21332
    },
    {
      "epoch": 4.990059642147117,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7308,
      "step": 21333
    },
    {
      "epoch": 4.990293532920126,
      "grad_norm": 9.3125,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 21334
    },
    {
      "epoch": 4.990527423693135,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 21335
    },
    {
      "epoch": 4.990761314466145,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 21336
    },
    {
      "epoch": 4.9909952052391535,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 21337
    },
    {
      "epoch": 4.991229096012162,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5806,
      "step": 21338
    },
    {
      "epoch": 4.991462986785171,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3689,
      "step": 21339
    },
    {
      "epoch": 4.99169687755818,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 21340
    },
    {
      "epoch": 4.991930768331189,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5187,
      "step": 21341
    },
    {
      "epoch": 4.992164659104199,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 21342
    },
    {
      "epoch": 4.992398549877207,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7321,
      "step": 21343
    },
    {
      "epoch": 4.992632440650216,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.677,
      "step": 21344
    },
    {
      "epoch": 4.992866331423225,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5328,
      "step": 21345
    },
    {
      "epoch": 4.993100222196235,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7093,
      "step": 21346
    },
    {
      "epoch": 4.993334112969244,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 21347
    },
    {
      "epoch": 4.9935680037422525,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 21348
    },
    {
      "epoch": 4.993801894515261,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 21349
    },
    {
      "epoch": 4.99403578528827,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5741,
      "step": 21350
    },
    {
      "epoch": 4.994269676061279,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 21351
    },
    {
      "epoch": 4.994503566834289,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6323,
      "step": 21352
    },
    {
      "epoch": 4.994737457607298,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6657,
      "step": 21353
    },
    {
      "epoch": 4.994971348380306,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 21354
    },
    {
      "epoch": 4.995205239153315,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 21355
    },
    {
      "epoch": 4.995439129926324,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 21356
    },
    {
      "epoch": 4.995673020699334,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 21357
    },
    {
      "epoch": 4.995906911472343,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 21358
    },
    {
      "epoch": 4.9961408022453515,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.637,
      "step": 21359
    },
    {
      "epoch": 4.99637469301836,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 21360
    },
    {
      "epoch": 4.996608583791369,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 21361
    },
    {
      "epoch": 4.996842474564378,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 21362
    },
    {
      "epoch": 4.997076365337388,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 21363
    },
    {
      "epoch": 4.997310256110397,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 21364
    },
    {
      "epoch": 4.997544146883405,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8748,
      "step": 21365
    },
    {
      "epoch": 4.997778037656414,
      "grad_norm": 8.375,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 21366
    },
    {
      "epoch": 4.998011928429423,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 21367
    },
    {
      "epoch": 4.998245819202433,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 21368
    },
    {
      "epoch": 4.998479709975442,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 21369
    },
    {
      "epoch": 4.9987136007484505,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 21370
    },
    {
      "epoch": 4.998947491521459,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 21371
    },
    {
      "epoch": 4.999181382294468,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 21372
    },
    {
      "epoch": 4.999415273067477,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 21373
    },
    {
      "epoch": 4.999649163840487,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 21374
    },
    {
      "epoch": 4.999883054613496,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 21375
    },
    {
      "epoch": 5.000233890773009,
      "grad_norm": 10.1875,
      "learning_rate": 3e-05,
      "loss": 4.2077,
      "step": 21376
    },
    {
      "epoch": 5.000467781546018,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7101,
      "step": 21377
    },
    {
      "epoch": 5.000701672319027,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 21378
    },
    {
      "epoch": 5.000935563092036,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.8025,
      "step": 21379
    },
    {
      "epoch": 5.001169453865045,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 21380
    },
    {
      "epoch": 5.001403344638054,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7375,
      "step": 21381
    },
    {
      "epoch": 5.001637235411063,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 21382
    },
    {
      "epoch": 5.0018711261840725,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 21383
    },
    {
      "epoch": 5.002105016957081,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 21384
    },
    {
      "epoch": 5.00233890773009,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 21385
    },
    {
      "epoch": 5.002572798503099,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 21386
    },
    {
      "epoch": 5.002806689276108,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 21387
    },
    {
      "epoch": 5.003040580049117,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 21388
    },
    {
      "epoch": 5.003274470822126,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7741,
      "step": 21389
    },
    {
      "epoch": 5.003508361595135,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.4928,
      "step": 21390
    },
    {
      "epoch": 5.003742252368144,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 21391
    },
    {
      "epoch": 5.003976143141153,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 21392
    },
    {
      "epoch": 5.004210033914162,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 21393
    },
    {
      "epoch": 5.0044439246871715,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6948,
      "step": 21394
    },
    {
      "epoch": 5.00467781546018,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7634,
      "step": 21395
    },
    {
      "epoch": 5.004911706233189,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 21396
    },
    {
      "epoch": 5.005145597006198,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6401,
      "step": 21397
    },
    {
      "epoch": 5.005379487779207,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4341,
      "step": 21398
    },
    {
      "epoch": 5.0056133785522166,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 21399
    },
    {
      "epoch": 5.005847269325225,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 21400
    },
    {
      "epoch": 5.005847269325225,
      "eval_runtime": 4.5899,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 21400
    },
    {
      "epoch": 5.006081160098234,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 21401
    },
    {
      "epoch": 5.006315050871243,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 21402
    },
    {
      "epoch": 5.006548941644252,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 21403
    },
    {
      "epoch": 5.006782832417261,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 21404
    },
    {
      "epoch": 5.0070167231902705,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5071,
      "step": 21405
    },
    {
      "epoch": 5.007250613963279,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.922,
      "step": 21406
    },
    {
      "epoch": 5.007484504736288,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7407,
      "step": 21407
    },
    {
      "epoch": 5.007718395509297,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7609,
      "step": 21408
    },
    {
      "epoch": 5.007952286282306,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 21409
    },
    {
      "epoch": 5.0081861770553155,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 21410
    },
    {
      "epoch": 5.008420067828324,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 21411
    },
    {
      "epoch": 5.008653958601333,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 21412
    },
    {
      "epoch": 5.008887849374342,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8617,
      "step": 21413
    },
    {
      "epoch": 5.009121740147351,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 21414
    },
    {
      "epoch": 5.009355630920361,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 21415
    },
    {
      "epoch": 5.0095895216933695,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7175,
      "step": 21416
    },
    {
      "epoch": 5.009823412466378,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 21417
    },
    {
      "epoch": 5.010057303239387,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6828,
      "step": 21418
    },
    {
      "epoch": 5.010291194012396,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4675,
      "step": 21419
    },
    {
      "epoch": 5.010525084785405,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6301,
      "step": 21420
    },
    {
      "epoch": 5.0107589755584145,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 21421
    },
    {
      "epoch": 5.010992866331423,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 21422
    },
    {
      "epoch": 5.011226757104432,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7846,
      "step": 21423
    },
    {
      "epoch": 5.011460647877441,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 21424
    },
    {
      "epoch": 5.01169453865045,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 21425
    },
    {
      "epoch": 5.01192842942346,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6829,
      "step": 21426
    },
    {
      "epoch": 5.0121623201964685,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 21427
    },
    {
      "epoch": 5.012396210969477,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 21428
    },
    {
      "epoch": 5.012630101742486,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 21429
    },
    {
      "epoch": 5.012863992515495,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 21430
    },
    {
      "epoch": 5.013097883288505,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 21431
    },
    {
      "epoch": 5.0133317740615135,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 21432
    },
    {
      "epoch": 5.013565664834522,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6752,
      "step": 21433
    },
    {
      "epoch": 5.013799555607531,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 21434
    },
    {
      "epoch": 5.01403344638054,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 21435
    },
    {
      "epoch": 5.014267337153549,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 21436
    },
    {
      "epoch": 5.014501227926559,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 21437
    },
    {
      "epoch": 5.0147351186995675,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6679,
      "step": 21438
    },
    {
      "epoch": 5.014969009472576,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8166,
      "step": 21439
    },
    {
      "epoch": 5.015202900245585,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 21440
    },
    {
      "epoch": 5.015436791018594,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9442,
      "step": 21441
    },
    {
      "epoch": 5.015670681791604,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 21442
    },
    {
      "epoch": 5.0159045725646125,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5992,
      "step": 21443
    },
    {
      "epoch": 5.016138463337621,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5417,
      "step": 21444
    },
    {
      "epoch": 5.01637235411063,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.4562,
      "step": 21445
    },
    {
      "epoch": 5.016606244883639,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 21446
    },
    {
      "epoch": 5.016840135656649,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 21447
    },
    {
      "epoch": 5.017074026429658,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 21448
    },
    {
      "epoch": 5.0173079172026664,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 21449
    },
    {
      "epoch": 5.017541807975675,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 21450
    },
    {
      "epoch": 5.017775698748684,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7667,
      "step": 21451
    },
    {
      "epoch": 5.018009589521693,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 21452
    },
    {
      "epoch": 5.018243480294703,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 21453
    },
    {
      "epoch": 5.0184773710677115,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 21454
    },
    {
      "epoch": 5.01871126184072,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 21455
    },
    {
      "epoch": 5.018945152613729,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6373,
      "step": 21456
    },
    {
      "epoch": 5.019179043386738,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 21457
    },
    {
      "epoch": 5.019412934159748,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 21458
    },
    {
      "epoch": 5.019646824932757,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 21459
    },
    {
      "epoch": 5.019880715705765,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 21460
    },
    {
      "epoch": 5.020114606478774,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 21461
    },
    {
      "epoch": 5.020348497251783,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 21462
    },
    {
      "epoch": 5.020582388024793,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 21463
    },
    {
      "epoch": 5.020816278797802,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 21464
    },
    {
      "epoch": 5.0210501695708105,
      "grad_norm": 7.65625,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 21465
    },
    {
      "epoch": 5.021284060343819,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 21466
    },
    {
      "epoch": 5.021517951116828,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 21467
    },
    {
      "epoch": 5.021751841889837,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 21468
    },
    {
      "epoch": 5.021985732662847,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5446,
      "step": 21469
    },
    {
      "epoch": 5.022219623435856,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9381,
      "step": 21470
    },
    {
      "epoch": 5.022453514208864,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7745,
      "step": 21471
    },
    {
      "epoch": 5.022687404981873,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5761,
      "step": 21472
    },
    {
      "epoch": 5.022921295754882,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 21473
    },
    {
      "epoch": 5.023155186527892,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 21474
    },
    {
      "epoch": 5.023389077300901,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5474,
      "step": 21475
    },
    {
      "epoch": 5.0236229680739095,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 21476
    },
    {
      "epoch": 5.023856858846918,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 21477
    },
    {
      "epoch": 5.024090749619927,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 21478
    },
    {
      "epoch": 5.024324640392937,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 21479
    },
    {
      "epoch": 5.024558531165946,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5912,
      "step": 21480
    },
    {
      "epoch": 5.024792421938955,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5624,
      "step": 21481
    },
    {
      "epoch": 5.025026312711963,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 21482
    },
    {
      "epoch": 5.025260203484972,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4605,
      "step": 21483
    },
    {
      "epoch": 5.025494094257981,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 21484
    },
    {
      "epoch": 5.025727985030991,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 21485
    },
    {
      "epoch": 5.025961875804,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6687,
      "step": 21486
    },
    {
      "epoch": 5.0261957665770085,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 21487
    },
    {
      "epoch": 5.026429657350017,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 21488
    },
    {
      "epoch": 5.026663548123026,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.569,
      "step": 21489
    },
    {
      "epoch": 5.026897438896036,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3693,
      "step": 21490
    },
    {
      "epoch": 5.027131329669045,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 21491
    },
    {
      "epoch": 5.027365220442054,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 21492
    },
    {
      "epoch": 5.027599111215062,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6886,
      "step": 21493
    },
    {
      "epoch": 5.027833001988071,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 21494
    },
    {
      "epoch": 5.028066892761081,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 21495
    },
    {
      "epoch": 5.02830078353409,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 21496
    },
    {
      "epoch": 5.028534674307099,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5632,
      "step": 21497
    },
    {
      "epoch": 5.0287685650801075,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6167,
      "step": 21498
    },
    {
      "epoch": 5.029002455853116,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 21499
    },
    {
      "epoch": 5.029236346626125,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 21500
    },
    {
      "epoch": 5.029236346626125,
      "eval_runtime": 4.6296,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 21500
    },
    {
      "epoch": 5.029470237399135,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 21501
    },
    {
      "epoch": 5.029704128172144,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 21502
    },
    {
      "epoch": 5.029938018945153,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6087,
      "step": 21503
    },
    {
      "epoch": 5.030171909718161,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 21504
    },
    {
      "epoch": 5.03040580049117,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 21505
    },
    {
      "epoch": 5.03063969126418,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 21506
    },
    {
      "epoch": 5.030873582037189,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6628,
      "step": 21507
    },
    {
      "epoch": 5.031107472810198,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7073,
      "step": 21508
    },
    {
      "epoch": 5.0313413635832065,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 21509
    },
    {
      "epoch": 5.031575254356215,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 21510
    },
    {
      "epoch": 5.031809145129225,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5881,
      "step": 21511
    },
    {
      "epoch": 5.032043035902234,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5912,
      "step": 21512
    },
    {
      "epoch": 5.032276926675243,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6785,
      "step": 21513
    },
    {
      "epoch": 5.032510817448252,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5784,
      "step": 21514
    },
    {
      "epoch": 5.03274470822126,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 21515
    },
    {
      "epoch": 5.032978598994269,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7369,
      "step": 21516
    },
    {
      "epoch": 5.033212489767279,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6546,
      "step": 21517
    },
    {
      "epoch": 5.033446380540288,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 21518
    },
    {
      "epoch": 5.033680271313297,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 21519
    },
    {
      "epoch": 5.0339141620863055,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5824,
      "step": 21520
    },
    {
      "epoch": 5.034148052859314,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 21521
    },
    {
      "epoch": 5.034381943632324,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 21522
    },
    {
      "epoch": 5.034615834405333,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7495,
      "step": 21523
    },
    {
      "epoch": 5.034849725178342,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 21524
    },
    {
      "epoch": 5.035083615951351,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6245,
      "step": 21525
    },
    {
      "epoch": 5.035317506724359,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 21526
    },
    {
      "epoch": 5.035551397497369,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4364,
      "step": 21527
    },
    {
      "epoch": 5.035785288270378,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 21528
    },
    {
      "epoch": 5.036019179043387,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 21529
    },
    {
      "epoch": 5.036253069816396,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5755,
      "step": 21530
    },
    {
      "epoch": 5.0364869605894045,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 21531
    },
    {
      "epoch": 5.036720851362413,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6417,
      "step": 21532
    },
    {
      "epoch": 5.036954742135423,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 21533
    },
    {
      "epoch": 5.037188632908432,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 21534
    },
    {
      "epoch": 5.037422523681441,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 21535
    },
    {
      "epoch": 5.03765641445445,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 21536
    },
    {
      "epoch": 5.037890305227458,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 21537
    },
    {
      "epoch": 5.038124196000468,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 21538
    },
    {
      "epoch": 5.038358086773477,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 21539
    },
    {
      "epoch": 5.038591977546486,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6222,
      "step": 21540
    },
    {
      "epoch": 5.038825868319495,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.4641,
      "step": 21541
    },
    {
      "epoch": 5.0390597590925035,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 21542
    },
    {
      "epoch": 5.039293649865513,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.5497,
      "step": 21543
    },
    {
      "epoch": 5.039527540638522,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 21544
    },
    {
      "epoch": 5.039761431411531,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 21545
    },
    {
      "epoch": 5.03999532218454,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 21546
    },
    {
      "epoch": 5.0402292129575486,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 21547
    },
    {
      "epoch": 5.040463103730557,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7283,
      "step": 21548
    },
    {
      "epoch": 5.040696994503567,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6001,
      "step": 21549
    },
    {
      "epoch": 5.040930885276576,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6819,
      "step": 21550
    },
    {
      "epoch": 5.041164776049585,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 21551
    },
    {
      "epoch": 5.041398666822594,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 21552
    },
    {
      "epoch": 5.0416325575956025,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 21553
    },
    {
      "epoch": 5.041866448368612,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 21554
    },
    {
      "epoch": 5.042100339141621,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4493,
      "step": 21555
    },
    {
      "epoch": 5.04233422991463,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 21556
    },
    {
      "epoch": 5.042568120687639,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6359,
      "step": 21557
    },
    {
      "epoch": 5.0428020114606475,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5824,
      "step": 21558
    },
    {
      "epoch": 5.043035902233657,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6505,
      "step": 21559
    },
    {
      "epoch": 5.043269793006666,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.813,
      "step": 21560
    },
    {
      "epoch": 5.043503683779675,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 21561
    },
    {
      "epoch": 5.043737574552684,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 21562
    },
    {
      "epoch": 5.043971465325693,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.5339,
      "step": 21563
    },
    {
      "epoch": 5.044205356098702,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 21564
    },
    {
      "epoch": 5.044439246871711,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 21565
    },
    {
      "epoch": 5.04467313764472,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 21566
    },
    {
      "epoch": 5.044907028417729,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6536,
      "step": 21567
    },
    {
      "epoch": 5.045140919190738,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 21568
    },
    {
      "epoch": 5.0453748099637465,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 21569
    },
    {
      "epoch": 5.045608700736756,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 21570
    },
    {
      "epoch": 5.045842591509765,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 21571
    },
    {
      "epoch": 5.046076482282774,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6022,
      "step": 21572
    },
    {
      "epoch": 5.046310373055783,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7107,
      "step": 21573
    },
    {
      "epoch": 5.046544263828792,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 21574
    },
    {
      "epoch": 5.046778154601801,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 21575
    },
    {
      "epoch": 5.04701204537481,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 21576
    },
    {
      "epoch": 5.047245936147819,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 21577
    },
    {
      "epoch": 5.047479826920828,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.4921,
      "step": 21578
    },
    {
      "epoch": 5.047713717693837,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7128,
      "step": 21579
    },
    {
      "epoch": 5.0479476084668455,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 21580
    },
    {
      "epoch": 5.048181499239855,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 21581
    },
    {
      "epoch": 5.048415390012864,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 21582
    },
    {
      "epoch": 5.048649280785873,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 21583
    },
    {
      "epoch": 5.048883171558882,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7423,
      "step": 21584
    },
    {
      "epoch": 5.049117062331891,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 21585
    },
    {
      "epoch": 5.0493509531049,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 21586
    },
    {
      "epoch": 5.049584843877909,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6673,
      "step": 21587
    },
    {
      "epoch": 5.049818734650918,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 21588
    },
    {
      "epoch": 5.050052625423927,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8734,
      "step": 21589
    },
    {
      "epoch": 5.050286516196936,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 21590
    },
    {
      "epoch": 5.050520406969945,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.623,
      "step": 21591
    },
    {
      "epoch": 5.050754297742954,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5676,
      "step": 21592
    },
    {
      "epoch": 5.050988188515963,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 21593
    },
    {
      "epoch": 5.051222079288972,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7201,
      "step": 21594
    },
    {
      "epoch": 5.051455970061981,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 21595
    },
    {
      "epoch": 5.0516898608349905,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.518,
      "step": 21596
    },
    {
      "epoch": 5.051923751607999,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.565,
      "step": 21597
    },
    {
      "epoch": 5.052157642381008,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1983,
      "step": 21598
    },
    {
      "epoch": 5.052391533154017,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 21599
    },
    {
      "epoch": 5.052625423927026,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 21600
    },
    {
      "epoch": 5.052625423927026,
      "eval_runtime": 4.7315,
      "eval_samples_per_second": 0.211,
      "eval_steps_per_second": 0.211,
      "step": 21600
    },
    {
      "epoch": 5.052859314700035,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 21601
    },
    {
      "epoch": 5.053093205473044,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 21602
    },
    {
      "epoch": 5.053327096246053,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 21603
    },
    {
      "epoch": 5.053560987019062,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 21604
    },
    {
      "epoch": 5.053794877792071,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 21605
    },
    {
      "epoch": 5.05402876856508,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 21606
    },
    {
      "epoch": 5.0542626593380895,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8141,
      "step": 21607
    },
    {
      "epoch": 5.054496550111098,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 21608
    },
    {
      "epoch": 5.054730440884107,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 21609
    },
    {
      "epoch": 5.054964331657116,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 21610
    },
    {
      "epoch": 5.055198222430125,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 21611
    },
    {
      "epoch": 5.055432113203135,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 21612
    },
    {
      "epoch": 5.055666003976143,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 21613
    },
    {
      "epoch": 5.055899894749152,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 21614
    },
    {
      "epoch": 5.056133785522161,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 21615
    },
    {
      "epoch": 5.05636767629517,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 21616
    },
    {
      "epoch": 5.056601567068179,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5748,
      "step": 21617
    },
    {
      "epoch": 5.0568354578411885,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 21618
    },
    {
      "epoch": 5.057069348614197,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 21619
    },
    {
      "epoch": 5.057303239387206,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.6427,
      "step": 21620
    },
    {
      "epoch": 5.057537130160215,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 21621
    },
    {
      "epoch": 5.057771020933224,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 21622
    },
    {
      "epoch": 5.058004911706234,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 21623
    },
    {
      "epoch": 5.058238802479242,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 21624
    },
    {
      "epoch": 5.058472693252251,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 21625
    },
    {
      "epoch": 5.05870658402526,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 21626
    },
    {
      "epoch": 5.058940474798269,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 21627
    },
    {
      "epoch": 5.059174365571279,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 21628
    },
    {
      "epoch": 5.0594082563442875,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 21629
    },
    {
      "epoch": 5.059642147117296,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 21630
    },
    {
      "epoch": 5.059876037890305,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5669,
      "step": 21631
    },
    {
      "epoch": 5.060109928663314,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6624,
      "step": 21632
    },
    {
      "epoch": 5.060343819436323,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 21633
    },
    {
      "epoch": 5.060577710209333,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 21634
    },
    {
      "epoch": 5.060811600982341,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 21635
    },
    {
      "epoch": 5.06104549175535,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8676,
      "step": 21636
    },
    {
      "epoch": 5.061279382528359,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5594,
      "step": 21637
    },
    {
      "epoch": 5.061513273301368,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 21638
    },
    {
      "epoch": 5.061747164074378,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 21639
    },
    {
      "epoch": 5.0619810548473865,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 21640
    },
    {
      "epoch": 5.062214945620395,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6951,
      "step": 21641
    },
    {
      "epoch": 5.062448836393404,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5539,
      "step": 21642
    },
    {
      "epoch": 5.062682727166413,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7631,
      "step": 21643
    },
    {
      "epoch": 5.062916617939423,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 21644
    },
    {
      "epoch": 5.0631505087124316,
      "grad_norm": 2.46875,
      "learning_rate": 3e-05,
      "loss": 1.4924,
      "step": 21645
    },
    {
      "epoch": 5.06338439948544,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 21646
    },
    {
      "epoch": 5.063618290258449,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6497,
      "step": 21647
    },
    {
      "epoch": 5.063852181031458,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 21648
    },
    {
      "epoch": 5.064086071804467,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 21649
    },
    {
      "epoch": 5.064319962577477,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7078,
      "step": 21650
    },
    {
      "epoch": 5.0645538533504855,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.697,
      "step": 21651
    },
    {
      "epoch": 5.064787744123494,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 21652
    },
    {
      "epoch": 5.065021634896503,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 21653
    },
    {
      "epoch": 5.065255525669512,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8183,
      "step": 21654
    },
    {
      "epoch": 5.065489416442522,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 21655
    },
    {
      "epoch": 5.0657233072155305,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 21656
    },
    {
      "epoch": 5.065957197988539,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5689,
      "step": 21657
    },
    {
      "epoch": 5.066191088761548,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6424,
      "step": 21658
    },
    {
      "epoch": 5.066424979534557,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0032,
      "step": 21659
    },
    {
      "epoch": 5.066658870307567,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7624,
      "step": 21660
    },
    {
      "epoch": 5.066892761080576,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 21661
    },
    {
      "epoch": 5.0671266518535845,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4164,
      "step": 21662
    },
    {
      "epoch": 5.067360542626593,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6749,
      "step": 21663
    },
    {
      "epoch": 5.067594433399602,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6844,
      "step": 21664
    },
    {
      "epoch": 5.067828324172611,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7782,
      "step": 21665
    },
    {
      "epoch": 5.068062214945621,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 21666
    },
    {
      "epoch": 5.0682961057186295,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6495,
      "step": 21667
    },
    {
      "epoch": 5.068529996491638,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7971,
      "step": 21668
    },
    {
      "epoch": 5.068763887264647,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 21669
    },
    {
      "epoch": 5.068997778037656,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 21670
    },
    {
      "epoch": 5.069231668810666,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 21671
    },
    {
      "epoch": 5.069465559583675,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 21672
    },
    {
      "epoch": 5.0696994503566835,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 21673
    },
    {
      "epoch": 5.069933341129692,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7772,
      "step": 21674
    },
    {
      "epoch": 5.070167231902701,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 21675
    },
    {
      "epoch": 5.070401122675711,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 21676
    },
    {
      "epoch": 5.07063501344872,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 21677
    },
    {
      "epoch": 5.0708689042217285,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6796,
      "step": 21678
    },
    {
      "epoch": 5.071102794994737,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 21679
    },
    {
      "epoch": 5.071336685767746,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8122,
      "step": 21680
    },
    {
      "epoch": 5.071570576540755,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6179,
      "step": 21681
    },
    {
      "epoch": 5.071804467313765,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 21682
    },
    {
      "epoch": 5.072038358086774,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 21683
    },
    {
      "epoch": 5.0722722488597825,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8131,
      "step": 21684
    },
    {
      "epoch": 5.072506139632791,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 21685
    },
    {
      "epoch": 5.0727400304058,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 21686
    },
    {
      "epoch": 5.07297392117881,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6402,
      "step": 21687
    },
    {
      "epoch": 5.073207811951819,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 21688
    },
    {
      "epoch": 5.0734417027248275,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 21689
    },
    {
      "epoch": 5.073675593497836,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 21690
    },
    {
      "epoch": 5.073909484270845,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 21691
    },
    {
      "epoch": 5.074143375043855,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 21692
    },
    {
      "epoch": 5.074377265816864,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.483,
      "step": 21693
    },
    {
      "epoch": 5.074611156589873,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 21694
    },
    {
      "epoch": 5.0748450473628814,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 21695
    },
    {
      "epoch": 5.07507893813589,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7292,
      "step": 21696
    },
    {
      "epoch": 5.075312828908899,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 21697
    },
    {
      "epoch": 5.075546719681909,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5856,
      "step": 21698
    },
    {
      "epoch": 5.075780610454918,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 21699
    },
    {
      "epoch": 5.0760145012279265,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8002,
      "step": 21700
    },
    {
      "epoch": 5.0760145012279265,
      "eval_runtime": 4.6058,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 21700
    },
    {
      "epoch": 5.076248392000935,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 21701
    },
    {
      "epoch": 5.076482282773944,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5404,
      "step": 21702
    },
    {
      "epoch": 5.076716173546954,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 21703
    },
    {
      "epoch": 5.076950064319963,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 21704
    },
    {
      "epoch": 5.077183955092972,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 21705
    },
    {
      "epoch": 5.07741784586598,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5638,
      "step": 21706
    },
    {
      "epoch": 5.077651736638989,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 21707
    },
    {
      "epoch": 5.077885627411999,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 21708
    },
    {
      "epoch": 5.078119518185008,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 21709
    },
    {
      "epoch": 5.078353408958017,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 21710
    },
    {
      "epoch": 5.0785872997310255,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5211,
      "step": 21711
    },
    {
      "epoch": 5.078821190504034,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 21712
    },
    {
      "epoch": 5.079055081277043,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5213,
      "step": 21713
    },
    {
      "epoch": 5.079288972050053,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7416,
      "step": 21714
    },
    {
      "epoch": 5.079522862823062,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 21715
    },
    {
      "epoch": 5.079756753596071,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 21716
    },
    {
      "epoch": 5.079990644369079,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7161,
      "step": 21717
    },
    {
      "epoch": 5.080224535142088,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 21718
    },
    {
      "epoch": 5.080458425915098,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 21719
    },
    {
      "epoch": 5.080692316688107,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 21720
    },
    {
      "epoch": 5.080926207461116,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 21721
    },
    {
      "epoch": 5.0811600982341245,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7483,
      "step": 21722
    },
    {
      "epoch": 5.081393989007133,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7367,
      "step": 21723
    },
    {
      "epoch": 5.081627879780143,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 21724
    },
    {
      "epoch": 5.081861770553152,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 21725
    },
    {
      "epoch": 5.082095661326161,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 21726
    },
    {
      "epoch": 5.08232955209917,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 21727
    },
    {
      "epoch": 5.082563442872178,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5075,
      "step": 21728
    },
    {
      "epoch": 5.082797333645187,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5933,
      "step": 21729
    },
    {
      "epoch": 5.083031224418197,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 21730
    },
    {
      "epoch": 5.083265115191206,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7682,
      "step": 21731
    },
    {
      "epoch": 5.083499005964215,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 21732
    },
    {
      "epoch": 5.0837328967372235,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 21733
    },
    {
      "epoch": 5.083966787510232,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6082,
      "step": 21734
    },
    {
      "epoch": 5.084200678283242,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 21735
    },
    {
      "epoch": 5.084434569056251,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 21736
    },
    {
      "epoch": 5.08466845982926,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 21737
    },
    {
      "epoch": 5.084902350602269,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 21738
    },
    {
      "epoch": 5.085136241375277,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 21739
    },
    {
      "epoch": 5.085370132148287,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 21740
    },
    {
      "epoch": 5.085604022921296,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 21741
    },
    {
      "epoch": 5.085837913694305,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 21742
    },
    {
      "epoch": 5.086071804467314,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5964,
      "step": 21743
    },
    {
      "epoch": 5.0863056952403225,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 21744
    },
    {
      "epoch": 5.086539586013331,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.886,
      "step": 21745
    },
    {
      "epoch": 5.086773476786341,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 21746
    },
    {
      "epoch": 5.08700736755935,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 21747
    },
    {
      "epoch": 5.087241258332359,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 21748
    },
    {
      "epoch": 5.087475149105368,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6777,
      "step": 21749
    },
    {
      "epoch": 5.087709039878376,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7897,
      "step": 21750
    },
    {
      "epoch": 5.087942930651386,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8276,
      "step": 21751
    },
    {
      "epoch": 5.088176821424395,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 21752
    },
    {
      "epoch": 5.088410712197404,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 21753
    },
    {
      "epoch": 5.088644602970413,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 21754
    },
    {
      "epoch": 5.0888784937434215,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 21755
    },
    {
      "epoch": 5.089112384516431,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 21756
    },
    {
      "epoch": 5.08934627528944,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 21757
    },
    {
      "epoch": 5.089580166062449,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 21758
    },
    {
      "epoch": 5.089814056835458,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 21759
    },
    {
      "epoch": 5.090047947608467,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 21760
    },
    {
      "epoch": 5.090281838381475,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 21761
    },
    {
      "epoch": 5.090515729154485,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 21762
    },
    {
      "epoch": 5.090749619927494,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 21763
    },
    {
      "epoch": 5.090983510700503,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 21764
    },
    {
      "epoch": 5.091217401473512,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 21765
    },
    {
      "epoch": 5.0914512922465205,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.77,
      "step": 21766
    },
    {
      "epoch": 5.09168518301953,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6531,
      "step": 21767
    },
    {
      "epoch": 5.091919073792539,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3067,
      "step": 21768
    },
    {
      "epoch": 5.092152964565548,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 21769
    },
    {
      "epoch": 5.092386855338557,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 21770
    },
    {
      "epoch": 5.092620746111566,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 21771
    },
    {
      "epoch": 5.092854636884575,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 21772
    },
    {
      "epoch": 5.093088527657584,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 21773
    },
    {
      "epoch": 5.093322418430593,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 21774
    },
    {
      "epoch": 5.093556309203602,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 21775
    },
    {
      "epoch": 5.093790199976611,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7244,
      "step": 21776
    },
    {
      "epoch": 5.09402409074962,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 21777
    },
    {
      "epoch": 5.094257981522629,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 21778
    },
    {
      "epoch": 5.094491872295638,
      "grad_norm": 14.75,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 21779
    },
    {
      "epoch": 5.094725763068647,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6033,
      "step": 21780
    },
    {
      "epoch": 5.094959653841656,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7401,
      "step": 21781
    },
    {
      "epoch": 5.095193544614665,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7851,
      "step": 21782
    },
    {
      "epoch": 5.095427435387674,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6434,
      "step": 21783
    },
    {
      "epoch": 5.095661326160683,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8174,
      "step": 21784
    },
    {
      "epoch": 5.095895216933692,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.6484,
      "step": 21785
    },
    {
      "epoch": 5.096129107706701,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6377,
      "step": 21786
    },
    {
      "epoch": 5.09636299847971,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 21787
    },
    {
      "epoch": 5.096596889252719,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9001,
      "step": 21788
    },
    {
      "epoch": 5.096830780025728,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 21789
    },
    {
      "epoch": 5.097064670798737,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 21790
    },
    {
      "epoch": 5.097298561571746,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 21791
    },
    {
      "epoch": 5.097532452344755,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 21792
    },
    {
      "epoch": 5.0977663431177636,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 21793
    },
    {
      "epoch": 5.098000233890773,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7616,
      "step": 21794
    },
    {
      "epoch": 5.098234124663782,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 21795
    },
    {
      "epoch": 5.098468015436791,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 21796
    },
    {
      "epoch": 5.0987019062098,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 21797
    },
    {
      "epoch": 5.098935796982809,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 21798
    },
    {
      "epoch": 5.099169687755818,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 21799
    },
    {
      "epoch": 5.099403578528827,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 21800
    },
    {
      "epoch": 5.099403578528827,
      "eval_runtime": 4.5895,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 21800
    },
    {
      "epoch": 5.099637469301836,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5523,
      "step": 21801
    },
    {
      "epoch": 5.099871360074845,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5914,
      "step": 21802
    },
    {
      "epoch": 5.100105250847854,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 21803
    },
    {
      "epoch": 5.100339141620863,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 21804
    },
    {
      "epoch": 5.100573032393872,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 21805
    },
    {
      "epoch": 5.100806923166881,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 21806
    },
    {
      "epoch": 5.10104081393989,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6647,
      "step": 21807
    },
    {
      "epoch": 5.101274704712899,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7,
      "step": 21808
    },
    {
      "epoch": 5.1015085954859085,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6648,
      "step": 21809
    },
    {
      "epoch": 5.101742486258917,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 21810
    },
    {
      "epoch": 5.101976377031926,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 21811
    },
    {
      "epoch": 5.102210267804935,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 21812
    },
    {
      "epoch": 5.102444158577944,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8719,
      "step": 21813
    },
    {
      "epoch": 5.102678049350953,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 21814
    },
    {
      "epoch": 5.102911940123962,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 21815
    },
    {
      "epoch": 5.103145830896971,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 21816
    },
    {
      "epoch": 5.10337972166998,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 21817
    },
    {
      "epoch": 5.103613612442989,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7589,
      "step": 21818
    },
    {
      "epoch": 5.103847503215998,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4556,
      "step": 21819
    },
    {
      "epoch": 5.1040813939890075,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.866,
      "step": 21820
    },
    {
      "epoch": 5.104315284762016,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6449,
      "step": 21821
    },
    {
      "epoch": 5.104549175535025,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 21822
    },
    {
      "epoch": 5.104783066308034,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 21823
    },
    {
      "epoch": 5.105016957081043,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 21824
    },
    {
      "epoch": 5.105250847854053,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 21825
    },
    {
      "epoch": 5.105484738627061,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 21826
    },
    {
      "epoch": 5.10571862940007,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5533,
      "step": 21827
    },
    {
      "epoch": 5.105952520173079,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.4145,
      "step": 21828
    },
    {
      "epoch": 5.106186410946088,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7123,
      "step": 21829
    },
    {
      "epoch": 5.106420301719097,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5953,
      "step": 21830
    },
    {
      "epoch": 5.1066541924921065,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7851,
      "step": 21831
    },
    {
      "epoch": 5.106888083265115,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 21832
    },
    {
      "epoch": 5.107121974038124,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 21833
    },
    {
      "epoch": 5.107355864811133,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 21834
    },
    {
      "epoch": 5.107589755584142,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 21835
    },
    {
      "epoch": 5.107823646357152,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 21836
    },
    {
      "epoch": 5.10805753713016,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 21837
    },
    {
      "epoch": 5.108291427903169,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 21838
    },
    {
      "epoch": 5.108525318676178,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 21839
    },
    {
      "epoch": 5.108759209449187,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 21840
    },
    {
      "epoch": 5.108993100222197,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 21841
    },
    {
      "epoch": 5.1092269909952055,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 21842
    },
    {
      "epoch": 5.109460881768214,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6363,
      "step": 21843
    },
    {
      "epoch": 5.109694772541223,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 21844
    },
    {
      "epoch": 5.109928663314232,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 21845
    },
    {
      "epoch": 5.110162554087241,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5516,
      "step": 21846
    },
    {
      "epoch": 5.110396444860251,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 21847
    },
    {
      "epoch": 5.110630335633259,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 21848
    },
    {
      "epoch": 5.110864226406268,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7407,
      "step": 21849
    },
    {
      "epoch": 5.111098117179277,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6502,
      "step": 21850
    },
    {
      "epoch": 5.111332007952286,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8205,
      "step": 21851
    },
    {
      "epoch": 5.111565898725296,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 21852
    },
    {
      "epoch": 5.1117997894983045,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9976,
      "step": 21853
    },
    {
      "epoch": 5.112033680271313,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6882,
      "step": 21854
    },
    {
      "epoch": 5.112267571044322,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5913,
      "step": 21855
    },
    {
      "epoch": 5.112501461817331,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5661,
      "step": 21856
    },
    {
      "epoch": 5.112735352590341,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6696,
      "step": 21857
    },
    {
      "epoch": 5.11296924336335,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 21858
    },
    {
      "epoch": 5.113203134136358,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 21859
    },
    {
      "epoch": 5.113437024909367,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 21860
    },
    {
      "epoch": 5.113670915682376,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.512,
      "step": 21861
    },
    {
      "epoch": 5.113904806455385,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5651,
      "step": 21862
    },
    {
      "epoch": 5.114138697228395,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8726,
      "step": 21863
    },
    {
      "epoch": 5.1143725880014035,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 21864
    },
    {
      "epoch": 5.114606478774412,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 21865
    },
    {
      "epoch": 5.114840369547421,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 21866
    },
    {
      "epoch": 5.11507426032043,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6225,
      "step": 21867
    },
    {
      "epoch": 5.11530815109344,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6904,
      "step": 21868
    },
    {
      "epoch": 5.115542041866449,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 21869
    },
    {
      "epoch": 5.115775932639457,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 21870
    },
    {
      "epoch": 5.116009823412466,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6154,
      "step": 21871
    },
    {
      "epoch": 5.116243714185475,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6987,
      "step": 21872
    },
    {
      "epoch": 5.116477604958485,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 21873
    },
    {
      "epoch": 5.116711495731494,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 21874
    },
    {
      "epoch": 5.1169453865045025,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5664,
      "step": 21875
    },
    {
      "epoch": 5.117179277277511,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 21876
    },
    {
      "epoch": 5.11741316805052,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 21877
    },
    {
      "epoch": 5.117647058823529,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6622,
      "step": 21878
    },
    {
      "epoch": 5.117880949596539,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 21879
    },
    {
      "epoch": 5.118114840369548,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 21880
    },
    {
      "epoch": 5.118348731142556,
      "grad_norm": 7.5625,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 21881
    },
    {
      "epoch": 5.118582621915565,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 21882
    },
    {
      "epoch": 5.118816512688574,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4632,
      "step": 21883
    },
    {
      "epoch": 5.119050403461584,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6589,
      "step": 21884
    },
    {
      "epoch": 5.119284294234593,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.4623,
      "step": 21885
    },
    {
      "epoch": 5.1195181850076015,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 21886
    },
    {
      "epoch": 5.11975207578061,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5276,
      "step": 21887
    },
    {
      "epoch": 5.119985966553619,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 21888
    },
    {
      "epoch": 5.120219857326629,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.696,
      "step": 21889
    },
    {
      "epoch": 5.120453748099638,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6713,
      "step": 21890
    },
    {
      "epoch": 5.1206876388726466,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5267,
      "step": 21891
    },
    {
      "epoch": 5.120921529645655,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6253,
      "step": 21892
    },
    {
      "epoch": 5.121155420418664,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 21893
    },
    {
      "epoch": 5.121389311191673,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8684,
      "step": 21894
    },
    {
      "epoch": 5.121623201964683,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5692,
      "step": 21895
    },
    {
      "epoch": 5.121857092737692,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6052,
      "step": 21896
    },
    {
      "epoch": 5.1220909835107005,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 21897
    },
    {
      "epoch": 5.122324874283709,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6784,
      "step": 21898
    },
    {
      "epoch": 5.122558765056718,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 21899
    },
    {
      "epoch": 5.122792655829728,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 21900
    },
    {
      "epoch": 5.122792655829728,
      "eval_runtime": 4.6336,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 21900
    },
    {
      "epoch": 5.123026546602737,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8136,
      "step": 21901
    },
    {
      "epoch": 5.1232604373757455,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 21902
    },
    {
      "epoch": 5.123494328148754,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 21903
    },
    {
      "epoch": 5.123728218921763,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 21904
    },
    {
      "epoch": 5.123962109694773,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 21905
    },
    {
      "epoch": 5.124196000467782,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7075,
      "step": 21906
    },
    {
      "epoch": 5.124429891240791,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 21907
    },
    {
      "epoch": 5.1246637820137995,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5866,
      "step": 21908
    },
    {
      "epoch": 5.124897672786808,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7184,
      "step": 21909
    },
    {
      "epoch": 5.125131563559817,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4226,
      "step": 21910
    },
    {
      "epoch": 5.125365454332827,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 21911
    },
    {
      "epoch": 5.125599345105836,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 21912
    },
    {
      "epoch": 5.1258332358788445,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.4645,
      "step": 21913
    },
    {
      "epoch": 5.126067126651853,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 21914
    },
    {
      "epoch": 5.126301017424862,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 21915
    },
    {
      "epoch": 5.126534908197872,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.5772,
      "step": 21916
    },
    {
      "epoch": 5.126768798970881,
      "grad_norm": 16.75,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 21917
    },
    {
      "epoch": 5.12700268974389,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 21918
    },
    {
      "epoch": 5.1272365805168985,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 21919
    },
    {
      "epoch": 5.127470471289907,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7252,
      "step": 21920
    },
    {
      "epoch": 5.127704362062917,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6393,
      "step": 21921
    },
    {
      "epoch": 5.127938252835926,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 21922
    },
    {
      "epoch": 5.128172143608935,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 21923
    },
    {
      "epoch": 5.1284060343819435,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6072,
      "step": 21924
    },
    {
      "epoch": 5.128639925154952,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 21925
    },
    {
      "epoch": 5.128873815927962,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 21926
    },
    {
      "epoch": 5.129107706700971,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 21927
    },
    {
      "epoch": 5.12934159747398,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6172,
      "step": 21928
    },
    {
      "epoch": 5.129575488246989,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5798,
      "step": 21929
    },
    {
      "epoch": 5.1298093790199975,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 21930
    },
    {
      "epoch": 5.130043269793006,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 21931
    },
    {
      "epoch": 5.130277160566016,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 21932
    },
    {
      "epoch": 5.130511051339025,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 21933
    },
    {
      "epoch": 5.130744942112034,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 21934
    },
    {
      "epoch": 5.1309788328850425,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5435,
      "step": 21935
    },
    {
      "epoch": 5.131212723658051,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 21936
    },
    {
      "epoch": 5.131446614431061,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 21937
    },
    {
      "epoch": 5.13168050520407,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6195,
      "step": 21938
    },
    {
      "epoch": 5.131914395977079,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 21939
    },
    {
      "epoch": 5.132148286750088,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 21940
    },
    {
      "epoch": 5.1323821775230964,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 21941
    },
    {
      "epoch": 5.132616068296105,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 21942
    },
    {
      "epoch": 5.132849959069115,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 21943
    },
    {
      "epoch": 5.133083849842124,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 21944
    },
    {
      "epoch": 5.133317740615133,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.4702,
      "step": 21945
    },
    {
      "epoch": 5.1335516313881415,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 21946
    },
    {
      "epoch": 5.13378552216115,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 21947
    },
    {
      "epoch": 5.13401941293416,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4834,
      "step": 21948
    },
    {
      "epoch": 5.134253303707169,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6824,
      "step": 21949
    },
    {
      "epoch": 5.134487194480178,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 21950
    },
    {
      "epoch": 5.134721085253187,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 21951
    },
    {
      "epoch": 5.134954976026195,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 21952
    },
    {
      "epoch": 5.135188866799205,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 21953
    },
    {
      "epoch": 5.135422757572214,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 21954
    },
    {
      "epoch": 5.135656648345223,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7259,
      "step": 21955
    },
    {
      "epoch": 5.135890539118232,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 21956
    },
    {
      "epoch": 5.1361244298912405,
      "grad_norm": 8.9375,
      "learning_rate": 3e-05,
      "loss": 1.6735,
      "step": 21957
    },
    {
      "epoch": 5.13635832066425,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6068,
      "step": 21958
    },
    {
      "epoch": 5.136592211437259,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2942,
      "step": 21959
    },
    {
      "epoch": 5.136826102210268,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 21960
    },
    {
      "epoch": 5.137059992983277,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6986,
      "step": 21961
    },
    {
      "epoch": 5.137293883756286,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8945,
      "step": 21962
    },
    {
      "epoch": 5.137527774529294,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 21963
    },
    {
      "epoch": 5.137761665302304,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 21964
    },
    {
      "epoch": 5.137995556075313,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 21965
    },
    {
      "epoch": 5.138229446848322,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 21966
    },
    {
      "epoch": 5.138463337621331,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 21967
    },
    {
      "epoch": 5.1386972283943395,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 21968
    },
    {
      "epoch": 5.138931119167349,
      "grad_norm": 10.1875,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 21969
    },
    {
      "epoch": 5.139165009940358,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6937,
      "step": 21970
    },
    {
      "epoch": 5.139398900713367,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 21971
    },
    {
      "epoch": 5.139632791486376,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5136,
      "step": 21972
    },
    {
      "epoch": 5.139866682259385,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 21973
    },
    {
      "epoch": 5.140100573032393,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 21974
    },
    {
      "epoch": 5.140334463805403,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 21975
    },
    {
      "epoch": 5.140568354578412,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6458,
      "step": 21976
    },
    {
      "epoch": 5.140802245351421,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 21977
    },
    {
      "epoch": 5.14103613612443,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7352,
      "step": 21978
    },
    {
      "epoch": 5.1412700268974385,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4756,
      "step": 21979
    },
    {
      "epoch": 5.141503917670448,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 21980
    },
    {
      "epoch": 5.141737808443457,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 21981
    },
    {
      "epoch": 5.141971699216466,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6094,
      "step": 21982
    },
    {
      "epoch": 5.142205589989475,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 21983
    },
    {
      "epoch": 5.142439480762484,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4713,
      "step": 21984
    },
    {
      "epoch": 5.142673371535493,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6175,
      "step": 21985
    },
    {
      "epoch": 5.142907262308502,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 21986
    },
    {
      "epoch": 5.143141153081511,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 21987
    },
    {
      "epoch": 5.14337504385452,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 21988
    },
    {
      "epoch": 5.143608934627529,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7134,
      "step": 21989
    },
    {
      "epoch": 5.143842825400538,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 21990
    },
    {
      "epoch": 5.144076716173547,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 21991
    },
    {
      "epoch": 5.144310606946556,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.594,
      "step": 21992
    },
    {
      "epoch": 5.144544497719565,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.682,
      "step": 21993
    },
    {
      "epoch": 5.144778388492574,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 21994
    },
    {
      "epoch": 5.145012279265583,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.5221,
      "step": 21995
    },
    {
      "epoch": 5.145246170038592,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6463,
      "step": 21996
    },
    {
      "epoch": 5.145480060811601,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 21997
    },
    {
      "epoch": 5.14571395158461,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6283,
      "step": 21998
    },
    {
      "epoch": 5.145947842357619,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5306,
      "step": 21999
    },
    {
      "epoch": 5.146181733130628,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6738,
      "step": 22000
    },
    {
      "epoch": 5.146181733130628,
      "eval_runtime": 4.6112,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 22000
    },
    {
      "epoch": 5.146415623903637,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 22001
    },
    {
      "epoch": 5.146649514676646,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 22002
    },
    {
      "epoch": 5.146883405449655,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 22003
    },
    {
      "epoch": 5.147117296222664,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6467,
      "step": 22004
    },
    {
      "epoch": 5.147351186995673,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4777,
      "step": 22005
    },
    {
      "epoch": 5.147585077768682,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7131,
      "step": 22006
    },
    {
      "epoch": 5.147818968541691,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6726,
      "step": 22007
    },
    {
      "epoch": 5.1480528593147,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6443,
      "step": 22008
    },
    {
      "epoch": 5.148286750087709,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6255,
      "step": 22009
    },
    {
      "epoch": 5.148520640860718,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 22010
    },
    {
      "epoch": 5.148754531633727,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 22011
    },
    {
      "epoch": 5.148988422406736,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4677,
      "step": 22012
    },
    {
      "epoch": 5.149222313179745,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 22013
    },
    {
      "epoch": 5.149456203952754,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.788,
      "step": 22014
    },
    {
      "epoch": 5.149690094725763,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 22015
    },
    {
      "epoch": 5.149923985498772,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 22016
    },
    {
      "epoch": 5.1501578762717815,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 22017
    },
    {
      "epoch": 5.15039176704479,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 22018
    },
    {
      "epoch": 5.150625657817799,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 22019
    },
    {
      "epoch": 5.150859548590808,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 22020
    },
    {
      "epoch": 5.151093439363817,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 22021
    },
    {
      "epoch": 5.1513273301368265,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0932,
      "step": 22022
    },
    {
      "epoch": 5.151561220909835,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 22023
    },
    {
      "epoch": 5.151795111682844,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 22024
    },
    {
      "epoch": 5.152029002455853,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 22025
    },
    {
      "epoch": 5.152262893228862,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 22026
    },
    {
      "epoch": 5.152496784001871,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 22027
    },
    {
      "epoch": 5.1527306747748804,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.798,
      "step": 22028
    },
    {
      "epoch": 5.152964565547889,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 22029
    },
    {
      "epoch": 5.153198456320898,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 22030
    },
    {
      "epoch": 5.153432347093907,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5565,
      "step": 22031
    },
    {
      "epoch": 5.153666237866916,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.71,
      "step": 22032
    },
    {
      "epoch": 5.1539001286399255,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 22033
    },
    {
      "epoch": 5.154134019412934,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6594,
      "step": 22034
    },
    {
      "epoch": 5.154367910185943,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 22035
    },
    {
      "epoch": 5.154601800958952,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 22036
    },
    {
      "epoch": 5.154835691731961,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6076,
      "step": 22037
    },
    {
      "epoch": 5.155069582504971,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 22038
    },
    {
      "epoch": 5.155303473277979,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 22039
    },
    {
      "epoch": 5.155537364050988,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 22040
    },
    {
      "epoch": 5.155771254823997,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 22041
    },
    {
      "epoch": 5.156005145597006,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 22042
    },
    {
      "epoch": 5.156239036370015,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8536,
      "step": 22043
    },
    {
      "epoch": 5.1564729271430245,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 22044
    },
    {
      "epoch": 5.156706817916033,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8101,
      "step": 22045
    },
    {
      "epoch": 5.156940708689042,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4866,
      "step": 22046
    },
    {
      "epoch": 5.157174599462051,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7865,
      "step": 22047
    },
    {
      "epoch": 5.15740849023506,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5318,
      "step": 22048
    },
    {
      "epoch": 5.15764238100807,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 22049
    },
    {
      "epoch": 5.157876271781078,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 22050
    },
    {
      "epoch": 5.158110162554087,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 22051
    },
    {
      "epoch": 5.158344053327096,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 22052
    },
    {
      "epoch": 5.158577944100105,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 22053
    },
    {
      "epoch": 5.158811834873115,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 22054
    },
    {
      "epoch": 5.1590457256461235,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 22055
    },
    {
      "epoch": 5.159279616419132,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7369,
      "step": 22056
    },
    {
      "epoch": 5.159513507192141,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6446,
      "step": 22057
    },
    {
      "epoch": 5.15974739796515,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 22058
    },
    {
      "epoch": 5.159981288738159,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 22059
    },
    {
      "epoch": 5.160215179511169,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 22060
    },
    {
      "epoch": 5.160449070284177,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 22061
    },
    {
      "epoch": 5.160682961057186,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 22062
    },
    {
      "epoch": 5.160916851830195,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5361,
      "step": 22063
    },
    {
      "epoch": 5.161150742603204,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 22064
    },
    {
      "epoch": 5.161384633376214,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 22065
    },
    {
      "epoch": 5.1616185241492225,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 22066
    },
    {
      "epoch": 5.161852414922231,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 22067
    },
    {
      "epoch": 5.16208630569524,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 22068
    },
    {
      "epoch": 5.162320196468249,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 22069
    },
    {
      "epoch": 5.162554087241259,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 22070
    },
    {
      "epoch": 5.162787978014268,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8008,
      "step": 22071
    },
    {
      "epoch": 5.163021868787276,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 22072
    },
    {
      "epoch": 5.163255759560285,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 22073
    },
    {
      "epoch": 5.163489650333294,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 22074
    },
    {
      "epoch": 5.163723541106303,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 22075
    },
    {
      "epoch": 5.163957431879313,
      "grad_norm": 9.5,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 22076
    },
    {
      "epoch": 5.1641913226523215,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 22077
    },
    {
      "epoch": 5.16442521342533,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.4563,
      "step": 22078
    },
    {
      "epoch": 5.164659104198339,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6983,
      "step": 22079
    },
    {
      "epoch": 5.164892994971348,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 22080
    },
    {
      "epoch": 5.165126885744358,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 22081
    },
    {
      "epoch": 5.165360776517367,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 22082
    },
    {
      "epoch": 5.165594667290375,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6772,
      "step": 22083
    },
    {
      "epoch": 5.165828558063384,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5622,
      "step": 22084
    },
    {
      "epoch": 5.166062448836393,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 22085
    },
    {
      "epoch": 5.166296339609403,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5802,
      "step": 22086
    },
    {
      "epoch": 5.166530230382412,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 22087
    },
    {
      "epoch": 5.1667641211554205,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 22088
    },
    {
      "epoch": 5.166998011928429,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 22089
    },
    {
      "epoch": 5.167231902701438,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 22090
    },
    {
      "epoch": 5.167465793474447,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 22091
    },
    {
      "epoch": 5.167699684247457,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 22092
    },
    {
      "epoch": 5.167933575020466,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4061,
      "step": 22093
    },
    {
      "epoch": 5.168167465793474,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 22094
    },
    {
      "epoch": 5.168401356566483,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6385,
      "step": 22095
    },
    {
      "epoch": 5.168635247339492,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 22096
    },
    {
      "epoch": 5.168869138112502,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 22097
    },
    {
      "epoch": 5.169103028885511,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 22098
    },
    {
      "epoch": 5.1693369196585195,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 22099
    },
    {
      "epoch": 5.169570810431528,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7862,
      "step": 22100
    },
    {
      "epoch": 5.169570810431528,
      "eval_runtime": 4.6241,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 22100
    },
    {
      "epoch": 5.169804701204537,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 22101
    },
    {
      "epoch": 5.170038591977547,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7616,
      "step": 22102
    },
    {
      "epoch": 5.170272482750556,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9094,
      "step": 22103
    },
    {
      "epoch": 5.170506373523565,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 22104
    },
    {
      "epoch": 5.170740264296573,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 22105
    },
    {
      "epoch": 5.170974155069582,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 22106
    },
    {
      "epoch": 5.171208045842591,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 22107
    },
    {
      "epoch": 5.171441936615601,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6858,
      "step": 22108
    },
    {
      "epoch": 5.17167582738861,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 22109
    },
    {
      "epoch": 5.1719097181616185,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 22110
    },
    {
      "epoch": 5.172143608934627,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 22111
    },
    {
      "epoch": 5.172377499707636,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 22112
    },
    {
      "epoch": 5.172611390480646,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 22113
    },
    {
      "epoch": 5.172845281253655,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 22114
    },
    {
      "epoch": 5.173079172026664,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7421,
      "step": 22115
    },
    {
      "epoch": 5.173313062799672,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 22116
    },
    {
      "epoch": 5.173546953572681,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 22117
    },
    {
      "epoch": 5.173780844345691,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 22118
    },
    {
      "epoch": 5.1740147351187,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 22119
    },
    {
      "epoch": 5.174248625891709,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 22120
    },
    {
      "epoch": 5.1744825166647175,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8166,
      "step": 22121
    },
    {
      "epoch": 5.174716407437726,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5983,
      "step": 22122
    },
    {
      "epoch": 5.174950298210735,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 22123
    },
    {
      "epoch": 5.175184188983745,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 22124
    },
    {
      "epoch": 5.175418079756754,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7045,
      "step": 22125
    },
    {
      "epoch": 5.175651970529763,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6138,
      "step": 22126
    },
    {
      "epoch": 5.175885861302771,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 22127
    },
    {
      "epoch": 5.17611975207578,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 22128
    },
    {
      "epoch": 5.17635364284879,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 22129
    },
    {
      "epoch": 5.176587533621799,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9388,
      "step": 22130
    },
    {
      "epoch": 5.176821424394808,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 22131
    },
    {
      "epoch": 5.1770553151678165,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 22132
    },
    {
      "epoch": 5.177289205940825,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 22133
    },
    {
      "epoch": 5.177523096713835,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 22134
    },
    {
      "epoch": 5.177756987486844,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 22135
    },
    {
      "epoch": 5.177990878259853,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 22136
    },
    {
      "epoch": 5.1782247690328616,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 22137
    },
    {
      "epoch": 5.17845865980587,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 22138
    },
    {
      "epoch": 5.17869255057888,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 22139
    },
    {
      "epoch": 5.178926441351889,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 22140
    },
    {
      "epoch": 5.179160332124898,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 22141
    },
    {
      "epoch": 5.179394222897907,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5808,
      "step": 22142
    },
    {
      "epoch": 5.1796281136709155,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 22143
    },
    {
      "epoch": 5.179862004443924,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 22144
    },
    {
      "epoch": 5.180095895216934,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 22145
    },
    {
      "epoch": 5.180329785989943,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 22146
    },
    {
      "epoch": 5.180563676762952,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 22147
    },
    {
      "epoch": 5.1807975675359605,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 22148
    },
    {
      "epoch": 5.181031458308969,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 22149
    },
    {
      "epoch": 5.181265349081979,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5804,
      "step": 22150
    },
    {
      "epoch": 5.181499239854988,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6099,
      "step": 22151
    },
    {
      "epoch": 5.181733130627997,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6465,
      "step": 22152
    },
    {
      "epoch": 5.181967021401006,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6857,
      "step": 22153
    },
    {
      "epoch": 5.1822009121740145,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 22154
    },
    {
      "epoch": 5.182434802947023,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.5405,
      "step": 22155
    },
    {
      "epoch": 5.182668693720033,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 22156
    },
    {
      "epoch": 5.182902584493042,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8318,
      "step": 22157
    },
    {
      "epoch": 5.183136475266051,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6504,
      "step": 22158
    },
    {
      "epoch": 5.1833703660390595,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 22159
    },
    {
      "epoch": 5.183604256812068,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 22160
    },
    {
      "epoch": 5.183838147585078,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.623,
      "step": 22161
    },
    {
      "epoch": 5.184072038358087,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8075,
      "step": 22162
    },
    {
      "epoch": 5.184305929131096,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6117,
      "step": 22163
    },
    {
      "epoch": 5.184539819904105,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7468,
      "step": 22164
    },
    {
      "epoch": 5.1847737106771135,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 22165
    },
    {
      "epoch": 5.185007601450123,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 22166
    },
    {
      "epoch": 5.185241492223132,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 22167
    },
    {
      "epoch": 5.185475382996141,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4848,
      "step": 22168
    },
    {
      "epoch": 5.18570927376915,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 22169
    },
    {
      "epoch": 5.1859431645421585,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 22170
    },
    {
      "epoch": 5.186177055315168,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 22171
    },
    {
      "epoch": 5.186410946088177,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.78,
      "step": 22172
    },
    {
      "epoch": 5.186644836861186,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 22173
    },
    {
      "epoch": 5.186878727634195,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 22174
    },
    {
      "epoch": 5.187112618407204,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 22175
    },
    {
      "epoch": 5.1873465091802125,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 22176
    },
    {
      "epoch": 5.187580399953222,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7444,
      "step": 22177
    },
    {
      "epoch": 5.187814290726231,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6651,
      "step": 22178
    },
    {
      "epoch": 5.18804818149924,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7178,
      "step": 22179
    },
    {
      "epoch": 5.188282072272249,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 22180
    },
    {
      "epoch": 5.1885159630452575,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6327,
      "step": 22181
    },
    {
      "epoch": 5.188749853818267,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 22182
    },
    {
      "epoch": 5.188983744591276,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 22183
    },
    {
      "epoch": 5.189217635364285,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4243,
      "step": 22184
    },
    {
      "epoch": 5.189451526137294,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 22185
    },
    {
      "epoch": 5.189685416910303,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3874,
      "step": 22186
    },
    {
      "epoch": 5.1899193076833114,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 22187
    },
    {
      "epoch": 5.190153198456321,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 22188
    },
    {
      "epoch": 5.19038708922933,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 22189
    },
    {
      "epoch": 5.190620980002339,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 22190
    },
    {
      "epoch": 5.190854870775348,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 22191
    },
    {
      "epoch": 5.1910887615483565,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 22192
    },
    {
      "epoch": 5.191322652321366,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7643,
      "step": 22193
    },
    {
      "epoch": 5.191556543094375,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 22194
    },
    {
      "epoch": 5.191790433867384,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 22195
    },
    {
      "epoch": 5.192024324640393,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6174,
      "step": 22196
    },
    {
      "epoch": 5.192258215413402,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 22197
    },
    {
      "epoch": 5.192492106186411,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6893,
      "step": 22198
    },
    {
      "epoch": 5.19272599695942,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 22199
    },
    {
      "epoch": 5.192959887732429,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 22200
    },
    {
      "epoch": 5.192959887732429,
      "eval_runtime": 4.6117,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 22200
    },
    {
      "epoch": 5.193193778505438,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7135,
      "step": 22201
    },
    {
      "epoch": 5.193427669278447,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4539,
      "step": 22202
    },
    {
      "epoch": 5.193661560051456,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5116,
      "step": 22203
    },
    {
      "epoch": 5.193895450824465,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6868,
      "step": 22204
    },
    {
      "epoch": 5.194129341597474,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.73,
      "step": 22205
    },
    {
      "epoch": 5.194363232370483,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6409,
      "step": 22206
    },
    {
      "epoch": 5.194597123143492,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6278,
      "step": 22207
    },
    {
      "epoch": 5.194831013916501,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4497,
      "step": 22208
    },
    {
      "epoch": 5.19506490468951,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6469,
      "step": 22209
    },
    {
      "epoch": 5.195298795462519,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 22210
    },
    {
      "epoch": 5.195532686235528,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 22211
    },
    {
      "epoch": 5.195766577008537,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 22212
    },
    {
      "epoch": 5.196000467781546,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 22213
    },
    {
      "epoch": 5.196234358554555,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 22214
    },
    {
      "epoch": 5.196468249327564,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 22215
    },
    {
      "epoch": 5.196702140100573,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7411,
      "step": 22216
    },
    {
      "epoch": 5.196936030873582,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6997,
      "step": 22217
    },
    {
      "epoch": 5.197169921646591,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6313,
      "step": 22218
    },
    {
      "epoch": 5.1974038124196,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 22219
    },
    {
      "epoch": 5.197637703192609,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 22220
    },
    {
      "epoch": 5.197871593965618,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9824,
      "step": 22221
    },
    {
      "epoch": 5.198105484738627,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 22222
    },
    {
      "epoch": 5.198339375511636,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 22223
    },
    {
      "epoch": 5.198573266284645,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 22224
    },
    {
      "epoch": 5.198807157057654,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 22225
    },
    {
      "epoch": 5.199041047830663,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.646,
      "step": 22226
    },
    {
      "epoch": 5.199274938603672,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 22227
    },
    {
      "epoch": 5.199508829376681,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 22228
    },
    {
      "epoch": 5.19974272014969,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 22229
    },
    {
      "epoch": 5.1999766109226995,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5119,
      "step": 22230
    },
    {
      "epoch": 5.200210501695708,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 22231
    },
    {
      "epoch": 5.200444392468717,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 22232
    },
    {
      "epoch": 5.200678283241726,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 22233
    },
    {
      "epoch": 5.200912174014735,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 22234
    },
    {
      "epoch": 5.2011460647877445,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 22235
    },
    {
      "epoch": 5.201379955560753,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6345,
      "step": 22236
    },
    {
      "epoch": 5.201613846333762,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 22237
    },
    {
      "epoch": 5.201847737106771,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 22238
    },
    {
      "epoch": 5.20208162787978,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 22239
    },
    {
      "epoch": 5.202315518652789,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 22240
    },
    {
      "epoch": 5.2025494094257985,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 22241
    },
    {
      "epoch": 5.202783300198807,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 22242
    },
    {
      "epoch": 5.203017190971816,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 22243
    },
    {
      "epoch": 5.203251081744825,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 22244
    },
    {
      "epoch": 5.203484972517834,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7075,
      "step": 22245
    },
    {
      "epoch": 5.2037188632908435,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 22246
    },
    {
      "epoch": 5.203952754063852,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 22247
    },
    {
      "epoch": 5.204186644836861,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 22248
    },
    {
      "epoch": 5.20442053560987,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 22249
    },
    {
      "epoch": 5.204654426382879,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 22250
    },
    {
      "epoch": 5.204888317155889,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 22251
    },
    {
      "epoch": 5.2051222079288975,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 22252
    },
    {
      "epoch": 5.205356098701906,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 22253
    },
    {
      "epoch": 5.205589989474915,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6765,
      "step": 22254
    },
    {
      "epoch": 5.205823880247924,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7095,
      "step": 22255
    },
    {
      "epoch": 5.206057771020933,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 22256
    },
    {
      "epoch": 5.2062916617939425,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5941,
      "step": 22257
    },
    {
      "epoch": 5.206525552566951,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 22258
    },
    {
      "epoch": 5.20675944333996,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 22259
    },
    {
      "epoch": 5.206993334112969,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8456,
      "step": 22260
    },
    {
      "epoch": 5.207227224885978,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6108,
      "step": 22261
    },
    {
      "epoch": 5.207461115658988,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 22262
    },
    {
      "epoch": 5.2076950064319965,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.576,
      "step": 22263
    },
    {
      "epoch": 5.207928897205005,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6394,
      "step": 22264
    },
    {
      "epoch": 5.208162787978014,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 22265
    },
    {
      "epoch": 5.208396678751023,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 22266
    },
    {
      "epoch": 5.208630569524033,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6483,
      "step": 22267
    },
    {
      "epoch": 5.2088644602970415,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 22268
    },
    {
      "epoch": 5.20909835107005,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 22269
    },
    {
      "epoch": 5.209332241843059,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.571,
      "step": 22270
    },
    {
      "epoch": 5.209566132616068,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 22271
    },
    {
      "epoch": 5.209800023389077,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 22272
    },
    {
      "epoch": 5.210033914162087,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5973,
      "step": 22273
    },
    {
      "epoch": 5.2102678049350954,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.951,
      "step": 22274
    },
    {
      "epoch": 5.210501695708104,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 22275
    },
    {
      "epoch": 5.210735586481113,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.819,
      "step": 22276
    },
    {
      "epoch": 5.210969477254122,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.557,
      "step": 22277
    },
    {
      "epoch": 5.211203368027132,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 22278
    },
    {
      "epoch": 5.2114372588001405,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 22279
    },
    {
      "epoch": 5.211671149573149,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7767,
      "step": 22280
    },
    {
      "epoch": 5.211905040346158,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8175,
      "step": 22281
    },
    {
      "epoch": 5.212138931119167,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 22282
    },
    {
      "epoch": 5.212372821892177,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 22283
    },
    {
      "epoch": 5.212606712665186,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 22284
    },
    {
      "epoch": 5.212840603438194,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 22285
    },
    {
      "epoch": 5.213074494211203,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 22286
    },
    {
      "epoch": 5.213308384984212,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7113,
      "step": 22287
    },
    {
      "epoch": 5.213542275757221,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 22288
    },
    {
      "epoch": 5.213776166530231,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7083,
      "step": 22289
    },
    {
      "epoch": 5.2140100573032395,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 22290
    },
    {
      "epoch": 5.214243948076248,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 22291
    },
    {
      "epoch": 5.214477838849257,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 22292
    },
    {
      "epoch": 5.214711729622266,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8434,
      "step": 22293
    },
    {
      "epoch": 5.214945620395276,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 22294
    },
    {
      "epoch": 5.215179511168285,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 22295
    },
    {
      "epoch": 5.215413401941293,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 22296
    },
    {
      "epoch": 5.215647292714302,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 22297
    },
    {
      "epoch": 5.215881183487311,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6561,
      "step": 22298
    },
    {
      "epoch": 5.216115074260321,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5675,
      "step": 22299
    },
    {
      "epoch": 5.21634896503333,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 22300
    },
    {
      "epoch": 5.21634896503333,
      "eval_runtime": 4.578,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 22300
    },
    {
      "epoch": 5.2165828558063385,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 22301
    },
    {
      "epoch": 5.216816746579347,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 22302
    },
    {
      "epoch": 5.217050637352356,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8196,
      "step": 22303
    },
    {
      "epoch": 5.217284528125365,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 22304
    },
    {
      "epoch": 5.217518418898375,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7897,
      "step": 22305
    },
    {
      "epoch": 5.217752309671384,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 22306
    },
    {
      "epoch": 5.217986200444392,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5048,
      "step": 22307
    },
    {
      "epoch": 5.218220091217401,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4687,
      "step": 22308
    },
    {
      "epoch": 5.21845398199041,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5144,
      "step": 22309
    },
    {
      "epoch": 5.21868787276342,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 22310
    },
    {
      "epoch": 5.218921763536429,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 22311
    },
    {
      "epoch": 5.2191556543094375,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.3415,
      "step": 22312
    },
    {
      "epoch": 5.219389545082446,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7296,
      "step": 22313
    },
    {
      "epoch": 5.219623435855455,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6738,
      "step": 22314
    },
    {
      "epoch": 5.219857326628465,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7171,
      "step": 22315
    },
    {
      "epoch": 5.220091217401474,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 22316
    },
    {
      "epoch": 5.220325108174483,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 22317
    },
    {
      "epoch": 5.220558998947491,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6465,
      "step": 22318
    },
    {
      "epoch": 5.2207928897205,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 22319
    },
    {
      "epoch": 5.22102678049351,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7236,
      "step": 22320
    },
    {
      "epoch": 5.221260671266519,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 22321
    },
    {
      "epoch": 5.221494562039528,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 22322
    },
    {
      "epoch": 5.2217284528125365,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8359,
      "step": 22323
    },
    {
      "epoch": 5.221962343585545,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6306,
      "step": 22324
    },
    {
      "epoch": 5.222196234358554,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 22325
    },
    {
      "epoch": 5.222430125131564,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 22326
    },
    {
      "epoch": 5.222664015904573,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5379,
      "step": 22327
    },
    {
      "epoch": 5.222897906677582,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 22328
    },
    {
      "epoch": 5.22313179745059,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 22329
    },
    {
      "epoch": 5.223365688223599,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7527,
      "step": 22330
    },
    {
      "epoch": 5.223599578996609,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 22331
    },
    {
      "epoch": 5.223833469769618,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 22332
    },
    {
      "epoch": 5.224067360542627,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7826,
      "step": 22333
    },
    {
      "epoch": 5.2243012513156355,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 22334
    },
    {
      "epoch": 5.224535142088644,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8801,
      "step": 22335
    },
    {
      "epoch": 5.224769032861653,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 22336
    },
    {
      "epoch": 5.225002923634663,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 22337
    },
    {
      "epoch": 5.225236814407672,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7268,
      "step": 22338
    },
    {
      "epoch": 5.225470705180681,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6817,
      "step": 22339
    },
    {
      "epoch": 5.225704595953689,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6272,
      "step": 22340
    },
    {
      "epoch": 5.225938486726698,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 22341
    },
    {
      "epoch": 5.226172377499708,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 22342
    },
    {
      "epoch": 5.226406268272717,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5522,
      "step": 22343
    },
    {
      "epoch": 5.226640159045726,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 22344
    },
    {
      "epoch": 5.2268740498187345,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 22345
    },
    {
      "epoch": 5.227107940591743,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7615,
      "step": 22346
    },
    {
      "epoch": 5.227341831364753,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 22347
    },
    {
      "epoch": 5.227575722137762,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 22348
    },
    {
      "epoch": 5.227809612910771,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 22349
    },
    {
      "epoch": 5.22804350368378,
      "grad_norm": 14.1875,
      "learning_rate": 3e-05,
      "loss": 2.3432,
      "step": 22350
    },
    {
      "epoch": 5.228277394456788,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6551,
      "step": 22351
    },
    {
      "epoch": 5.228511285229798,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 22352
    },
    {
      "epoch": 5.228745176002807,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 22353
    },
    {
      "epoch": 5.228979066775816,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 22354
    },
    {
      "epoch": 5.229212957548825,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 22355
    },
    {
      "epoch": 5.2294468483218335,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5565,
      "step": 22356
    },
    {
      "epoch": 5.229680739094842,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 22357
    },
    {
      "epoch": 5.229914629867852,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 22358
    },
    {
      "epoch": 5.230148520640861,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 22359
    },
    {
      "epoch": 5.23038241141387,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 22360
    },
    {
      "epoch": 5.230616302186879,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6507,
      "step": 22361
    },
    {
      "epoch": 5.230850192959887,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.616,
      "step": 22362
    },
    {
      "epoch": 5.231084083732897,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5014,
      "step": 22363
    },
    {
      "epoch": 5.231317974505906,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5137,
      "step": 22364
    },
    {
      "epoch": 5.231551865278915,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5957,
      "step": 22365
    },
    {
      "epoch": 5.231785756051924,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 22366
    },
    {
      "epoch": 5.2320196468249325,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 22367
    },
    {
      "epoch": 5.232253537597941,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 22368
    },
    {
      "epoch": 5.232487428370951,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 22369
    },
    {
      "epoch": 5.23272131914396,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.5455,
      "step": 22370
    },
    {
      "epoch": 5.232955209916969,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.4261,
      "step": 22371
    },
    {
      "epoch": 5.233189100689978,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 22372
    },
    {
      "epoch": 5.233422991462986,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0032,
      "step": 22373
    },
    {
      "epoch": 5.233656882235996,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7485,
      "step": 22374
    },
    {
      "epoch": 5.233890773009005,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6727,
      "step": 22375
    },
    {
      "epoch": 5.234124663782014,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6219,
      "step": 22376
    },
    {
      "epoch": 5.234358554555023,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 22377
    },
    {
      "epoch": 5.2345924453280315,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5565,
      "step": 22378
    },
    {
      "epoch": 5.234826336101041,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 22379
    },
    {
      "epoch": 5.23506022687405,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 22380
    },
    {
      "epoch": 5.235294117647059,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 22381
    },
    {
      "epoch": 5.235528008420068,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 22382
    },
    {
      "epoch": 5.2357618991930766,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6619,
      "step": 22383
    },
    {
      "epoch": 5.235995789966086,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 22384
    },
    {
      "epoch": 5.236229680739095,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 22385
    },
    {
      "epoch": 5.236463571512104,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 22386
    },
    {
      "epoch": 5.236697462285113,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 22387
    },
    {
      "epoch": 5.236931353058122,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.532,
      "step": 22388
    },
    {
      "epoch": 5.2371652438311305,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 22389
    },
    {
      "epoch": 5.23739913460414,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 22390
    },
    {
      "epoch": 5.237633025377149,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 22391
    },
    {
      "epoch": 5.237866916150158,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5394,
      "step": 22392
    },
    {
      "epoch": 5.238100806923167,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 22393
    },
    {
      "epoch": 5.2383346976961755,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6251,
      "step": 22394
    },
    {
      "epoch": 5.238568588469185,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 22395
    },
    {
      "epoch": 5.238802479242194,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 22396
    },
    {
      "epoch": 5.239036370015203,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 22397
    },
    {
      "epoch": 5.239270260788212,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 22398
    },
    {
      "epoch": 5.239504151561221,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 22399
    },
    {
      "epoch": 5.2397380423342295,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 22400
    },
    {
      "epoch": 5.2397380423342295,
      "eval_runtime": 4.6129,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 22400
    },
    {
      "epoch": 5.239971933107239,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 22401
    },
    {
      "epoch": 5.240205823880248,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 22402
    },
    {
      "epoch": 5.240439714653257,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6304,
      "step": 22403
    },
    {
      "epoch": 5.240673605426266,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 22404
    },
    {
      "epoch": 5.2409074961992745,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5569,
      "step": 22405
    },
    {
      "epoch": 5.241141386972284,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 22406
    },
    {
      "epoch": 5.241375277745293,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6489,
      "step": 22407
    },
    {
      "epoch": 5.241609168518302,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 22408
    },
    {
      "epoch": 5.241843059291311,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8132,
      "step": 22409
    },
    {
      "epoch": 5.24207695006432,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6945,
      "step": 22410
    },
    {
      "epoch": 5.242310840837329,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 22411
    },
    {
      "epoch": 5.242544731610338,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 22412
    },
    {
      "epoch": 5.242778622383347,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 22413
    },
    {
      "epoch": 5.243012513156356,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 22414
    },
    {
      "epoch": 5.243246403929365,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 22415
    },
    {
      "epoch": 5.243480294702374,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 22416
    },
    {
      "epoch": 5.243714185475383,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 22417
    },
    {
      "epoch": 5.243948076248392,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 22418
    },
    {
      "epoch": 5.244181967021401,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 22419
    },
    {
      "epoch": 5.24441585779441,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 22420
    },
    {
      "epoch": 5.244649748567419,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 22421
    },
    {
      "epoch": 5.244883639340428,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 22422
    },
    {
      "epoch": 5.245117530113437,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7298,
      "step": 22423
    },
    {
      "epoch": 5.245351420886446,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7472,
      "step": 22424
    },
    {
      "epoch": 5.245585311659455,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 22425
    },
    {
      "epoch": 5.245819202432464,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5498,
      "step": 22426
    },
    {
      "epoch": 5.246053093205473,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4973,
      "step": 22427
    },
    {
      "epoch": 5.246286983978482,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.4604,
      "step": 22428
    },
    {
      "epoch": 5.246520874751491,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 22429
    },
    {
      "epoch": 5.2467547655245,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 22430
    },
    {
      "epoch": 5.246988656297509,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 22431
    },
    {
      "epoch": 5.2472225470705185,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 22432
    },
    {
      "epoch": 5.247456437843527,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 22433
    },
    {
      "epoch": 5.247690328616536,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 22434
    },
    {
      "epoch": 5.247924219389545,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 22435
    },
    {
      "epoch": 5.248158110162554,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.4647,
      "step": 22436
    },
    {
      "epoch": 5.248392000935563,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7528,
      "step": 22437
    },
    {
      "epoch": 5.248625891708572,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 22438
    },
    {
      "epoch": 5.248859782481581,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 22439
    },
    {
      "epoch": 5.24909367325459,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6436,
      "step": 22440
    },
    {
      "epoch": 5.249327564027599,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 22441
    },
    {
      "epoch": 5.249561454800608,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 22442
    },
    {
      "epoch": 5.2497953455736175,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6295,
      "step": 22443
    },
    {
      "epoch": 5.250029236346626,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.64,
      "step": 22444
    },
    {
      "epoch": 5.250263127119635,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 22445
    },
    {
      "epoch": 5.250497017892644,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 22446
    },
    {
      "epoch": 5.250730908665653,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 22447
    },
    {
      "epoch": 5.250964799438663,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 22448
    },
    {
      "epoch": 5.251198690211671,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5922,
      "step": 22449
    },
    {
      "epoch": 5.25143258098468,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6872,
      "step": 22450
    },
    {
      "epoch": 5.251666471757689,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 22451
    },
    {
      "epoch": 5.251900362530698,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 22452
    },
    {
      "epoch": 5.252134253303707,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 22453
    },
    {
      "epoch": 5.2523681440767165,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 22454
    },
    {
      "epoch": 5.252602034849725,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 22455
    },
    {
      "epoch": 5.252835925622734,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5841,
      "step": 22456
    },
    {
      "epoch": 5.253069816395743,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 22457
    },
    {
      "epoch": 5.253303707168752,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 22458
    },
    {
      "epoch": 5.253537597941762,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6852,
      "step": 22459
    },
    {
      "epoch": 5.25377148871477,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 22460
    },
    {
      "epoch": 5.254005379487779,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 22461
    },
    {
      "epoch": 5.254239270260788,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7527,
      "step": 22462
    },
    {
      "epoch": 5.254473161033797,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 22463
    },
    {
      "epoch": 5.254707051806806,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 22464
    },
    {
      "epoch": 5.2549409425798155,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 22465
    },
    {
      "epoch": 5.255174833352824,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 22466
    },
    {
      "epoch": 5.255408724125833,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7757,
      "step": 22467
    },
    {
      "epoch": 5.255642614898842,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 22468
    },
    {
      "epoch": 5.255876505671852,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 22469
    },
    {
      "epoch": 5.2561103964448606,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6887,
      "step": 22470
    },
    {
      "epoch": 5.256344287217869,
      "grad_norm": 7.6875,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 22471
    },
    {
      "epoch": 5.256578177990878,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6256,
      "step": 22472
    },
    {
      "epoch": 5.256812068763887,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 22473
    },
    {
      "epoch": 5.257045959536896,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 22474
    },
    {
      "epoch": 5.257279850309906,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 22475
    },
    {
      "epoch": 5.2575137410829145,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6288,
      "step": 22476
    },
    {
      "epoch": 5.257747631855923,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7337,
      "step": 22477
    },
    {
      "epoch": 5.257981522628932,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 22478
    },
    {
      "epoch": 5.258215413401941,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8151,
      "step": 22479
    },
    {
      "epoch": 5.258449304174951,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7507,
      "step": 22480
    },
    {
      "epoch": 5.2586831949479595,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 22481
    },
    {
      "epoch": 5.258917085720968,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 22482
    },
    {
      "epoch": 5.259150976493977,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 22483
    },
    {
      "epoch": 5.259384867266986,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 22484
    },
    {
      "epoch": 5.259618758039995,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 22485
    },
    {
      "epoch": 5.259852648813005,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 22486
    },
    {
      "epoch": 5.2600865395860135,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 22487
    },
    {
      "epoch": 5.260320430359022,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 22488
    },
    {
      "epoch": 5.260554321132031,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.6957,
      "step": 22489
    },
    {
      "epoch": 5.26078821190504,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 22490
    },
    {
      "epoch": 5.26102210267805,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 22491
    },
    {
      "epoch": 5.2612559934510585,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 22492
    },
    {
      "epoch": 5.261489884224067,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 22493
    },
    {
      "epoch": 5.261723774997076,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 22494
    },
    {
      "epoch": 5.261957665770085,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.574,
      "step": 22495
    },
    {
      "epoch": 5.262191556543095,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8657,
      "step": 22496
    },
    {
      "epoch": 5.262425447316104,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 22497
    },
    {
      "epoch": 5.2626593380891125,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.4713,
      "step": 22498
    },
    {
      "epoch": 5.262893228862121,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 22499
    },
    {
      "epoch": 5.26312711963513,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 22500
    },
    {
      "epoch": 5.26312711963513,
      "eval_runtime": 4.6036,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 22500
    },
    {
      "epoch": 5.26336101040814,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 22501
    },
    {
      "epoch": 5.263594901181149,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.648,
      "step": 22502
    },
    {
      "epoch": 5.2638287919541575,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 22503
    },
    {
      "epoch": 5.264062682727166,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 22504
    },
    {
      "epoch": 5.264296573500175,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6636,
      "step": 22505
    },
    {
      "epoch": 5.264530464273184,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6409,
      "step": 22506
    },
    {
      "epoch": 5.264764355046194,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 22507
    },
    {
      "epoch": 5.264998245819203,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 22508
    },
    {
      "epoch": 5.2652321365922115,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 22509
    },
    {
      "epoch": 5.26546602736522,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 22510
    },
    {
      "epoch": 5.265699918138229,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4936,
      "step": 22511
    },
    {
      "epoch": 5.265933808911239,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 22512
    },
    {
      "epoch": 5.266167699684248,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5612,
      "step": 22513
    },
    {
      "epoch": 5.2664015904572565,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6736,
      "step": 22514
    },
    {
      "epoch": 5.266635481230265,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5818,
      "step": 22515
    },
    {
      "epoch": 5.266869372003274,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4453,
      "step": 22516
    },
    {
      "epoch": 5.267103262776283,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7876,
      "step": 22517
    },
    {
      "epoch": 5.267337153549293,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 22518
    },
    {
      "epoch": 5.267571044322302,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 22519
    },
    {
      "epoch": 5.2678049350953104,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 22520
    },
    {
      "epoch": 5.268038825868319,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 22521
    },
    {
      "epoch": 5.268272716641328,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 22522
    },
    {
      "epoch": 5.268506607414338,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 22523
    },
    {
      "epoch": 5.268740498187347,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5854,
      "step": 22524
    },
    {
      "epoch": 5.2689743889603555,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.694,
      "step": 22525
    },
    {
      "epoch": 5.269208279733364,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 22526
    },
    {
      "epoch": 5.269442170506373,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 22527
    },
    {
      "epoch": 5.269676061279383,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 22528
    },
    {
      "epoch": 5.269909952052392,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 22529
    },
    {
      "epoch": 5.270143842825401,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 22530
    },
    {
      "epoch": 5.270377733598409,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8349,
      "step": 22531
    },
    {
      "epoch": 5.270611624371418,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6733,
      "step": 22532
    },
    {
      "epoch": 5.270845515144428,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 22533
    },
    {
      "epoch": 5.271079405917437,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.7163,
      "step": 22534
    },
    {
      "epoch": 5.271313296690446,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 22535
    },
    {
      "epoch": 5.2715471874634545,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 22536
    },
    {
      "epoch": 5.271781078236463,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 22537
    },
    {
      "epoch": 5.272014969009472,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.4396,
      "step": 22538
    },
    {
      "epoch": 5.272248859782482,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 22539
    },
    {
      "epoch": 5.272482750555491,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 22540
    },
    {
      "epoch": 5.2727166413285,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 22541
    },
    {
      "epoch": 5.272950532101508,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9638,
      "step": 22542
    },
    {
      "epoch": 5.273184422874517,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 22543
    },
    {
      "epoch": 5.273418313647527,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7403,
      "step": 22544
    },
    {
      "epoch": 5.273652204420536,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6988,
      "step": 22545
    },
    {
      "epoch": 5.273886095193545,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 22546
    },
    {
      "epoch": 5.2741199859665535,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 22547
    },
    {
      "epoch": 5.274353876739562,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 22548
    },
    {
      "epoch": 5.274587767512571,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 22549
    },
    {
      "epoch": 5.274821658285581,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 22550
    },
    {
      "epoch": 5.27505554905859,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 22551
    },
    {
      "epoch": 5.275289439831599,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 22552
    },
    {
      "epoch": 5.275523330604607,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 22553
    },
    {
      "epoch": 5.275757221377616,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6261,
      "step": 22554
    },
    {
      "epoch": 5.275991112150626,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 22555
    },
    {
      "epoch": 5.276225002923635,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 22556
    },
    {
      "epoch": 5.276458893696644,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7855,
      "step": 22557
    },
    {
      "epoch": 5.2766927844696525,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6494,
      "step": 22558
    },
    {
      "epoch": 5.276926675242661,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6144,
      "step": 22559
    },
    {
      "epoch": 5.277160566015671,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 22560
    },
    {
      "epoch": 5.27739445678868,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6563,
      "step": 22561
    },
    {
      "epoch": 5.277628347561689,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4588,
      "step": 22562
    },
    {
      "epoch": 5.277862238334698,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 22563
    },
    {
      "epoch": 5.278096129107706,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 22564
    },
    {
      "epoch": 5.278330019880716,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6585,
      "step": 22565
    },
    {
      "epoch": 5.278563910653725,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6743,
      "step": 22566
    },
    {
      "epoch": 5.278797801426734,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 22567
    },
    {
      "epoch": 5.279031692199743,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 22568
    },
    {
      "epoch": 5.2792655829727515,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 22569
    },
    {
      "epoch": 5.27949947374576,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 22570
    },
    {
      "epoch": 5.27973336451877,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 22571
    },
    {
      "epoch": 5.279967255291779,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 22572
    },
    {
      "epoch": 5.280201146064788,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8303,
      "step": 22573
    },
    {
      "epoch": 5.280435036837797,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 22574
    },
    {
      "epoch": 5.280668927610805,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 22575
    },
    {
      "epoch": 5.280902818383815,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.652,
      "step": 22576
    },
    {
      "epoch": 5.281136709156824,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8061,
      "step": 22577
    },
    {
      "epoch": 5.281370599929833,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 22578
    },
    {
      "epoch": 5.281604490702842,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 22579
    },
    {
      "epoch": 5.2818383814758505,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6381,
      "step": 22580
    },
    {
      "epoch": 5.282072272248859,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6836,
      "step": 22581
    },
    {
      "epoch": 5.282306163021869,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 22582
    },
    {
      "epoch": 5.282540053794878,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 22583
    },
    {
      "epoch": 5.282773944567887,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 22584
    },
    {
      "epoch": 5.283007835340896,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 22585
    },
    {
      "epoch": 5.283241726113904,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 22586
    },
    {
      "epoch": 5.283475616886914,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 22587
    },
    {
      "epoch": 5.283709507659923,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 22588
    },
    {
      "epoch": 5.283943398432932,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.64,
      "step": 22589
    },
    {
      "epoch": 5.284177289205941,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6138,
      "step": 22590
    },
    {
      "epoch": 5.2844111799789495,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5525,
      "step": 22591
    },
    {
      "epoch": 5.284645070751959,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 22592
    },
    {
      "epoch": 5.284878961524968,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 22593
    },
    {
      "epoch": 5.285112852297977,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 22594
    },
    {
      "epoch": 5.285346743070986,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 22595
    },
    {
      "epoch": 5.285580633843995,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.415,
      "step": 22596
    },
    {
      "epoch": 5.285814524617004,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7447,
      "step": 22597
    },
    {
      "epoch": 5.286048415390013,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 22598
    },
    {
      "epoch": 5.286282306163022,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 22599
    },
    {
      "epoch": 5.286516196936031,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 22600
    },
    {
      "epoch": 5.286516196936031,
      "eval_runtime": 4.6392,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 22600
    },
    {
      "epoch": 5.28675008770904,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 22601
    },
    {
      "epoch": 5.2869839784820485,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7969,
      "step": 22602
    },
    {
      "epoch": 5.287217869255058,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7067,
      "step": 22603
    },
    {
      "epoch": 5.287451760028067,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7509,
      "step": 22604
    },
    {
      "epoch": 5.287685650801076,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7255,
      "step": 22605
    },
    {
      "epoch": 5.287919541574085,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6776,
      "step": 22606
    },
    {
      "epoch": 5.288153432347094,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8354,
      "step": 22607
    },
    {
      "epoch": 5.288387323120103,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6613,
      "step": 22608
    },
    {
      "epoch": 5.288621213893112,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5707,
      "step": 22609
    },
    {
      "epoch": 5.288855104666121,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 22610
    },
    {
      "epoch": 5.28908899543913,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7207,
      "step": 22611
    },
    {
      "epoch": 5.289322886212139,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 22612
    },
    {
      "epoch": 5.2895567769851475,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 22613
    },
    {
      "epoch": 5.289790667758157,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9998,
      "step": 22614
    },
    {
      "epoch": 5.290024558531166,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 22615
    },
    {
      "epoch": 5.290258449304175,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6648,
      "step": 22616
    },
    {
      "epoch": 5.290492340077184,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 22617
    },
    {
      "epoch": 5.290726230850193,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 22618
    },
    {
      "epoch": 5.290960121623202,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 22619
    },
    {
      "epoch": 5.291194012396211,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 22620
    },
    {
      "epoch": 5.29142790316922,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 22621
    },
    {
      "epoch": 5.291661793942229,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6975,
      "step": 22622
    },
    {
      "epoch": 5.291895684715238,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5376,
      "step": 22623
    },
    {
      "epoch": 5.292129575488247,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 22624
    },
    {
      "epoch": 5.292363466261256,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 22625
    },
    {
      "epoch": 5.292597357034265,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 22626
    },
    {
      "epoch": 5.292831247807274,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.442,
      "step": 22627
    },
    {
      "epoch": 5.293065138580283,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5062,
      "step": 22628
    },
    {
      "epoch": 5.293299029353292,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 22629
    },
    {
      "epoch": 5.293532920126301,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5859,
      "step": 22630
    },
    {
      "epoch": 5.29376681089931,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 22631
    },
    {
      "epoch": 5.294000701672319,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6195,
      "step": 22632
    },
    {
      "epoch": 5.294234592445328,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 22633
    },
    {
      "epoch": 5.294468483218337,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 22634
    },
    {
      "epoch": 5.294702373991346,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 22635
    },
    {
      "epoch": 5.294936264764355,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 22636
    },
    {
      "epoch": 5.295170155537364,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 22637
    },
    {
      "epoch": 5.295404046310373,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7802,
      "step": 22638
    },
    {
      "epoch": 5.295637937083382,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 22639
    },
    {
      "epoch": 5.295871827856391,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7117,
      "step": 22640
    },
    {
      "epoch": 5.2961057186294,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 22641
    },
    {
      "epoch": 5.296339609402409,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 22642
    },
    {
      "epoch": 5.296573500175418,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 22643
    },
    {
      "epoch": 5.296807390948427,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 22644
    },
    {
      "epoch": 5.297041281721436,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 22645
    },
    {
      "epoch": 5.297275172494445,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 22646
    },
    {
      "epoch": 5.297509063267454,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7587,
      "step": 22647
    },
    {
      "epoch": 5.297742954040463,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 22648
    },
    {
      "epoch": 5.297976844813472,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 22649
    },
    {
      "epoch": 5.298210735586481,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 22650
    },
    {
      "epoch": 5.29844462635949,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 22651
    },
    {
      "epoch": 5.298678517132499,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 22652
    },
    {
      "epoch": 5.298912407905508,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 22653
    },
    {
      "epoch": 5.299146298678517,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 22654
    },
    {
      "epoch": 5.299380189451526,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 22655
    },
    {
      "epoch": 5.2996140802245355,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7737,
      "step": 22656
    },
    {
      "epoch": 5.299847970997544,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5454,
      "step": 22657
    },
    {
      "epoch": 5.300081861770553,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9831,
      "step": 22658
    },
    {
      "epoch": 5.300315752543562,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 22659
    },
    {
      "epoch": 5.300549643316571,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 22660
    },
    {
      "epoch": 5.300783534089581,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5407,
      "step": 22661
    },
    {
      "epoch": 5.301017424862589,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 22662
    },
    {
      "epoch": 5.301251315635598,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 22663
    },
    {
      "epoch": 5.301485206408607,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 22664
    },
    {
      "epoch": 5.301719097181616,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 22665
    },
    {
      "epoch": 5.301952987954625,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 22666
    },
    {
      "epoch": 5.3021868787276345,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.762,
      "step": 22667
    },
    {
      "epoch": 5.302420769500643,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 22668
    },
    {
      "epoch": 5.302654660273652,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 22669
    },
    {
      "epoch": 5.302888551046661,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7858,
      "step": 22670
    },
    {
      "epoch": 5.30312244181967,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 22671
    },
    {
      "epoch": 5.30335633259268,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 22672
    },
    {
      "epoch": 5.303590223365688,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 22673
    },
    {
      "epoch": 5.303824114138697,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 22674
    },
    {
      "epoch": 5.304058004911706,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 22675
    },
    {
      "epoch": 5.304291895684715,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 22676
    },
    {
      "epoch": 5.304525786457725,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 22677
    },
    {
      "epoch": 5.3047596772307335,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 22678
    },
    {
      "epoch": 5.304993568003742,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 22679
    },
    {
      "epoch": 5.305227458776751,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 22680
    },
    {
      "epoch": 5.30546134954976,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5205,
      "step": 22681
    },
    {
      "epoch": 5.30569524032277,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 22682
    },
    {
      "epoch": 5.305929131095779,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 22683
    },
    {
      "epoch": 5.306163021868787,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4715,
      "step": 22684
    },
    {
      "epoch": 5.306396912641796,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 22685
    },
    {
      "epoch": 5.306630803414805,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7166,
      "step": 22686
    },
    {
      "epoch": 5.306864694187814,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 22687
    },
    {
      "epoch": 5.307098584960824,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7232,
      "step": 22688
    },
    {
      "epoch": 5.3073324757338325,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 2.2731,
      "step": 22689
    },
    {
      "epoch": 5.307566366506841,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 22690
    },
    {
      "epoch": 5.30780025727985,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 22691
    },
    {
      "epoch": 5.308034148052859,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6059,
      "step": 22692
    },
    {
      "epoch": 5.308268038825869,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6733,
      "step": 22693
    },
    {
      "epoch": 5.308501929598878,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5257,
      "step": 22694
    },
    {
      "epoch": 5.308735820371886,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 22695
    },
    {
      "epoch": 5.308969711144895,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7384,
      "step": 22696
    },
    {
      "epoch": 5.309203601917904,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 22697
    },
    {
      "epoch": 5.309437492690913,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 22698
    },
    {
      "epoch": 5.309671383463923,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 22699
    },
    {
      "epoch": 5.3099052742369315,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8251,
      "step": 22700
    },
    {
      "epoch": 5.3099052742369315,
      "eval_runtime": 4.6636,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 22700
    },
    {
      "epoch": 5.31013916500994,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 22701
    },
    {
      "epoch": 5.310373055782949,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 22702
    },
    {
      "epoch": 5.310606946555958,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 22703
    },
    {
      "epoch": 5.310840837328968,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 22704
    },
    {
      "epoch": 5.311074728101977,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 22705
    },
    {
      "epoch": 5.311308618874985,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 22706
    },
    {
      "epoch": 5.311542509647994,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 22707
    },
    {
      "epoch": 5.311776400421003,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 22708
    },
    {
      "epoch": 5.312010291194013,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 22709
    },
    {
      "epoch": 5.312244181967022,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 22710
    },
    {
      "epoch": 5.3124780727400305,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 22711
    },
    {
      "epoch": 5.312711963513039,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7352,
      "step": 22712
    },
    {
      "epoch": 5.312945854286048,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 22713
    },
    {
      "epoch": 5.313179745059058,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 22714
    },
    {
      "epoch": 5.313413635832067,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 22715
    },
    {
      "epoch": 5.3136475266050756,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7392,
      "step": 22716
    },
    {
      "epoch": 5.313881417378084,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 22717
    },
    {
      "epoch": 5.314115308151093,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 22718
    },
    {
      "epoch": 5.314349198924102,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 22719
    },
    {
      "epoch": 5.314583089697112,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 22720
    },
    {
      "epoch": 5.314816980470121,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 22721
    },
    {
      "epoch": 5.3150508712431295,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7147,
      "step": 22722
    },
    {
      "epoch": 5.315284762016138,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 22723
    },
    {
      "epoch": 5.315518652789147,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 22724
    },
    {
      "epoch": 5.315752543562157,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 22725
    },
    {
      "epoch": 5.315986434335166,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 22726
    },
    {
      "epoch": 5.3162203251081745,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 22727
    },
    {
      "epoch": 5.316454215881183,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 22728
    },
    {
      "epoch": 5.316688106654192,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5611,
      "step": 22729
    },
    {
      "epoch": 5.316921997427201,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5821,
      "step": 22730
    },
    {
      "epoch": 5.317155888200211,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 22731
    },
    {
      "epoch": 5.31738977897322,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.685,
      "step": 22732
    },
    {
      "epoch": 5.3176236697462285,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7411,
      "step": 22733
    },
    {
      "epoch": 5.317857560519237,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6266,
      "step": 22734
    },
    {
      "epoch": 5.318091451292246,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 22735
    },
    {
      "epoch": 5.318325342065256,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 22736
    },
    {
      "epoch": 5.318559232838265,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 22737
    },
    {
      "epoch": 5.3187931236112735,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 22738
    },
    {
      "epoch": 5.319027014384282,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 22739
    },
    {
      "epoch": 5.319260905157291,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8304,
      "step": 22740
    },
    {
      "epoch": 5.319494795930301,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 22741
    },
    {
      "epoch": 5.31972868670331,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 22742
    },
    {
      "epoch": 5.319962577476319,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7056,
      "step": 22743
    },
    {
      "epoch": 5.3201964682493275,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5253,
      "step": 22744
    },
    {
      "epoch": 5.320430359022336,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 22745
    },
    {
      "epoch": 5.320664249795346,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 22746
    },
    {
      "epoch": 5.320898140568355,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 22747
    },
    {
      "epoch": 5.321132031341364,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 22748
    },
    {
      "epoch": 5.3213659221143725,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2037,
      "step": 22749
    },
    {
      "epoch": 5.321599812887381,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 22750
    },
    {
      "epoch": 5.32183370366039,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 22751
    },
    {
      "epoch": 5.3220675944334,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 22752
    },
    {
      "epoch": 5.322301485206409,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 22753
    },
    {
      "epoch": 5.322535375979418,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 22754
    },
    {
      "epoch": 5.3227692667524265,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5371,
      "step": 22755
    },
    {
      "epoch": 5.323003157525435,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 22756
    },
    {
      "epoch": 5.323237048298445,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 22757
    },
    {
      "epoch": 5.323470939071454,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 22758
    },
    {
      "epoch": 5.323704829844463,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6295,
      "step": 22759
    },
    {
      "epoch": 5.3239387206174715,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 22760
    },
    {
      "epoch": 5.32417261139048,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 22761
    },
    {
      "epoch": 5.324406502163489,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8305,
      "step": 22762
    },
    {
      "epoch": 5.324640392936499,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 22763
    },
    {
      "epoch": 5.324874283709508,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 22764
    },
    {
      "epoch": 5.325108174482517,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.5733,
      "step": 22765
    },
    {
      "epoch": 5.3253420652555254,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 22766
    },
    {
      "epoch": 5.325575956028534,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 22767
    },
    {
      "epoch": 5.325809846801544,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 22768
    },
    {
      "epoch": 5.326043737574553,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.761,
      "step": 22769
    },
    {
      "epoch": 5.326277628347562,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6071,
      "step": 22770
    },
    {
      "epoch": 5.3265115191205705,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9642,
      "step": 22771
    },
    {
      "epoch": 5.326745409893579,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 22772
    },
    {
      "epoch": 5.326979300666589,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 22773
    },
    {
      "epoch": 5.327213191439598,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 22774
    },
    {
      "epoch": 5.327447082212607,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 22775
    },
    {
      "epoch": 5.327680972985616,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 22776
    },
    {
      "epoch": 5.327914863758624,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4932,
      "step": 22777
    },
    {
      "epoch": 5.328148754531634,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7906,
      "step": 22778
    },
    {
      "epoch": 5.328382645304643,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5548,
      "step": 22779
    },
    {
      "epoch": 5.328616536077652,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7579,
      "step": 22780
    },
    {
      "epoch": 5.328850426850661,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 22781
    },
    {
      "epoch": 5.3290843176236695,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 22782
    },
    {
      "epoch": 5.329318208396678,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 22783
    },
    {
      "epoch": 5.329552099169688,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 22784
    },
    {
      "epoch": 5.329785989942697,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 22785
    },
    {
      "epoch": 5.330019880715706,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.726,
      "step": 22786
    },
    {
      "epoch": 5.330253771488715,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 22787
    },
    {
      "epoch": 5.330487662261723,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 22788
    },
    {
      "epoch": 5.330721553034733,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6211,
      "step": 22789
    },
    {
      "epoch": 5.330955443807742,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6592,
      "step": 22790
    },
    {
      "epoch": 5.331189334580751,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7228,
      "step": 22791
    },
    {
      "epoch": 5.33142322535376,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.48,
      "step": 22792
    },
    {
      "epoch": 5.3316571161267685,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 22793
    },
    {
      "epoch": 5.331891006899777,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 22794
    },
    {
      "epoch": 5.332124897672787,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 22795
    },
    {
      "epoch": 5.332358788445796,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 22796
    },
    {
      "epoch": 5.332592679218805,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 22797
    },
    {
      "epoch": 5.332826569991814,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 22798
    },
    {
      "epoch": 5.333060460764822,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.534,
      "step": 22799
    },
    {
      "epoch": 5.333294351537832,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5339,
      "step": 22800
    },
    {
      "epoch": 5.333294351537832,
      "eval_runtime": 4.6456,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 22800
    },
    {
      "epoch": 5.333528242310841,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6879,
      "step": 22801
    },
    {
      "epoch": 5.33376213308385,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 22802
    },
    {
      "epoch": 5.333996023856859,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 22803
    },
    {
      "epoch": 5.3342299146298675,
      "grad_norm": 7.90625,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 22804
    },
    {
      "epoch": 5.334463805402877,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 22805
    },
    {
      "epoch": 5.334697696175886,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 22806
    },
    {
      "epoch": 5.334931586948895,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7509,
      "step": 22807
    },
    {
      "epoch": 5.335165477721904,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7827,
      "step": 22808
    },
    {
      "epoch": 5.335399368494913,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 22809
    },
    {
      "epoch": 5.335633259267922,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 22810
    },
    {
      "epoch": 5.335867150040931,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5613,
      "step": 22811
    },
    {
      "epoch": 5.33610104081394,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.7609,
      "step": 22812
    },
    {
      "epoch": 5.336334931586949,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 22813
    },
    {
      "epoch": 5.336568822359958,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 22814
    },
    {
      "epoch": 5.3368027131329665,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 22815
    },
    {
      "epoch": 5.337036603905976,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 22816
    },
    {
      "epoch": 5.337270494678985,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 22817
    },
    {
      "epoch": 5.337504385451994,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 22818
    },
    {
      "epoch": 5.337738276225003,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8363,
      "step": 22819
    },
    {
      "epoch": 5.337972166998012,
      "grad_norm": 2.453125,
      "learning_rate": 3e-05,
      "loss": 1.4948,
      "step": 22820
    },
    {
      "epoch": 5.338206057771021,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 22821
    },
    {
      "epoch": 5.33843994854403,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 22822
    },
    {
      "epoch": 5.338673839317039,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 22823
    },
    {
      "epoch": 5.338907730090048,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 22824
    },
    {
      "epoch": 5.339141620863057,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 22825
    },
    {
      "epoch": 5.3393755116360655,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 22826
    },
    {
      "epoch": 5.339609402409075,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 22827
    },
    {
      "epoch": 5.339843293182084,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 22828
    },
    {
      "epoch": 5.340077183955093,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 22829
    },
    {
      "epoch": 5.340311074728102,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 22830
    },
    {
      "epoch": 5.340544965501111,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 22831
    },
    {
      "epoch": 5.34077885627412,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 22832
    },
    {
      "epoch": 5.341012747047129,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5951,
      "step": 22833
    },
    {
      "epoch": 5.341246637820138,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9605,
      "step": 22834
    },
    {
      "epoch": 5.341480528593147,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8976,
      "step": 22835
    },
    {
      "epoch": 5.341714419366156,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5651,
      "step": 22836
    },
    {
      "epoch": 5.341948310139165,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 22837
    },
    {
      "epoch": 5.342182200912174,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8976,
      "step": 22838
    },
    {
      "epoch": 5.342416091685183,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5048,
      "step": 22839
    },
    {
      "epoch": 5.342649982458192,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7116,
      "step": 22840
    },
    {
      "epoch": 5.342883873231201,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5744,
      "step": 22841
    },
    {
      "epoch": 5.3431177640042105,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 22842
    },
    {
      "epoch": 5.343351654777219,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6877,
      "step": 22843
    },
    {
      "epoch": 5.343585545550228,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 22844
    },
    {
      "epoch": 5.343819436323237,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 22845
    },
    {
      "epoch": 5.344053327096246,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 22846
    },
    {
      "epoch": 5.344287217869255,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7666,
      "step": 22847
    },
    {
      "epoch": 5.344521108642264,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6771,
      "step": 22848
    },
    {
      "epoch": 5.344754999415273,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 22849
    },
    {
      "epoch": 5.344988890188282,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 22850
    },
    {
      "epoch": 5.345222780961291,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 22851
    },
    {
      "epoch": 5.3454566717343,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 22852
    },
    {
      "epoch": 5.3456905625073095,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 22853
    },
    {
      "epoch": 5.345924453280318,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5681,
      "step": 22854
    },
    {
      "epoch": 5.346158344053327,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 22855
    },
    {
      "epoch": 5.346392234826336,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 22856
    },
    {
      "epoch": 5.346626125599345,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.4969,
      "step": 22857
    },
    {
      "epoch": 5.346860016372354,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 22858
    },
    {
      "epoch": 5.347093907145363,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5745,
      "step": 22859
    },
    {
      "epoch": 5.347327797918372,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2814,
      "step": 22860
    },
    {
      "epoch": 5.347561688691381,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2865,
      "step": 22861
    },
    {
      "epoch": 5.34779557946439,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 22862
    },
    {
      "epoch": 5.348029470237399,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 22863
    },
    {
      "epoch": 5.3482633610104084,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 22864
    },
    {
      "epoch": 5.348497251783417,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 22865
    },
    {
      "epoch": 5.348731142556426,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 22866
    },
    {
      "epoch": 5.348965033329435,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 22867
    },
    {
      "epoch": 5.349198924102444,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 22868
    },
    {
      "epoch": 5.3494328148754535,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 22869
    },
    {
      "epoch": 5.349666705648462,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 22870
    },
    {
      "epoch": 5.349900596421471,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.793,
      "step": 22871
    },
    {
      "epoch": 5.35013448719448,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 22872
    },
    {
      "epoch": 5.350368377967489,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 22873
    },
    {
      "epoch": 5.350602268740499,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 22874
    },
    {
      "epoch": 5.350836159513507,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 22875
    },
    {
      "epoch": 5.351070050286516,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 22876
    },
    {
      "epoch": 5.351303941059525,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 22877
    },
    {
      "epoch": 5.351537831832534,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6548,
      "step": 22878
    },
    {
      "epoch": 5.351771722605543,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6877,
      "step": 22879
    },
    {
      "epoch": 5.3520056133785525,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 22880
    },
    {
      "epoch": 5.352239504151561,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7421,
      "step": 22881
    },
    {
      "epoch": 5.35247339492457,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 22882
    },
    {
      "epoch": 5.352707285697579,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6255,
      "step": 22883
    },
    {
      "epoch": 5.352941176470588,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 22884
    },
    {
      "epoch": 5.353175067243598,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4421,
      "step": 22885
    },
    {
      "epoch": 5.353408958016606,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 22886
    },
    {
      "epoch": 5.353642848789615,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 22887
    },
    {
      "epoch": 5.353876739562624,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6431,
      "step": 22888
    },
    {
      "epoch": 5.354110630335633,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 22889
    },
    {
      "epoch": 5.354344521108643,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 22890
    },
    {
      "epoch": 5.3545784118816515,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 22891
    },
    {
      "epoch": 5.35481230265466,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 22892
    },
    {
      "epoch": 5.355046193427669,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 22893
    },
    {
      "epoch": 5.355280084200678,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 22894
    },
    {
      "epoch": 5.355513974973688,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 22895
    },
    {
      "epoch": 5.355747865746697,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 22896
    },
    {
      "epoch": 5.355981756519705,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6038,
      "step": 22897
    },
    {
      "epoch": 5.356215647292714,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 22898
    },
    {
      "epoch": 5.356449538065723,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 22899
    },
    {
      "epoch": 5.356683428838732,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 22900
    },
    {
      "epoch": 5.356683428838732,
      "eval_runtime": 4.69,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 22900
    },
    {
      "epoch": 5.356917319611742,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 22901
    },
    {
      "epoch": 5.3571512103847505,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.5307,
      "step": 22902
    },
    {
      "epoch": 5.357385101157759,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6457,
      "step": 22903
    },
    {
      "epoch": 5.357618991930768,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7923,
      "step": 22904
    },
    {
      "epoch": 5.357852882703777,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 22905
    },
    {
      "epoch": 5.358086773476787,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 22906
    },
    {
      "epoch": 5.358320664249796,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 22907
    },
    {
      "epoch": 5.358554555022804,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 22908
    },
    {
      "epoch": 5.358788445795813,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 22909
    },
    {
      "epoch": 5.359022336568822,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 22910
    },
    {
      "epoch": 5.359256227341831,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 22911
    },
    {
      "epoch": 5.359490118114841,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5941,
      "step": 22912
    },
    {
      "epoch": 5.3597240088878495,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 22913
    },
    {
      "epoch": 5.359957899660858,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 22914
    },
    {
      "epoch": 5.360191790433867,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 22915
    },
    {
      "epoch": 5.360425681206876,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 22916
    },
    {
      "epoch": 5.360659571979886,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.616,
      "step": 22917
    },
    {
      "epoch": 5.360893462752895,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 22918
    },
    {
      "epoch": 5.361127353525903,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6541,
      "step": 22919
    },
    {
      "epoch": 5.361361244298912,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 22920
    },
    {
      "epoch": 5.361595135071921,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.851,
      "step": 22921
    },
    {
      "epoch": 5.361829025844931,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5706,
      "step": 22922
    },
    {
      "epoch": 5.36206291661794,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8424,
      "step": 22923
    },
    {
      "epoch": 5.3622968073909485,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 22924
    },
    {
      "epoch": 5.362530698163957,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5516,
      "step": 22925
    },
    {
      "epoch": 5.362764588936966,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7994,
      "step": 22926
    },
    {
      "epoch": 5.362998479709976,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6714,
      "step": 22927
    },
    {
      "epoch": 5.363232370482985,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 22928
    },
    {
      "epoch": 5.363466261255994,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 22929
    },
    {
      "epoch": 5.363700152029002,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 22930
    },
    {
      "epoch": 5.363934042802011,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 22931
    },
    {
      "epoch": 5.36416793357502,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7573,
      "step": 22932
    },
    {
      "epoch": 5.36440182434803,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 22933
    },
    {
      "epoch": 5.364635715121039,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6894,
      "step": 22934
    },
    {
      "epoch": 5.3648696058940475,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 22935
    },
    {
      "epoch": 5.365103496667056,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 22936
    },
    {
      "epoch": 5.365337387440065,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6294,
      "step": 22937
    },
    {
      "epoch": 5.365571278213075,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 22938
    },
    {
      "epoch": 5.365805168986084,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 22939
    },
    {
      "epoch": 5.366039059759093,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6334,
      "step": 22940
    },
    {
      "epoch": 5.366272950532101,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 22941
    },
    {
      "epoch": 5.36650684130511,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 22942
    },
    {
      "epoch": 5.366740732078119,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.5355,
      "step": 22943
    },
    {
      "epoch": 5.366974622851129,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 22944
    },
    {
      "epoch": 5.367208513624138,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 22945
    },
    {
      "epoch": 5.3674424043971465,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6437,
      "step": 22946
    },
    {
      "epoch": 5.367676295170155,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2461,
      "step": 22947
    },
    {
      "epoch": 5.367910185943164,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 22948
    },
    {
      "epoch": 5.368144076716174,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 22949
    },
    {
      "epoch": 5.368377967489183,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 22950
    },
    {
      "epoch": 5.368611858262192,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 22951
    },
    {
      "epoch": 5.3688457490352,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 22952
    },
    {
      "epoch": 5.369079639808209,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6928,
      "step": 22953
    },
    {
      "epoch": 5.369313530581219,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4113,
      "step": 22954
    },
    {
      "epoch": 5.369547421354228,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.3178,
      "step": 22955
    },
    {
      "epoch": 5.369781312127237,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 22956
    },
    {
      "epoch": 5.3700152029002455,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 22957
    },
    {
      "epoch": 5.370249093673254,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 22958
    },
    {
      "epoch": 5.370482984446264,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 22959
    },
    {
      "epoch": 5.370716875219273,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 22960
    },
    {
      "epoch": 5.370950765992282,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 22961
    },
    {
      "epoch": 5.3711846567652906,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 22962
    },
    {
      "epoch": 5.371418547538299,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.532,
      "step": 22963
    },
    {
      "epoch": 5.371652438311308,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6062,
      "step": 22964
    },
    {
      "epoch": 5.371886329084318,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 22965
    },
    {
      "epoch": 5.372120219857327,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7442,
      "step": 22966
    },
    {
      "epoch": 5.372354110630336,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 22967
    },
    {
      "epoch": 5.3725880014033445,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 22968
    },
    {
      "epoch": 5.372821892176353,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 22969
    },
    {
      "epoch": 5.373055782949363,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 22970
    },
    {
      "epoch": 5.373289673722372,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 22971
    },
    {
      "epoch": 5.373523564495381,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 22972
    },
    {
      "epoch": 5.3737574552683895,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 22973
    },
    {
      "epoch": 5.373991346041398,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 22974
    },
    {
      "epoch": 5.374225236814407,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4419,
      "step": 22975
    },
    {
      "epoch": 5.374459127587417,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 22976
    },
    {
      "epoch": 5.374693018360426,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7586,
      "step": 22977
    },
    {
      "epoch": 5.374926909133435,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 22978
    },
    {
      "epoch": 5.3751607999064435,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 22979
    },
    {
      "epoch": 5.375394690679452,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.4621,
      "step": 22980
    },
    {
      "epoch": 5.375628581452462,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 22981
    },
    {
      "epoch": 5.375862472225471,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 22982
    },
    {
      "epoch": 5.37609636299848,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 22983
    },
    {
      "epoch": 5.3763302537714885,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7655,
      "step": 22984
    },
    {
      "epoch": 5.376564144544497,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 22985
    },
    {
      "epoch": 5.376798035317507,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8284,
      "step": 22986
    },
    {
      "epoch": 5.377031926090516,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.4653,
      "step": 22987
    },
    {
      "epoch": 5.377265816863525,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5525,
      "step": 22988
    },
    {
      "epoch": 5.377499707636534,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 22989
    },
    {
      "epoch": 5.3777335984095425,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7188,
      "step": 22990
    },
    {
      "epoch": 5.377967489182552,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 22991
    },
    {
      "epoch": 5.378201379955561,
      "grad_norm": 2.484375,
      "learning_rate": 3e-05,
      "loss": 1.6389,
      "step": 22992
    },
    {
      "epoch": 5.37843527072857,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.675,
      "step": 22993
    },
    {
      "epoch": 5.378669161501579,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 22994
    },
    {
      "epoch": 5.3789030522745875,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 22995
    },
    {
      "epoch": 5.379136943047596,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 22996
    },
    {
      "epoch": 5.379370833820606,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 22997
    },
    {
      "epoch": 5.379604724593615,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7186,
      "step": 22998
    },
    {
      "epoch": 5.379838615366624,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7982,
      "step": 22999
    },
    {
      "epoch": 5.380072506139633,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.595,
      "step": 23000
    },
    {
      "epoch": 5.380072506139633,
      "eval_runtime": 4.6106,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 23000
    },
    {
      "epoch": 5.3803063969126415,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 23001
    },
    {
      "epoch": 5.380540287685651,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 23002
    },
    {
      "epoch": 5.38077417845866,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 23003
    },
    {
      "epoch": 5.381008069231669,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 23004
    },
    {
      "epoch": 5.381241960004678,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 23005
    },
    {
      "epoch": 5.3814758507776865,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 23006
    },
    {
      "epoch": 5.381709741550695,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 23007
    },
    {
      "epoch": 5.381943632323705,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6227,
      "step": 23008
    },
    {
      "epoch": 5.382177523096714,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 23009
    },
    {
      "epoch": 5.382411413869723,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 23010
    },
    {
      "epoch": 5.382645304642732,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 23011
    },
    {
      "epoch": 5.3828791954157404,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 23012
    },
    {
      "epoch": 5.38311308618875,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6956,
      "step": 23013
    },
    {
      "epoch": 5.383346976961759,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 23014
    },
    {
      "epoch": 5.383580867734768,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 23015
    },
    {
      "epoch": 5.383814758507777,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 23016
    },
    {
      "epoch": 5.3840486492807855,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7003,
      "step": 23017
    },
    {
      "epoch": 5.384282540053795,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5897,
      "step": 23018
    },
    {
      "epoch": 5.384516430826804,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 23019
    },
    {
      "epoch": 5.384750321599813,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 23020
    },
    {
      "epoch": 5.384984212372822,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 23021
    },
    {
      "epoch": 5.385218103145831,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6996,
      "step": 23022
    },
    {
      "epoch": 5.38545199391884,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 23023
    },
    {
      "epoch": 5.385685884691849,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 23024
    },
    {
      "epoch": 5.385919775464858,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 23025
    },
    {
      "epoch": 5.386153666237867,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6334,
      "step": 23026
    },
    {
      "epoch": 5.386387557010876,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.5924,
      "step": 23027
    },
    {
      "epoch": 5.3866214477838845,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7946,
      "step": 23028
    },
    {
      "epoch": 5.386855338556894,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7219,
      "step": 23029
    },
    {
      "epoch": 5.387089229329903,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6693,
      "step": 23030
    },
    {
      "epoch": 5.387323120102912,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.562,
      "step": 23031
    },
    {
      "epoch": 5.387557010875921,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6519,
      "step": 23032
    },
    {
      "epoch": 5.38779090164893,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 23033
    },
    {
      "epoch": 5.388024792421939,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 23034
    },
    {
      "epoch": 5.388258683194948,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 23035
    },
    {
      "epoch": 5.388492573967957,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 23036
    },
    {
      "epoch": 5.388726464740966,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 23037
    },
    {
      "epoch": 5.388960355513975,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6508,
      "step": 23038
    },
    {
      "epoch": 5.3891942462869835,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 23039
    },
    {
      "epoch": 5.389428137059993,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 23040
    },
    {
      "epoch": 5.389662027833002,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4909,
      "step": 23041
    },
    {
      "epoch": 5.389895918606011,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4617,
      "step": 23042
    },
    {
      "epoch": 5.39012980937902,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 23043
    },
    {
      "epoch": 5.390363700152029,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7232,
      "step": 23044
    },
    {
      "epoch": 5.390597590925038,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 23045
    },
    {
      "epoch": 5.390831481698047,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 23046
    },
    {
      "epoch": 5.391065372471056,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 23047
    },
    {
      "epoch": 5.391299263244065,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 23048
    },
    {
      "epoch": 5.391533154017074,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6668,
      "step": 23049
    },
    {
      "epoch": 5.391767044790083,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 23050
    },
    {
      "epoch": 5.392000935563092,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 23051
    },
    {
      "epoch": 5.392234826336101,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 23052
    },
    {
      "epoch": 5.39246871710911,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 23053
    },
    {
      "epoch": 5.392702607882119,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 23054
    },
    {
      "epoch": 5.3929364986551285,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 23055
    },
    {
      "epoch": 5.393170389428137,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 23056
    },
    {
      "epoch": 5.393404280201146,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 23057
    },
    {
      "epoch": 5.393638170974155,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5968,
      "step": 23058
    },
    {
      "epoch": 5.393872061747164,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 23059
    },
    {
      "epoch": 5.394105952520173,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 23060
    },
    {
      "epoch": 5.394339843293182,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7931,
      "step": 23061
    },
    {
      "epoch": 5.394573734066191,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 23062
    },
    {
      "epoch": 5.3948076248392,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 23063
    },
    {
      "epoch": 5.395041515612209,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2217,
      "step": 23064
    },
    {
      "epoch": 5.395275406385218,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7018,
      "step": 23065
    },
    {
      "epoch": 5.3955092971582275,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 23066
    },
    {
      "epoch": 5.395743187931236,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.662,
      "step": 23067
    },
    {
      "epoch": 5.395977078704245,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 23068
    },
    {
      "epoch": 5.396210969477254,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 23069
    },
    {
      "epoch": 5.396444860250263,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 23070
    },
    {
      "epoch": 5.3966787510232725,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 23071
    },
    {
      "epoch": 5.396912641796281,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 23072
    },
    {
      "epoch": 5.39714653256929,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 23073
    },
    {
      "epoch": 5.397380423342299,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6923,
      "step": 23074
    },
    {
      "epoch": 5.397614314115308,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.5752,
      "step": 23075
    },
    {
      "epoch": 5.397848204888318,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 23076
    },
    {
      "epoch": 5.3980820956613265,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3664,
      "step": 23077
    },
    {
      "epoch": 5.398315986434335,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 23078
    },
    {
      "epoch": 5.398549877207344,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7068,
      "step": 23079
    },
    {
      "epoch": 5.398783767980353,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 23080
    },
    {
      "epoch": 5.399017658753362,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 23081
    },
    {
      "epoch": 5.3992515495263715,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7724,
      "step": 23082
    },
    {
      "epoch": 5.39948544029938,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 23083
    },
    {
      "epoch": 5.399719331072389,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6599,
      "step": 23084
    },
    {
      "epoch": 5.399953221845398,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 23085
    },
    {
      "epoch": 5.400187112618407,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 23086
    },
    {
      "epoch": 5.400421003391417,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 23087
    },
    {
      "epoch": 5.4006548941644255,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 23088
    },
    {
      "epoch": 5.400888784937434,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 23089
    },
    {
      "epoch": 5.401122675710443,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 23090
    },
    {
      "epoch": 5.401356566483452,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 23091
    },
    {
      "epoch": 5.401590457256461,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 23092
    },
    {
      "epoch": 5.4018243480294705,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.648,
      "step": 23093
    },
    {
      "epoch": 5.402058238802479,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 23094
    },
    {
      "epoch": 5.402292129575488,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.6738,
      "step": 23095
    },
    {
      "epoch": 5.402526020348497,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 23096
    },
    {
      "epoch": 5.402759911121506,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 23097
    },
    {
      "epoch": 5.402993801894516,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 23098
    },
    {
      "epoch": 5.4032276926675245,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 23099
    },
    {
      "epoch": 5.403461583440533,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 23100
    },
    {
      "epoch": 5.403461583440533,
      "eval_runtime": 4.6246,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 23100
    },
    {
      "epoch": 5.403695474213542,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5955,
      "step": 23101
    },
    {
      "epoch": 5.403929364986551,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6479,
      "step": 23102
    },
    {
      "epoch": 5.404163255759561,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6325,
      "step": 23103
    },
    {
      "epoch": 5.4043971465325695,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5031,
      "step": 23104
    },
    {
      "epoch": 5.404631037305578,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 23105
    },
    {
      "epoch": 5.404864928078587,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 23106
    },
    {
      "epoch": 5.405098818851596,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 23107
    },
    {
      "epoch": 5.405332709624606,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 23108
    },
    {
      "epoch": 5.405566600397615,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 23109
    },
    {
      "epoch": 5.4058004911706234,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 23110
    },
    {
      "epoch": 5.406034381943632,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 23111
    },
    {
      "epoch": 5.406268272716641,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 23112
    },
    {
      "epoch": 5.40650216348965,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8369,
      "step": 23113
    },
    {
      "epoch": 5.40673605426266,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 23114
    },
    {
      "epoch": 5.4069699450356685,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 23115
    },
    {
      "epoch": 5.407203835808677,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.672,
      "step": 23116
    },
    {
      "epoch": 5.407437726581686,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 23117
    },
    {
      "epoch": 5.407671617354695,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 23118
    },
    {
      "epoch": 5.407905508127705,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 23119
    },
    {
      "epoch": 5.408139398900714,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 23120
    },
    {
      "epoch": 5.408373289673722,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3452,
      "step": 23121
    },
    {
      "epoch": 5.408607180446731,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2001,
      "step": 23122
    },
    {
      "epoch": 5.40884107121974,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 23123
    },
    {
      "epoch": 5.409074961992749,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 23124
    },
    {
      "epoch": 5.409308852765759,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 23125
    },
    {
      "epoch": 5.4095427435387675,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 23126
    },
    {
      "epoch": 5.409776634311776,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7127,
      "step": 23127
    },
    {
      "epoch": 5.410010525084785,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 23128
    },
    {
      "epoch": 5.410244415857794,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 23129
    },
    {
      "epoch": 5.410478306630804,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8303,
      "step": 23130
    },
    {
      "epoch": 5.410712197403813,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6892,
      "step": 23131
    },
    {
      "epoch": 5.410946088176821,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 23132
    },
    {
      "epoch": 5.41117997894983,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 23133
    },
    {
      "epoch": 5.411413869722839,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6144,
      "step": 23134
    },
    {
      "epoch": 5.411647760495849,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 23135
    },
    {
      "epoch": 5.411881651268858,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7472,
      "step": 23136
    },
    {
      "epoch": 5.4121155420418665,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.7255,
      "step": 23137
    },
    {
      "epoch": 5.412349432814875,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 23138
    },
    {
      "epoch": 5.412583323587884,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 23139
    },
    {
      "epoch": 5.412817214360894,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 23140
    },
    {
      "epoch": 5.413051105133903,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 23141
    },
    {
      "epoch": 5.413284995906912,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 23142
    },
    {
      "epoch": 5.41351888667992,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 23143
    },
    {
      "epoch": 5.413752777452929,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5907,
      "step": 23144
    },
    {
      "epoch": 5.413986668225938,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5029,
      "step": 23145
    },
    {
      "epoch": 5.414220558998948,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 23146
    },
    {
      "epoch": 5.414454449771957,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7108,
      "step": 23147
    },
    {
      "epoch": 5.4146883405449655,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 23148
    },
    {
      "epoch": 5.414922231317974,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 23149
    },
    {
      "epoch": 5.415156122090983,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 23150
    },
    {
      "epoch": 5.415390012863993,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6355,
      "step": 23151
    },
    {
      "epoch": 5.415623903637002,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 23152
    },
    {
      "epoch": 5.415857794410011,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 23153
    },
    {
      "epoch": 5.416091685183019,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 23154
    },
    {
      "epoch": 5.416325575956028,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 23155
    },
    {
      "epoch": 5.416559466729037,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 23156
    },
    {
      "epoch": 5.416793357502047,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 23157
    },
    {
      "epoch": 5.417027248275056,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7317,
      "step": 23158
    },
    {
      "epoch": 5.4172611390480645,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.6286,
      "step": 23159
    },
    {
      "epoch": 5.417495029821073,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 23160
    },
    {
      "epoch": 5.417728920594082,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5298,
      "step": 23161
    },
    {
      "epoch": 5.417962811367092,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7946,
      "step": 23162
    },
    {
      "epoch": 5.418196702140101,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 23163
    },
    {
      "epoch": 5.41843059291311,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6539,
      "step": 23164
    },
    {
      "epoch": 5.418664483686118,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 23165
    },
    {
      "epoch": 5.418898374459127,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 23166
    },
    {
      "epoch": 5.419132265232137,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 23167
    },
    {
      "epoch": 5.419366156005146,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.613,
      "step": 23168
    },
    {
      "epoch": 5.419600046778155,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 23169
    },
    {
      "epoch": 5.4198339375511635,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 23170
    },
    {
      "epoch": 5.420067828324172,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5245,
      "step": 23171
    },
    {
      "epoch": 5.420301719097182,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 23172
    },
    {
      "epoch": 5.420535609870191,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 23173
    },
    {
      "epoch": 5.4207695006432,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4634,
      "step": 23174
    },
    {
      "epoch": 5.421003391416209,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 23175
    },
    {
      "epoch": 5.421237282189217,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7794,
      "step": 23176
    },
    {
      "epoch": 5.421471172962226,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 23177
    },
    {
      "epoch": 5.421705063735236,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6226,
      "step": 23178
    },
    {
      "epoch": 5.421938954508245,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 23179
    },
    {
      "epoch": 5.422172845281254,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 23180
    },
    {
      "epoch": 5.4224067360542625,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 23181
    },
    {
      "epoch": 5.422640626827271,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 23182
    },
    {
      "epoch": 5.422874517600281,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 23183
    },
    {
      "epoch": 5.42310840837329,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6599,
      "step": 23184
    },
    {
      "epoch": 5.423342299146299,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 23185
    },
    {
      "epoch": 5.423576189919308,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 23186
    },
    {
      "epoch": 5.423810080692316,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 23187
    },
    {
      "epoch": 5.424043971465325,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 23188
    },
    {
      "epoch": 5.424277862238335,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 23189
    },
    {
      "epoch": 5.424511753011344,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4984,
      "step": 23190
    },
    {
      "epoch": 5.424745643784353,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 23191
    },
    {
      "epoch": 5.4249795345573615,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 23192
    },
    {
      "epoch": 5.42521342533037,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 23193
    },
    {
      "epoch": 5.42544731610338,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7448,
      "step": 23194
    },
    {
      "epoch": 5.425681206876389,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 23195
    },
    {
      "epoch": 5.425915097649398,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 23196
    },
    {
      "epoch": 5.426148988422407,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 23197
    },
    {
      "epoch": 5.426382879195415,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 23198
    },
    {
      "epoch": 5.426616769968425,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 23199
    },
    {
      "epoch": 5.426850660741434,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 23200
    },
    {
      "epoch": 5.426850660741434,
      "eval_runtime": 4.5786,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 23200
    },
    {
      "epoch": 5.427084551514443,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6725,
      "step": 23201
    },
    {
      "epoch": 5.427318442287452,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3089,
      "step": 23202
    },
    {
      "epoch": 5.4275523330604605,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 23203
    },
    {
      "epoch": 5.42778622383347,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2774,
      "step": 23204
    },
    {
      "epoch": 5.428020114606479,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.078,
      "step": 23205
    },
    {
      "epoch": 5.428254005379488,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7723,
      "step": 23206
    },
    {
      "epoch": 5.428487896152497,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 23207
    },
    {
      "epoch": 5.4287217869255056,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 23208
    },
    {
      "epoch": 5.428955677698514,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 23209
    },
    {
      "epoch": 5.429189568471524,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 23210
    },
    {
      "epoch": 5.429423459244533,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8684,
      "step": 23211
    },
    {
      "epoch": 5.429657350017542,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 23212
    },
    {
      "epoch": 5.429891240790551,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 23213
    },
    {
      "epoch": 5.4301251315635595,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 23214
    },
    {
      "epoch": 5.430359022336569,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 23215
    },
    {
      "epoch": 5.430592913109578,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8903,
      "step": 23216
    },
    {
      "epoch": 5.430826803882587,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 23217
    },
    {
      "epoch": 5.431060694655596,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7687,
      "step": 23218
    },
    {
      "epoch": 5.4312945854286045,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 23219
    },
    {
      "epoch": 5.431528476201613,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.55,
      "step": 23220
    },
    {
      "epoch": 5.431762366974623,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 23221
    },
    {
      "epoch": 5.431996257747632,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 23222
    },
    {
      "epoch": 5.432230148520641,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 23223
    },
    {
      "epoch": 5.43246403929365,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5895,
      "step": 23224
    },
    {
      "epoch": 5.4326979300666585,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 23225
    },
    {
      "epoch": 5.432931820839668,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 23226
    },
    {
      "epoch": 5.433165711612677,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7171,
      "step": 23227
    },
    {
      "epoch": 5.433399602385686,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8553,
      "step": 23228
    },
    {
      "epoch": 5.433633493158695,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 23229
    },
    {
      "epoch": 5.4338673839317035,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.3304,
      "step": 23230
    },
    {
      "epoch": 5.434101274704713,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 23231
    },
    {
      "epoch": 5.434335165477722,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 23232
    },
    {
      "epoch": 5.434569056250731,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 23233
    },
    {
      "epoch": 5.43480294702374,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5922,
      "step": 23234
    },
    {
      "epoch": 5.435036837796749,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 23235
    },
    {
      "epoch": 5.435270728569758,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 23236
    },
    {
      "epoch": 5.435504619342767,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 23237
    },
    {
      "epoch": 5.435738510115776,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 23238
    },
    {
      "epoch": 5.435972400888785,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 23239
    },
    {
      "epoch": 5.436206291661794,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 23240
    },
    {
      "epoch": 5.4364401824348025,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 23241
    },
    {
      "epoch": 5.436674073207812,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 23242
    },
    {
      "epoch": 5.436907963980821,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 23243
    },
    {
      "epoch": 5.43714185475383,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 23244
    },
    {
      "epoch": 5.437375745526839,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6644,
      "step": 23245
    },
    {
      "epoch": 5.437609636299848,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 23246
    },
    {
      "epoch": 5.437843527072857,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 23247
    },
    {
      "epoch": 5.438077417845866,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4137,
      "step": 23248
    },
    {
      "epoch": 5.438311308618875,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4411,
      "step": 23249
    },
    {
      "epoch": 5.438545199391884,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8101,
      "step": 23250
    },
    {
      "epoch": 5.438779090164893,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 23251
    },
    {
      "epoch": 5.4390129809379015,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6236,
      "step": 23252
    },
    {
      "epoch": 5.439246871710911,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 23253
    },
    {
      "epoch": 5.43948076248392,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 23254
    },
    {
      "epoch": 5.439714653256929,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 23255
    },
    {
      "epoch": 5.439948544029938,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 23256
    },
    {
      "epoch": 5.440182434802947,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 23257
    },
    {
      "epoch": 5.440416325575956,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.5486,
      "step": 23258
    },
    {
      "epoch": 5.440650216348965,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 23259
    },
    {
      "epoch": 5.440884107121974,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 23260
    },
    {
      "epoch": 5.441117997894983,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 23261
    },
    {
      "epoch": 5.441351888667992,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 23262
    },
    {
      "epoch": 5.441585779441001,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6222,
      "step": 23263
    },
    {
      "epoch": 5.44181967021401,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 23264
    },
    {
      "epoch": 5.442053560987019,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0599,
      "step": 23265
    },
    {
      "epoch": 5.442287451760028,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 23266
    },
    {
      "epoch": 5.442521342533037,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 23267
    },
    {
      "epoch": 5.4427552333060465,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 23268
    },
    {
      "epoch": 5.442989124079055,
      "grad_norm": 12.5625,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 23269
    },
    {
      "epoch": 5.443223014852064,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6223,
      "step": 23270
    },
    {
      "epoch": 5.443456905625073,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 23271
    },
    {
      "epoch": 5.443690796398082,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 23272
    },
    {
      "epoch": 5.443924687171091,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 23273
    },
    {
      "epoch": 5.4441585779441,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 23274
    },
    {
      "epoch": 5.444392468717109,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 23275
    },
    {
      "epoch": 5.444626359490118,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 23276
    },
    {
      "epoch": 5.444860250263127,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 23277
    },
    {
      "epoch": 5.445094141036136,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 23278
    },
    {
      "epoch": 5.4453280318091455,
      "grad_norm": 7.375,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 23279
    },
    {
      "epoch": 5.445561922582154,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 23280
    },
    {
      "epoch": 5.445795813355163,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 23281
    },
    {
      "epoch": 5.446029704128172,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 23282
    },
    {
      "epoch": 5.446263594901181,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 23283
    },
    {
      "epoch": 5.446497485674191,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 23284
    },
    {
      "epoch": 5.446731376447199,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.726,
      "step": 23285
    },
    {
      "epoch": 5.446965267220208,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6856,
      "step": 23286
    },
    {
      "epoch": 5.447199157993217,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 23287
    },
    {
      "epoch": 5.447433048766226,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 23288
    },
    {
      "epoch": 5.447666939539236,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 23289
    },
    {
      "epoch": 5.4479008303122445,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 23290
    },
    {
      "epoch": 5.448134721085253,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 23291
    },
    {
      "epoch": 5.448368611858262,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 23292
    },
    {
      "epoch": 5.448602502631271,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8629,
      "step": 23293
    },
    {
      "epoch": 5.44883639340428,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 23294
    },
    {
      "epoch": 5.44907028417729,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 23295
    },
    {
      "epoch": 5.449304174950298,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.4248,
      "step": 23296
    },
    {
      "epoch": 5.449538065723307,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 23297
    },
    {
      "epoch": 5.449771956496316,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 23298
    },
    {
      "epoch": 5.450005847269325,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 23299
    },
    {
      "epoch": 5.450239738042335,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 23300
    },
    {
      "epoch": 5.450239738042335,
      "eval_runtime": 4.61,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 23300
    },
    {
      "epoch": 5.4504736288153435,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 23301
    },
    {
      "epoch": 5.450707519588352,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 23302
    },
    {
      "epoch": 5.450941410361361,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 23303
    },
    {
      "epoch": 5.45117530113437,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6005,
      "step": 23304
    },
    {
      "epoch": 5.451409191907379,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 23305
    },
    {
      "epoch": 5.4516430826803886,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 23306
    },
    {
      "epoch": 5.451876973453397,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 23307
    },
    {
      "epoch": 5.452110864226406,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 23308
    },
    {
      "epoch": 5.452344754999415,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 23309
    },
    {
      "epoch": 5.452578645772424,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 23310
    },
    {
      "epoch": 5.452812536545434,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 23311
    },
    {
      "epoch": 5.4530464273184425,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5902,
      "step": 23312
    },
    {
      "epoch": 5.453280318091451,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 23313
    },
    {
      "epoch": 5.45351420886446,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7714,
      "step": 23314
    },
    {
      "epoch": 5.453748099637469,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 23315
    },
    {
      "epoch": 5.453981990410479,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 23316
    },
    {
      "epoch": 5.4542158811834875,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 23317
    },
    {
      "epoch": 5.454449771956496,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 23318
    },
    {
      "epoch": 5.454683662729505,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 23319
    },
    {
      "epoch": 5.454917553502514,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 23320
    },
    {
      "epoch": 5.455151444275524,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.743,
      "step": 23321
    },
    {
      "epoch": 5.455385335048533,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4855,
      "step": 23322
    },
    {
      "epoch": 5.4556192258215415,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 23323
    },
    {
      "epoch": 5.45585311659455,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6877,
      "step": 23324
    },
    {
      "epoch": 5.456087007367559,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.4421,
      "step": 23325
    },
    {
      "epoch": 5.456320898140568,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 23326
    },
    {
      "epoch": 5.456554788913578,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 23327
    },
    {
      "epoch": 5.4567886796865865,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 23328
    },
    {
      "epoch": 5.457022570459595,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 23329
    },
    {
      "epoch": 5.457256461232604,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 23330
    },
    {
      "epoch": 5.457490352005613,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 23331
    },
    {
      "epoch": 5.457724242778623,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 23332
    },
    {
      "epoch": 5.457958133551632,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 23333
    },
    {
      "epoch": 5.4581920243246405,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 23334
    },
    {
      "epoch": 5.458425915097649,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 23335
    },
    {
      "epoch": 5.458659805870658,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 23336
    },
    {
      "epoch": 5.458893696643667,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 23337
    },
    {
      "epoch": 5.459127587416677,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 23338
    },
    {
      "epoch": 5.4593614781896855,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 23339
    },
    {
      "epoch": 5.459595368962694,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 23340
    },
    {
      "epoch": 5.459829259735703,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.676,
      "step": 23341
    },
    {
      "epoch": 5.460063150508712,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6696,
      "step": 23342
    },
    {
      "epoch": 5.460297041281722,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 23343
    },
    {
      "epoch": 5.460530932054731,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 23344
    },
    {
      "epoch": 5.4607648228277395,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 23345
    },
    {
      "epoch": 5.460998713600748,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 23346
    },
    {
      "epoch": 5.461232604373757,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 23347
    },
    {
      "epoch": 5.461466495146767,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 23348
    },
    {
      "epoch": 5.461700385919776,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 23349
    },
    {
      "epoch": 5.4619342766927845,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7478,
      "step": 23350
    },
    {
      "epoch": 5.462168167465793,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7356,
      "step": 23351
    },
    {
      "epoch": 5.462402058238802,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4939,
      "step": 23352
    },
    {
      "epoch": 5.462635949011812,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9733,
      "step": 23353
    },
    {
      "epoch": 5.462869839784821,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 23354
    },
    {
      "epoch": 5.46310373055783,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 23355
    },
    {
      "epoch": 5.4633376213308384,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 23356
    },
    {
      "epoch": 5.463571512103847,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5929,
      "step": 23357
    },
    {
      "epoch": 5.463805402876856,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 23358
    },
    {
      "epoch": 5.464039293649866,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7093,
      "step": 23359
    },
    {
      "epoch": 5.464273184422875,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 23360
    },
    {
      "epoch": 5.4645070751958835,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 23361
    },
    {
      "epoch": 5.464740965968892,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 23362
    },
    {
      "epoch": 5.464974856741901,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 23363
    },
    {
      "epoch": 5.465208747514911,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 23364
    },
    {
      "epoch": 5.46544263828792,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 23365
    },
    {
      "epoch": 5.465676529060929,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 23366
    },
    {
      "epoch": 5.465910419833937,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 23367
    },
    {
      "epoch": 5.466144310606946,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 23368
    },
    {
      "epoch": 5.466378201379955,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 23369
    },
    {
      "epoch": 5.466612092152965,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 23370
    },
    {
      "epoch": 5.466845982925974,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5642,
      "step": 23371
    },
    {
      "epoch": 5.4670798736989825,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7667,
      "step": 23372
    },
    {
      "epoch": 5.467313764471991,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 23373
    },
    {
      "epoch": 5.467547655245,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6061,
      "step": 23374
    },
    {
      "epoch": 5.46778154601801,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 23375
    },
    {
      "epoch": 5.468015436791019,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 23376
    },
    {
      "epoch": 5.468249327564028,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 23377
    },
    {
      "epoch": 5.468483218337036,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 23378
    },
    {
      "epoch": 5.468717109110045,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 23379
    },
    {
      "epoch": 5.468950999883055,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8251,
      "step": 23380
    },
    {
      "epoch": 5.469184890656064,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 23381
    },
    {
      "epoch": 5.469418781429073,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 23382
    },
    {
      "epoch": 5.4696526722020815,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 23383
    },
    {
      "epoch": 5.46988656297509,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 23384
    },
    {
      "epoch": 5.4701204537481,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 23385
    },
    {
      "epoch": 5.470354344521109,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7733,
      "step": 23386
    },
    {
      "epoch": 5.470588235294118,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7123,
      "step": 23387
    },
    {
      "epoch": 5.470822126067127,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 23388
    },
    {
      "epoch": 5.471056016840135,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 23389
    },
    {
      "epoch": 5.471289907613144,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 23390
    },
    {
      "epoch": 5.471523798386154,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 23391
    },
    {
      "epoch": 5.471757689159163,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 23392
    },
    {
      "epoch": 5.471991579932172,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 23393
    },
    {
      "epoch": 5.4722254707051805,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7669,
      "step": 23394
    },
    {
      "epoch": 5.472459361478189,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 23395
    },
    {
      "epoch": 5.472693252251199,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6492,
      "step": 23396
    },
    {
      "epoch": 5.472927143024208,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 23397
    },
    {
      "epoch": 5.473161033797217,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 23398
    },
    {
      "epoch": 5.473394924570226,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5844,
      "step": 23399
    },
    {
      "epoch": 5.473628815343234,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5624,
      "step": 23400
    },
    {
      "epoch": 5.473628815343234,
      "eval_runtime": 4.5992,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 23400
    },
    {
      "epoch": 5.473862706116243,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 23401
    },
    {
      "epoch": 5.474096596889253,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 23402
    },
    {
      "epoch": 5.474330487662262,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 23403
    },
    {
      "epoch": 5.474564378435271,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 23404
    },
    {
      "epoch": 5.4747982692082795,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 23405
    },
    {
      "epoch": 5.475032159981288,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 23406
    },
    {
      "epoch": 5.475266050754298,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6423,
      "step": 23407
    },
    {
      "epoch": 5.475499941527307,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 23408
    },
    {
      "epoch": 5.475733832300316,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6015,
      "step": 23409
    },
    {
      "epoch": 5.475967723073325,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5525,
      "step": 23410
    },
    {
      "epoch": 5.476201613846333,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6309,
      "step": 23411
    },
    {
      "epoch": 5.476435504619343,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 23412
    },
    {
      "epoch": 5.476669395392352,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 23413
    },
    {
      "epoch": 5.476903286165361,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6257,
      "step": 23414
    },
    {
      "epoch": 5.47713717693837,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 23415
    },
    {
      "epoch": 5.4773710677113785,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 23416
    },
    {
      "epoch": 5.477604958484388,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 23417
    },
    {
      "epoch": 5.477838849257397,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 23418
    },
    {
      "epoch": 5.478072740030406,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 23419
    },
    {
      "epoch": 5.478306630803415,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 23420
    },
    {
      "epoch": 5.478540521576424,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 23421
    },
    {
      "epoch": 5.478774412349432,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 23422
    },
    {
      "epoch": 5.479008303122442,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6841,
      "step": 23423
    },
    {
      "epoch": 5.479242193895451,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5171,
      "step": 23424
    },
    {
      "epoch": 5.47947608466846,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7022,
      "step": 23425
    },
    {
      "epoch": 5.479709975441469,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6381,
      "step": 23426
    },
    {
      "epoch": 5.4799438662144775,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 23427
    },
    {
      "epoch": 5.480177756987487,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 23428
    },
    {
      "epoch": 5.480411647760496,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 23429
    },
    {
      "epoch": 5.480645538533505,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 23430
    },
    {
      "epoch": 5.480879429306514,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1703,
      "step": 23431
    },
    {
      "epoch": 5.481113320079523,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.662,
      "step": 23432
    },
    {
      "epoch": 5.481347210852531,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 23433
    },
    {
      "epoch": 5.481581101625541,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 23434
    },
    {
      "epoch": 5.48181499239855,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 23435
    },
    {
      "epoch": 5.482048883171559,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 23436
    },
    {
      "epoch": 5.482282773944568,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 23437
    },
    {
      "epoch": 5.4825166647175765,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4471,
      "step": 23438
    },
    {
      "epoch": 5.482750555490586,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 23439
    },
    {
      "epoch": 5.482984446263595,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 23440
    },
    {
      "epoch": 5.483218337036604,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.813,
      "step": 23441
    },
    {
      "epoch": 5.483452227809613,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 2.4544,
      "step": 23442
    },
    {
      "epoch": 5.483686118582622,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6156,
      "step": 23443
    },
    {
      "epoch": 5.483920009355631,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 23444
    },
    {
      "epoch": 5.48415390012864,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 23445
    },
    {
      "epoch": 5.484387790901649,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 23446
    },
    {
      "epoch": 5.484621681674658,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 23447
    },
    {
      "epoch": 5.484855572447667,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 23448
    },
    {
      "epoch": 5.485089463220676,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6727,
      "step": 23449
    },
    {
      "epoch": 5.485323353993685,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 23450
    },
    {
      "epoch": 5.485557244766694,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 23451
    },
    {
      "epoch": 5.485791135539703,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 23452
    },
    {
      "epoch": 5.486025026312712,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 23453
    },
    {
      "epoch": 5.4862589170857206,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 23454
    },
    {
      "epoch": 5.48649280785873,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 23455
    },
    {
      "epoch": 5.486726698631739,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 23456
    },
    {
      "epoch": 5.486960589404748,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 23457
    },
    {
      "epoch": 5.487194480177757,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.7411,
      "step": 23458
    },
    {
      "epoch": 5.487428370950766,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 23459
    },
    {
      "epoch": 5.487662261723775,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 23460
    },
    {
      "epoch": 5.487896152496784,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 23461
    },
    {
      "epoch": 5.488130043269793,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 23462
    },
    {
      "epoch": 5.488363934042802,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 23463
    },
    {
      "epoch": 5.488597824815811,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 23464
    },
    {
      "epoch": 5.48883171558882,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 23465
    },
    {
      "epoch": 5.489065606361829,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 23466
    },
    {
      "epoch": 5.489299497134838,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 23467
    },
    {
      "epoch": 5.489533387907847,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 23468
    },
    {
      "epoch": 5.489767278680856,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 23469
    },
    {
      "epoch": 5.4900011694538655,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 23470
    },
    {
      "epoch": 5.490235060226874,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 23471
    },
    {
      "epoch": 5.490468950999883,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 1.8252,
      "step": 23472
    },
    {
      "epoch": 5.490702841772892,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5164,
      "step": 23473
    },
    {
      "epoch": 5.490936732545901,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 23474
    },
    {
      "epoch": 5.49117062331891,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 23475
    },
    {
      "epoch": 5.491404514091919,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 23476
    },
    {
      "epoch": 5.491638404864928,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5759,
      "step": 23477
    },
    {
      "epoch": 5.491872295637937,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 23478
    },
    {
      "epoch": 5.492106186410946,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 23479
    },
    {
      "epoch": 5.492340077183955,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 23480
    },
    {
      "epoch": 5.4925739679569645,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 23481
    },
    {
      "epoch": 5.492807858729973,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 23482
    },
    {
      "epoch": 5.493041749502982,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 23483
    },
    {
      "epoch": 5.493275640275991,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6157,
      "step": 23484
    },
    {
      "epoch": 5.493509531049,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6603,
      "step": 23485
    },
    {
      "epoch": 5.493743421822009,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 23486
    },
    {
      "epoch": 5.493977312595018,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5735,
      "step": 23487
    },
    {
      "epoch": 5.494211203368027,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 23488
    },
    {
      "epoch": 5.494445094141036,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 23489
    },
    {
      "epoch": 5.494678984914045,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7656,
      "step": 23490
    },
    {
      "epoch": 5.494912875687054,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4238,
      "step": 23491
    },
    {
      "epoch": 5.4951467664600635,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1556,
      "step": 23492
    },
    {
      "epoch": 5.495380657233072,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 23493
    },
    {
      "epoch": 5.495614548006081,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 23494
    },
    {
      "epoch": 5.49584843877909,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 23495
    },
    {
      "epoch": 5.496082329552099,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 23496
    },
    {
      "epoch": 5.496316220325109,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7202,
      "step": 23497
    },
    {
      "epoch": 5.496550111098117,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4261,
      "step": 23498
    },
    {
      "epoch": 5.496784001871126,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6809,
      "step": 23499
    },
    {
      "epoch": 5.497017892644135,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.685,
      "step": 23500
    },
    {
      "epoch": 5.497017892644135,
      "eval_runtime": 4.6729,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 23500
    },
    {
      "epoch": 5.497251783417144,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 23501
    },
    {
      "epoch": 5.497485674190154,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 23502
    },
    {
      "epoch": 5.4977195649631625,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 23503
    },
    {
      "epoch": 5.497953455736171,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 23504
    },
    {
      "epoch": 5.49818734650918,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6907,
      "step": 23505
    },
    {
      "epoch": 5.498421237282189,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 23506
    },
    {
      "epoch": 5.498655128055198,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4148,
      "step": 23507
    },
    {
      "epoch": 5.498889018828208,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 23508
    },
    {
      "epoch": 5.499122909601216,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 23509
    },
    {
      "epoch": 5.499356800374225,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 23510
    },
    {
      "epoch": 5.499590691147234,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 23511
    },
    {
      "epoch": 5.499824581920243,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 23512
    },
    {
      "epoch": 5.500058472693253,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 23513
    },
    {
      "epoch": 5.5002923634662615,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 23514
    },
    {
      "epoch": 5.50052625423927,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 23515
    },
    {
      "epoch": 5.500760145012279,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 23516
    },
    {
      "epoch": 5.500994035785288,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 23517
    },
    {
      "epoch": 5.501227926558297,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 23518
    },
    {
      "epoch": 5.501461817331307,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.656,
      "step": 23519
    },
    {
      "epoch": 5.501695708104315,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 23520
    },
    {
      "epoch": 5.501929598877324,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9087,
      "step": 23521
    },
    {
      "epoch": 5.502163489650333,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 23522
    },
    {
      "epoch": 5.502397380423342,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 23523
    },
    {
      "epoch": 5.502631271196352,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 23524
    },
    {
      "epoch": 5.5028651619693605,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 23525
    },
    {
      "epoch": 5.503099052742369,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8073,
      "step": 23526
    },
    {
      "epoch": 5.503332943515378,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 23527
    },
    {
      "epoch": 5.503566834288387,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 23528
    },
    {
      "epoch": 5.503800725061396,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5779,
      "step": 23529
    },
    {
      "epoch": 5.504034615834406,
      "grad_norm": 8.5,
      "learning_rate": 3e-05,
      "loss": 2.5847,
      "step": 23530
    },
    {
      "epoch": 5.504268506607414,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.675,
      "step": 23531
    },
    {
      "epoch": 5.504502397380423,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4334,
      "step": 23532
    },
    {
      "epoch": 5.504736288153432,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 23533
    },
    {
      "epoch": 5.504970178926442,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7454,
      "step": 23534
    },
    {
      "epoch": 5.505204069699451,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1277,
      "step": 23535
    },
    {
      "epoch": 5.5054379604724595,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 23536
    },
    {
      "epoch": 5.505671851245468,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 23537
    },
    {
      "epoch": 5.505905742018477,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 23538
    },
    {
      "epoch": 5.506139632791486,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 23539
    },
    {
      "epoch": 5.506373523564496,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 23540
    },
    {
      "epoch": 5.506607414337505,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9221,
      "step": 23541
    },
    {
      "epoch": 5.506841305110513,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 23542
    },
    {
      "epoch": 5.507075195883522,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6926,
      "step": 23543
    },
    {
      "epoch": 5.507309086656531,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 23544
    },
    {
      "epoch": 5.507542977429541,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6426,
      "step": 23545
    },
    {
      "epoch": 5.50777686820255,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 23546
    },
    {
      "epoch": 5.5080107589755585,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 23547
    },
    {
      "epoch": 5.508244649748567,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 23548
    },
    {
      "epoch": 5.508478540521576,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7019,
      "step": 23549
    },
    {
      "epoch": 5.508712431294585,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 23550
    },
    {
      "epoch": 5.508946322067595,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 23551
    },
    {
      "epoch": 5.5091802128406036,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 23552
    },
    {
      "epoch": 5.509414103613612,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6697,
      "step": 23553
    },
    {
      "epoch": 5.509647994386621,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5997,
      "step": 23554
    },
    {
      "epoch": 5.50988188515963,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 23555
    },
    {
      "epoch": 5.51011577593264,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 23556
    },
    {
      "epoch": 5.510349666705649,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6358,
      "step": 23557
    },
    {
      "epoch": 5.5105835574786575,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6437,
      "step": 23558
    },
    {
      "epoch": 5.510817448251666,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 23559
    },
    {
      "epoch": 5.511051339024675,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 23560
    },
    {
      "epoch": 5.511285229797684,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 23561
    },
    {
      "epoch": 5.511519120570694,
      "grad_norm": 7.625,
      "learning_rate": 3e-05,
      "loss": 1.7154,
      "step": 23562
    },
    {
      "epoch": 5.5117530113437025,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7068,
      "step": 23563
    },
    {
      "epoch": 5.511986902116711,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7536,
      "step": 23564
    },
    {
      "epoch": 5.51222079288972,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6909,
      "step": 23565
    },
    {
      "epoch": 5.51245468366273,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7727,
      "step": 23566
    },
    {
      "epoch": 5.512688574435739,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 23567
    },
    {
      "epoch": 5.512922465208748,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 23568
    },
    {
      "epoch": 5.5131563559817565,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 23569
    },
    {
      "epoch": 5.513390246754765,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4633,
      "step": 23570
    },
    {
      "epoch": 5.513624137527774,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5029,
      "step": 23571
    },
    {
      "epoch": 5.513858028300784,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.4846,
      "step": 23572
    },
    {
      "epoch": 5.514091919073793,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 23573
    },
    {
      "epoch": 5.5143258098468015,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 23574
    },
    {
      "epoch": 5.51455970061981,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 23575
    },
    {
      "epoch": 5.514793591392819,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 23576
    },
    {
      "epoch": 5.515027482165829,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 23577
    },
    {
      "epoch": 5.515261372938838,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 23578
    },
    {
      "epoch": 5.515495263711847,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 23579
    },
    {
      "epoch": 5.5157291544848555,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7614,
      "step": 23580
    },
    {
      "epoch": 5.515963045257864,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 23581
    },
    {
      "epoch": 5.516196936030873,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 23582
    },
    {
      "epoch": 5.516430826803883,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 23583
    },
    {
      "epoch": 5.516664717576892,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 23584
    },
    {
      "epoch": 5.5168986083499005,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 23585
    },
    {
      "epoch": 5.517132499122909,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6223,
      "step": 23586
    },
    {
      "epoch": 5.517366389895919,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6735,
      "step": 23587
    },
    {
      "epoch": 5.517600280668928,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 23588
    },
    {
      "epoch": 5.517834171441937,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 23589
    },
    {
      "epoch": 5.518068062214946,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 23590
    },
    {
      "epoch": 5.5183019529879545,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 23591
    },
    {
      "epoch": 5.518535843760963,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 23592
    },
    {
      "epoch": 5.518769734533973,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 23593
    },
    {
      "epoch": 5.519003625306982,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 23594
    },
    {
      "epoch": 5.519237516079991,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 23595
    },
    {
      "epoch": 5.5194714068529995,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 23596
    },
    {
      "epoch": 5.519705297626008,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 23597
    },
    {
      "epoch": 5.519939188399018,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 23598
    },
    {
      "epoch": 5.520173079172027,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7128,
      "step": 23599
    },
    {
      "epoch": 5.520406969945036,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 23600
    },
    {
      "epoch": 5.520406969945036,
      "eval_runtime": 4.6209,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 23600
    },
    {
      "epoch": 5.520640860718045,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 23601
    },
    {
      "epoch": 5.5208747514910534,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 23602
    },
    {
      "epoch": 5.521108642264062,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5602,
      "step": 23603
    },
    {
      "epoch": 5.521342533037072,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8688,
      "step": 23604
    },
    {
      "epoch": 5.521576423810081,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 23605
    },
    {
      "epoch": 5.52181031458309,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 23606
    },
    {
      "epoch": 5.5220442053560985,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 23607
    },
    {
      "epoch": 5.522278096129107,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 23608
    },
    {
      "epoch": 5.522511986902117,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 23609
    },
    {
      "epoch": 5.522745877675126,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 23610
    },
    {
      "epoch": 5.522979768448135,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 23611
    },
    {
      "epoch": 5.523213659221144,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 23612
    },
    {
      "epoch": 5.523447549994152,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6576,
      "step": 23613
    },
    {
      "epoch": 5.523681440767161,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 23614
    },
    {
      "epoch": 5.523915331540171,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8653,
      "step": 23615
    },
    {
      "epoch": 5.52414922231318,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 23616
    },
    {
      "epoch": 5.524383113086189,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 23617
    },
    {
      "epoch": 5.5246170038591975,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 23618
    },
    {
      "epoch": 5.524850894632207,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 23619
    },
    {
      "epoch": 5.525084785405216,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 23620
    },
    {
      "epoch": 5.525318676178225,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7413,
      "step": 23621
    },
    {
      "epoch": 5.525552566951234,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 23622
    },
    {
      "epoch": 5.525786457724243,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 23623
    },
    {
      "epoch": 5.526020348497251,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 23624
    },
    {
      "epoch": 5.526254239270261,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8047,
      "step": 23625
    },
    {
      "epoch": 5.52648813004327,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 23626
    },
    {
      "epoch": 5.526722020816279,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 23627
    },
    {
      "epoch": 5.526955911589288,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7339,
      "step": 23628
    },
    {
      "epoch": 5.5271898023622965,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 23629
    },
    {
      "epoch": 5.527423693135306,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 23630
    },
    {
      "epoch": 5.527657583908315,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 23631
    },
    {
      "epoch": 5.527891474681324,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6902,
      "step": 23632
    },
    {
      "epoch": 5.528125365454333,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 23633
    },
    {
      "epoch": 5.528359256227342,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6734,
      "step": 23634
    },
    {
      "epoch": 5.52859314700035,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 23635
    },
    {
      "epoch": 5.52882703777336,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 23636
    },
    {
      "epoch": 5.529060928546369,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 23637
    },
    {
      "epoch": 5.529294819319378,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 23638
    },
    {
      "epoch": 5.529528710092387,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 23639
    },
    {
      "epoch": 5.5297626008653955,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 23640
    },
    {
      "epoch": 5.529996491638405,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 23641
    },
    {
      "epoch": 5.530230382411414,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 23642
    },
    {
      "epoch": 5.530464273184423,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 23643
    },
    {
      "epoch": 5.530698163957432,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 23644
    },
    {
      "epoch": 5.530932054730441,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 23645
    },
    {
      "epoch": 5.531165945503449,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 23646
    },
    {
      "epoch": 5.531399836276459,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 23647
    },
    {
      "epoch": 5.531633727049468,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 23648
    },
    {
      "epoch": 5.531867617822477,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 23649
    },
    {
      "epoch": 5.532101508595486,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 23650
    },
    {
      "epoch": 5.532335399368495,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 23651
    },
    {
      "epoch": 5.532569290141504,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6587,
      "step": 23652
    },
    {
      "epoch": 5.532803180914513,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 23653
    },
    {
      "epoch": 5.533037071687522,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 23654
    },
    {
      "epoch": 5.533270962460531,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 23655
    },
    {
      "epoch": 5.53350485323354,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 23656
    },
    {
      "epoch": 5.533738744006549,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 23657
    },
    {
      "epoch": 5.533972634779558,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 23658
    },
    {
      "epoch": 5.534206525552567,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 23659
    },
    {
      "epoch": 5.534440416325576,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 23660
    },
    {
      "epoch": 5.534674307098585,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 23661
    },
    {
      "epoch": 5.534908197871594,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 23662
    },
    {
      "epoch": 5.535142088644603,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 23663
    },
    {
      "epoch": 5.535375979417612,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6884,
      "step": 23664
    },
    {
      "epoch": 5.535609870190621,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 23665
    },
    {
      "epoch": 5.53584376096363,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7085,
      "step": 23666
    },
    {
      "epoch": 5.536077651736639,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 23667
    },
    {
      "epoch": 5.536311542509648,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 23668
    },
    {
      "epoch": 5.536545433282657,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 23669
    },
    {
      "epoch": 5.536779324055666,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 23670
    },
    {
      "epoch": 5.537013214828675,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9388,
      "step": 23671
    },
    {
      "epoch": 5.537247105601684,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 23672
    },
    {
      "epoch": 5.537480996374693,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 23673
    },
    {
      "epoch": 5.537714887147702,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 23674
    },
    {
      "epoch": 5.537948777920711,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7818,
      "step": 23675
    },
    {
      "epoch": 5.53818266869372,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 23676
    },
    {
      "epoch": 5.538416559466729,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 23677
    },
    {
      "epoch": 5.538650450239738,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 23678
    },
    {
      "epoch": 5.538884341012747,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 23679
    },
    {
      "epoch": 5.539118231785756,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7724,
      "step": 23680
    },
    {
      "epoch": 5.539352122558765,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 23681
    },
    {
      "epoch": 5.539586013331774,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 23682
    },
    {
      "epoch": 5.5398199041047835,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 23683
    },
    {
      "epoch": 5.540053794877792,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 23684
    },
    {
      "epoch": 5.540287685650801,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 23685
    },
    {
      "epoch": 5.54052157642381,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 23686
    },
    {
      "epoch": 5.540755467196819,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6178,
      "step": 23687
    },
    {
      "epoch": 5.540989357969828,
      "grad_norm": 2.515625,
      "learning_rate": 3e-05,
      "loss": 1.6616,
      "step": 23688
    },
    {
      "epoch": 5.5412232487428374,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 23689
    },
    {
      "epoch": 5.541457139515846,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.3226,
      "step": 23690
    },
    {
      "epoch": 5.541691030288855,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 23691
    },
    {
      "epoch": 5.541924921061864,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 23692
    },
    {
      "epoch": 5.542158811834873,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.6413,
      "step": 23693
    },
    {
      "epoch": 5.5423927026078825,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 23694
    },
    {
      "epoch": 5.542626593380891,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3439,
      "step": 23695
    },
    {
      "epoch": 5.5428604841539,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 23696
    },
    {
      "epoch": 5.543094374926909,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 23697
    },
    {
      "epoch": 5.543328265699918,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 23698
    },
    {
      "epoch": 5.543562156472927,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6971,
      "step": 23699
    },
    {
      "epoch": 5.543796047245936,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5278,
      "step": 23700
    },
    {
      "epoch": 5.543796047245936,
      "eval_runtime": 4.6002,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 23700
    },
    {
      "epoch": 5.544029938018945,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 23701
    },
    {
      "epoch": 5.544263828791954,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 23702
    },
    {
      "epoch": 5.544497719564963,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 23703
    },
    {
      "epoch": 5.544731610337972,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5302,
      "step": 23704
    },
    {
      "epoch": 5.5449655011109815,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 23705
    },
    {
      "epoch": 5.54519939188399,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.636,
      "step": 23706
    },
    {
      "epoch": 5.545433282656999,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 23707
    },
    {
      "epoch": 5.545667173430008,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 23708
    },
    {
      "epoch": 5.545901064203017,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2004,
      "step": 23709
    },
    {
      "epoch": 5.546134954976026,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6936,
      "step": 23710
    },
    {
      "epoch": 5.546368845749035,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 23711
    },
    {
      "epoch": 5.546602736522044,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7148,
      "step": 23712
    },
    {
      "epoch": 5.546836627295053,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6692,
      "step": 23713
    },
    {
      "epoch": 5.547070518068062,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6938,
      "step": 23714
    },
    {
      "epoch": 5.547304408841072,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 23715
    },
    {
      "epoch": 5.5475382996140805,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6397,
      "step": 23716
    },
    {
      "epoch": 5.547772190387089,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 23717
    },
    {
      "epoch": 5.548006081160098,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7001,
      "step": 23718
    },
    {
      "epoch": 5.548239971933107,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 23719
    },
    {
      "epoch": 5.548473862706116,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7005,
      "step": 23720
    },
    {
      "epoch": 5.548707753479126,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5922,
      "step": 23721
    },
    {
      "epoch": 5.548941644252134,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5568,
      "step": 23722
    },
    {
      "epoch": 5.549175535025143,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 23723
    },
    {
      "epoch": 5.549409425798152,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 23724
    },
    {
      "epoch": 5.549643316571161,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 23725
    },
    {
      "epoch": 5.549877207344171,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8466,
      "step": 23726
    },
    {
      "epoch": 5.5501110981171795,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 23727
    },
    {
      "epoch": 5.550344988890188,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 23728
    },
    {
      "epoch": 5.550578879663197,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 23729
    },
    {
      "epoch": 5.550812770436206,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 23730
    },
    {
      "epoch": 5.551046661209215,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5997,
      "step": 23731
    },
    {
      "epoch": 5.551280551982225,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7655,
      "step": 23732
    },
    {
      "epoch": 5.551514442755233,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 23733
    },
    {
      "epoch": 5.551748333528242,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 23734
    },
    {
      "epoch": 5.551982224301251,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 23735
    },
    {
      "epoch": 5.55221611507426,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 23736
    },
    {
      "epoch": 5.55245000584727,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 23737
    },
    {
      "epoch": 5.5526838966202785,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 23738
    },
    {
      "epoch": 5.552917787393287,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 23739
    },
    {
      "epoch": 5.553151678166296,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.63,
      "step": 23740
    },
    {
      "epoch": 5.553385568939305,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 23741
    },
    {
      "epoch": 5.553619459712314,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 23742
    },
    {
      "epoch": 5.553853350485324,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 23743
    },
    {
      "epoch": 5.554087241258332,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 23744
    },
    {
      "epoch": 5.554321132031341,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5557,
      "step": 23745
    },
    {
      "epoch": 5.55455502280435,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 23746
    },
    {
      "epoch": 5.55478891357736,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 23747
    },
    {
      "epoch": 5.555022804350369,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 23748
    },
    {
      "epoch": 5.5552566951233775,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 23749
    },
    {
      "epoch": 5.555490585896386,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 23750
    },
    {
      "epoch": 5.555724476669395,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6546,
      "step": 23751
    },
    {
      "epoch": 5.555958367442404,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 23752
    },
    {
      "epoch": 5.556192258215414,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 23753
    },
    {
      "epoch": 5.556426148988423,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 23754
    },
    {
      "epoch": 5.556660039761431,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.685,
      "step": 23755
    },
    {
      "epoch": 5.55689393053444,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 23756
    },
    {
      "epoch": 5.557127821307449,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7987,
      "step": 23757
    },
    {
      "epoch": 5.557361712080459,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7413,
      "step": 23758
    },
    {
      "epoch": 5.557595602853468,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 23759
    },
    {
      "epoch": 5.5578294936264765,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6048,
      "step": 23760
    },
    {
      "epoch": 5.558063384399485,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4479,
      "step": 23761
    },
    {
      "epoch": 5.558297275172494,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 23762
    },
    {
      "epoch": 5.558531165945503,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 23763
    },
    {
      "epoch": 5.558765056718513,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7947,
      "step": 23764
    },
    {
      "epoch": 5.558998947491522,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 23765
    },
    {
      "epoch": 5.55923283826453,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 23766
    },
    {
      "epoch": 5.559466729037539,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 23767
    },
    {
      "epoch": 5.559700619810549,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 23768
    },
    {
      "epoch": 5.559934510583558,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8304,
      "step": 23769
    },
    {
      "epoch": 5.560168401356567,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6613,
      "step": 23770
    },
    {
      "epoch": 5.5604022921295755,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 23771
    },
    {
      "epoch": 5.560636182902584,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 23772
    },
    {
      "epoch": 5.560870073675593,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7789,
      "step": 23773
    },
    {
      "epoch": 5.561103964448603,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 23774
    },
    {
      "epoch": 5.561337855221612,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 23775
    },
    {
      "epoch": 5.561571745994621,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 23776
    },
    {
      "epoch": 5.561805636767629,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 23777
    },
    {
      "epoch": 5.562039527540638,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 23778
    },
    {
      "epoch": 5.562273418313648,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5769,
      "step": 23779
    },
    {
      "epoch": 5.562507309086657,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 23780
    },
    {
      "epoch": 5.562741199859666,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 23781
    },
    {
      "epoch": 5.5629750906326745,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 23782
    },
    {
      "epoch": 5.563208981405683,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 23783
    },
    {
      "epoch": 5.563442872178692,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 23784
    },
    {
      "epoch": 5.563676762951702,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 23785
    },
    {
      "epoch": 5.563910653724711,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 23786
    },
    {
      "epoch": 5.56414454449772,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8553,
      "step": 23787
    },
    {
      "epoch": 5.564378435270728,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 23788
    },
    {
      "epoch": 5.564612326043737,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 23789
    },
    {
      "epoch": 5.564846216816747,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 23790
    },
    {
      "epoch": 5.565080107589756,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 23791
    },
    {
      "epoch": 5.565313998362765,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 23792
    },
    {
      "epoch": 5.5655478891357735,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 23793
    },
    {
      "epoch": 5.565781779908782,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 23794
    },
    {
      "epoch": 5.566015670681791,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7934,
      "step": 23795
    },
    {
      "epoch": 5.566249561454801,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 23796
    },
    {
      "epoch": 5.56648345222781,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 23797
    },
    {
      "epoch": 5.5667173430008186,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 23798
    },
    {
      "epoch": 5.566951233773827,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 23799
    },
    {
      "epoch": 5.567185124546837,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 23800
    },
    {
      "epoch": 5.567185124546837,
      "eval_runtime": 4.646,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 23800
    },
    {
      "epoch": 5.567419015319846,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 23801
    },
    {
      "epoch": 5.567652906092855,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6872,
      "step": 23802
    },
    {
      "epoch": 5.567886796865864,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6236,
      "step": 23803
    },
    {
      "epoch": 5.5681206876388725,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 23804
    },
    {
      "epoch": 5.568354578411881,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 23805
    },
    {
      "epoch": 5.568588469184891,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 23806
    },
    {
      "epoch": 5.5688223599579,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7395,
      "step": 23807
    },
    {
      "epoch": 5.569056250730909,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 23808
    },
    {
      "epoch": 5.5692901415039175,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 23809
    },
    {
      "epoch": 5.569524032276926,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 23810
    },
    {
      "epoch": 5.569757923049936,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 23811
    },
    {
      "epoch": 5.569991813822945,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 23812
    },
    {
      "epoch": 5.570225704595954,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7907,
      "step": 23813
    },
    {
      "epoch": 5.570459595368963,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 23814
    },
    {
      "epoch": 5.5706934861419715,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 23815
    },
    {
      "epoch": 5.57092737691498,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6353,
      "step": 23816
    },
    {
      "epoch": 5.57116126768799,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8252,
      "step": 23817
    },
    {
      "epoch": 5.571395158460999,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9209,
      "step": 23818
    },
    {
      "epoch": 5.571629049234008,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 23819
    },
    {
      "epoch": 5.5718629400070165,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7375,
      "step": 23820
    },
    {
      "epoch": 5.572096830780025,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 23821
    },
    {
      "epoch": 5.572330721553035,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.3831,
      "step": 23822
    },
    {
      "epoch": 5.572564612326044,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 23823
    },
    {
      "epoch": 5.572798503099053,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.3318,
      "step": 23824
    },
    {
      "epoch": 5.573032393872062,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 23825
    },
    {
      "epoch": 5.5732662846450705,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 23826
    },
    {
      "epoch": 5.573500175418079,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7131,
      "step": 23827
    },
    {
      "epoch": 5.573734066191089,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6882,
      "step": 23828
    },
    {
      "epoch": 5.573967956964098,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 23829
    },
    {
      "epoch": 5.574201847737107,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 23830
    },
    {
      "epoch": 5.5744357385101155,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 23831
    },
    {
      "epoch": 5.574669629283125,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 23832
    },
    {
      "epoch": 5.574903520056134,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 23833
    },
    {
      "epoch": 5.575137410829143,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 23834
    },
    {
      "epoch": 5.575371301602152,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 23835
    },
    {
      "epoch": 5.575605192375161,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7598,
      "step": 23836
    },
    {
      "epoch": 5.5758390831481695,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5738,
      "step": 23837
    },
    {
      "epoch": 5.576072973921179,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 23838
    },
    {
      "epoch": 5.576306864694188,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 23839
    },
    {
      "epoch": 5.576540755467197,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.6701,
      "step": 23840
    },
    {
      "epoch": 5.576774646240206,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 23841
    },
    {
      "epoch": 5.5770085370132145,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5908,
      "step": 23842
    },
    {
      "epoch": 5.577242427786224,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 23843
    },
    {
      "epoch": 5.577476318559233,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 23844
    },
    {
      "epoch": 5.577710209332242,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 23845
    },
    {
      "epoch": 5.577944100105251,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 23846
    },
    {
      "epoch": 5.57817799087826,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6674,
      "step": 23847
    },
    {
      "epoch": 5.5784118816512684,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6416,
      "step": 23848
    },
    {
      "epoch": 5.578645772424278,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 23849
    },
    {
      "epoch": 5.578879663197287,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 23850
    },
    {
      "epoch": 5.579113553970296,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 23851
    },
    {
      "epoch": 5.579347444743305,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 23852
    },
    {
      "epoch": 5.5795813355163135,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 23853
    },
    {
      "epoch": 5.579815226289323,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6579,
      "step": 23854
    },
    {
      "epoch": 5.580049117062332,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6256,
      "step": 23855
    },
    {
      "epoch": 5.580283007835341,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 23856
    },
    {
      "epoch": 5.58051689860835,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 23857
    },
    {
      "epoch": 5.580750789381359,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 23858
    },
    {
      "epoch": 5.580984680154367,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7501,
      "step": 23859
    },
    {
      "epoch": 5.581218570927377,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6109,
      "step": 23860
    },
    {
      "epoch": 5.581452461700386,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.635,
      "step": 23861
    },
    {
      "epoch": 5.581686352473395,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5912,
      "step": 23862
    },
    {
      "epoch": 5.581920243246404,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 23863
    },
    {
      "epoch": 5.582154134019413,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 23864
    },
    {
      "epoch": 5.582388024792422,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 23865
    },
    {
      "epoch": 5.582621915565431,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 23866
    },
    {
      "epoch": 5.58285580633844,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 23867
    },
    {
      "epoch": 5.583089697111449,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 23868
    },
    {
      "epoch": 5.583323587884458,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5487,
      "step": 23869
    },
    {
      "epoch": 5.583557478657467,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8058,
      "step": 23870
    },
    {
      "epoch": 5.583791369430476,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 23871
    },
    {
      "epoch": 5.584025260203485,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 23872
    },
    {
      "epoch": 5.584259150976494,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6804,
      "step": 23873
    },
    {
      "epoch": 5.584493041749503,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7005,
      "step": 23874
    },
    {
      "epoch": 5.584726932522512,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 23875
    },
    {
      "epoch": 5.584960823295521,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8223,
      "step": 23876
    },
    {
      "epoch": 5.58519471406853,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 23877
    },
    {
      "epoch": 5.585428604841539,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6928,
      "step": 23878
    },
    {
      "epoch": 5.585662495614548,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 23879
    },
    {
      "epoch": 5.585896386387557,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8837,
      "step": 23880
    },
    {
      "epoch": 5.586130277160566,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 23881
    },
    {
      "epoch": 5.586364167933575,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 23882
    },
    {
      "epoch": 5.586598058706584,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 23883
    },
    {
      "epoch": 5.586831949479593,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.23,
      "step": 23884
    },
    {
      "epoch": 5.587065840252602,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 23885
    },
    {
      "epoch": 5.587299731025611,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 23886
    },
    {
      "epoch": 5.58753362179862,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 23887
    },
    {
      "epoch": 5.587767512571629,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6273,
      "step": 23888
    },
    {
      "epoch": 5.588001403344638,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 23889
    },
    {
      "epoch": 5.588235294117647,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7185,
      "step": 23890
    },
    {
      "epoch": 5.588469184890656,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 23891
    },
    {
      "epoch": 5.588703075663665,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 23892
    },
    {
      "epoch": 5.588936966436674,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 23893
    },
    {
      "epoch": 5.589170857209683,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 23894
    },
    {
      "epoch": 5.589404747982692,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.3922,
      "step": 23895
    },
    {
      "epoch": 5.5896386387557015,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 23896
    },
    {
      "epoch": 5.58987252952871,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 23897
    },
    {
      "epoch": 5.590106420301719,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 23898
    },
    {
      "epoch": 5.590340311074728,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 23899
    },
    {
      "epoch": 5.590574201847737,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6511,
      "step": 23900
    },
    {
      "epoch": 5.590574201847737,
      "eval_runtime": 4.5929,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 23900
    },
    {
      "epoch": 5.590808092620746,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 23901
    },
    {
      "epoch": 5.5910419833937555,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 23902
    },
    {
      "epoch": 5.591275874166764,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 23903
    },
    {
      "epoch": 5.591509764939773,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 23904
    },
    {
      "epoch": 5.591743655712782,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 23905
    },
    {
      "epoch": 5.591977546485791,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 23906
    },
    {
      "epoch": 5.5922114372588005,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 23907
    },
    {
      "epoch": 5.592445328031809,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 23908
    },
    {
      "epoch": 5.592679218804818,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 23909
    },
    {
      "epoch": 5.592913109577827,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 23910
    },
    {
      "epoch": 5.593147000350836,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 23911
    },
    {
      "epoch": 5.593380891123845,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 23912
    },
    {
      "epoch": 5.5936147818968545,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 23913
    },
    {
      "epoch": 5.593848672669863,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6562,
      "step": 23914
    },
    {
      "epoch": 5.594082563442872,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 23915
    },
    {
      "epoch": 5.594316454215881,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 23916
    },
    {
      "epoch": 5.59455034498889,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5875,
      "step": 23917
    },
    {
      "epoch": 5.5947842357618995,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 23918
    },
    {
      "epoch": 5.595018126534908,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 23919
    },
    {
      "epoch": 5.595252017307917,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 23920
    },
    {
      "epoch": 5.595485908080926,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8547,
      "step": 23921
    },
    {
      "epoch": 5.595719798853935,
      "grad_norm": 8.125,
      "learning_rate": 3e-05,
      "loss": 2.2565,
      "step": 23922
    },
    {
      "epoch": 5.595953689626944,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7489,
      "step": 23923
    },
    {
      "epoch": 5.5961875803999535,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 23924
    },
    {
      "epoch": 5.596421471172962,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 23925
    },
    {
      "epoch": 5.596655361945971,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5546,
      "step": 23926
    },
    {
      "epoch": 5.59688925271898,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7595,
      "step": 23927
    },
    {
      "epoch": 5.59712314349199,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5225,
      "step": 23928
    },
    {
      "epoch": 5.5973570342649985,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 23929
    },
    {
      "epoch": 5.597590925038007,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 23930
    },
    {
      "epoch": 5.597824815811016,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 23931
    },
    {
      "epoch": 5.598058706584025,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7644,
      "step": 23932
    },
    {
      "epoch": 5.598292597357034,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 23933
    },
    {
      "epoch": 5.598526488130044,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6776,
      "step": 23934
    },
    {
      "epoch": 5.5987603789030524,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6846,
      "step": 23935
    },
    {
      "epoch": 5.598994269676061,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 23936
    },
    {
      "epoch": 5.59922816044907,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4102,
      "step": 23937
    },
    {
      "epoch": 5.599462051222079,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 23938
    },
    {
      "epoch": 5.599695941995089,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 23939
    },
    {
      "epoch": 5.5999298327680975,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 23940
    },
    {
      "epoch": 5.600163723541106,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 23941
    },
    {
      "epoch": 5.600397614314115,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 23942
    },
    {
      "epoch": 5.600631505087124,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6088,
      "step": 23943
    },
    {
      "epoch": 5.600865395860133,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 23944
    },
    {
      "epoch": 5.601099286633143,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 23945
    },
    {
      "epoch": 5.601333177406151,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 23946
    },
    {
      "epoch": 5.60156706817916,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 23947
    },
    {
      "epoch": 5.601800958952169,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 23948
    },
    {
      "epoch": 5.602034849725178,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7664,
      "step": 23949
    },
    {
      "epoch": 5.602268740498188,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6691,
      "step": 23950
    },
    {
      "epoch": 5.6025026312711965,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 23951
    },
    {
      "epoch": 5.602736522044205,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 23952
    },
    {
      "epoch": 5.602970412817214,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 23953
    },
    {
      "epoch": 5.603204303590223,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 23954
    },
    {
      "epoch": 5.603438194363232,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 23955
    },
    {
      "epoch": 5.603672085136242,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 23956
    },
    {
      "epoch": 5.60390597590925,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 23957
    },
    {
      "epoch": 5.604139866682259,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 23958
    },
    {
      "epoch": 5.604373757455268,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 23959
    },
    {
      "epoch": 5.604607648228278,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5185,
      "step": 23960
    },
    {
      "epoch": 5.604841539001287,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 23961
    },
    {
      "epoch": 5.6050754297742955,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 23962
    },
    {
      "epoch": 5.605309320547304,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 23963
    },
    {
      "epoch": 5.605543211320313,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8028,
      "step": 23964
    },
    {
      "epoch": 5.605777102093322,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 23965
    },
    {
      "epoch": 5.606010992866332,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7807,
      "step": 23966
    },
    {
      "epoch": 5.606244883639341,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 23967
    },
    {
      "epoch": 5.606478774412349,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 23968
    },
    {
      "epoch": 5.606712665185358,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 23969
    },
    {
      "epoch": 5.606946555958367,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 23970
    },
    {
      "epoch": 5.607180446731377,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7982,
      "step": 23971
    },
    {
      "epoch": 5.607414337504386,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 23972
    },
    {
      "epoch": 5.6076482282773945,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 23973
    },
    {
      "epoch": 5.607882119050403,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 23974
    },
    {
      "epoch": 5.608116009823412,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6693,
      "step": 23975
    },
    {
      "epoch": 5.608349900596421,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 23976
    },
    {
      "epoch": 5.608583791369431,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 23977
    },
    {
      "epoch": 5.60881768214244,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 23978
    },
    {
      "epoch": 5.609051572915448,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7421,
      "step": 23979
    },
    {
      "epoch": 5.609285463688457,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6749,
      "step": 23980
    },
    {
      "epoch": 5.609519354461467,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 23981
    },
    {
      "epoch": 5.609753245234476,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6585,
      "step": 23982
    },
    {
      "epoch": 5.609987136007485,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 23983
    },
    {
      "epoch": 5.6102210267804935,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 23984
    },
    {
      "epoch": 5.610454917553502,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7103,
      "step": 23985
    },
    {
      "epoch": 5.610688808326511,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6053,
      "step": 23986
    },
    {
      "epoch": 5.610922699099521,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 23987
    },
    {
      "epoch": 5.61115658987253,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2206,
      "step": 23988
    },
    {
      "epoch": 5.611390480645539,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 23989
    },
    {
      "epoch": 5.611624371418547,
      "grad_norm": 7.46875,
      "learning_rate": 3e-05,
      "loss": 1.8089,
      "step": 23990
    },
    {
      "epoch": 5.611858262191556,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6958,
      "step": 23991
    },
    {
      "epoch": 5.612092152964566,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 23992
    },
    {
      "epoch": 5.612326043737575,
      "grad_norm": 8.1875,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 23993
    },
    {
      "epoch": 5.612559934510584,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5926,
      "step": 23994
    },
    {
      "epoch": 5.6127938252835925,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 23995
    },
    {
      "epoch": 5.613027716056601,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 23996
    },
    {
      "epoch": 5.61326160682961,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6105,
      "step": 23997
    },
    {
      "epoch": 5.61349549760262,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 23998
    },
    {
      "epoch": 5.613729388375629,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6461,
      "step": 23999
    },
    {
      "epoch": 5.613963279148638,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 24000
    },
    {
      "epoch": 5.613963279148638,
      "eval_runtime": 4.6654,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 24000
    },
    {
      "epoch": 5.614197169921646,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 24001
    },
    {
      "epoch": 5.614431060694655,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6047,
      "step": 24002
    },
    {
      "epoch": 5.614664951467665,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 24003
    },
    {
      "epoch": 5.614898842240674,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9831,
      "step": 24004
    },
    {
      "epoch": 5.615132733013683,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 24005
    },
    {
      "epoch": 5.6153666237866915,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 24006
    },
    {
      "epoch": 5.6156005145597,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7631,
      "step": 24007
    },
    {
      "epoch": 5.615834405332709,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 24008
    },
    {
      "epoch": 5.616068296105719,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 24009
    },
    {
      "epoch": 5.616302186878728,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 24010
    },
    {
      "epoch": 5.616536077651737,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 24011
    },
    {
      "epoch": 5.616769968424745,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7818,
      "step": 24012
    },
    {
      "epoch": 5.617003859197755,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6425,
      "step": 24013
    },
    {
      "epoch": 5.617237749970764,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 24014
    },
    {
      "epoch": 5.617471640743773,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6881,
      "step": 24015
    },
    {
      "epoch": 5.617705531516782,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 24016
    },
    {
      "epoch": 5.6179394222897905,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7067,
      "step": 24017
    },
    {
      "epoch": 5.618173313062799,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5208,
      "step": 24018
    },
    {
      "epoch": 5.618407203835809,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 24019
    },
    {
      "epoch": 5.618641094608818,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 24020
    },
    {
      "epoch": 5.618874985381827,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 24021
    },
    {
      "epoch": 5.619108876154836,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 24022
    },
    {
      "epoch": 5.619342766927844,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 24023
    },
    {
      "epoch": 5.619576657700854,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 24024
    },
    {
      "epoch": 5.619810548473863,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 24025
    },
    {
      "epoch": 5.620044439246872,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 24026
    },
    {
      "epoch": 5.620278330019881,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 24027
    },
    {
      "epoch": 5.6205122207928895,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 24028
    },
    {
      "epoch": 5.620746111565898,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 24029
    },
    {
      "epoch": 5.620980002338908,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 24030
    },
    {
      "epoch": 5.621213893111917,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 24031
    },
    {
      "epoch": 5.621447783884926,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 24032
    },
    {
      "epoch": 5.621681674657935,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 24033
    },
    {
      "epoch": 5.621915565430943,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.3085,
      "step": 24034
    },
    {
      "epoch": 5.622149456203953,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 24035
    },
    {
      "epoch": 5.622383346976962,
      "grad_norm": 10.25,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 24036
    },
    {
      "epoch": 5.622617237749971,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 24037
    },
    {
      "epoch": 5.62285112852298,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6932,
      "step": 24038
    },
    {
      "epoch": 5.6230850192959885,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 24039
    },
    {
      "epoch": 5.623318910068997,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8354,
      "step": 24040
    },
    {
      "epoch": 5.623552800842007,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6773,
      "step": 24041
    },
    {
      "epoch": 5.623786691615016,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6853,
      "step": 24042
    },
    {
      "epoch": 5.624020582388025,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 24043
    },
    {
      "epoch": 5.6242544731610336,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7587,
      "step": 24044
    },
    {
      "epoch": 5.624488363934043,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 24045
    },
    {
      "epoch": 5.624722254707052,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 24046
    },
    {
      "epoch": 5.624956145480061,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.901,
      "step": 24047
    },
    {
      "epoch": 5.62519003625307,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 24048
    },
    {
      "epoch": 5.625423927026079,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 24049
    },
    {
      "epoch": 5.6256578177990875,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 24050
    },
    {
      "epoch": 5.625891708572097,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3684,
      "step": 24051
    },
    {
      "epoch": 5.626125599345106,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 24052
    },
    {
      "epoch": 5.626359490118115,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 24053
    },
    {
      "epoch": 5.626593380891124,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 24054
    },
    {
      "epoch": 5.6268272716641325,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5074,
      "step": 24055
    },
    {
      "epoch": 5.627061162437142,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 24056
    },
    {
      "epoch": 5.627295053210151,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5862,
      "step": 24057
    },
    {
      "epoch": 5.62752894398316,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 24058
    },
    {
      "epoch": 5.627762834756169,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 24059
    },
    {
      "epoch": 5.627996725529178,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 24060
    },
    {
      "epoch": 5.6282306163021865,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9622,
      "step": 24061
    },
    {
      "epoch": 5.628464507075196,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 24062
    },
    {
      "epoch": 5.628698397848205,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5201,
      "step": 24063
    },
    {
      "epoch": 5.628932288621214,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 24064
    },
    {
      "epoch": 5.629166179394223,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 24065
    },
    {
      "epoch": 5.6294000701672315,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 24066
    },
    {
      "epoch": 5.629633960940241,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 24067
    },
    {
      "epoch": 5.62986785171325,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 24068
    },
    {
      "epoch": 5.630101742486259,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 24069
    },
    {
      "epoch": 5.630335633259268,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 24070
    },
    {
      "epoch": 5.630569524032277,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 24071
    },
    {
      "epoch": 5.6308034148052855,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 24072
    },
    {
      "epoch": 5.631037305578295,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5371,
      "step": 24073
    },
    {
      "epoch": 5.631271196351304,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 24074
    },
    {
      "epoch": 5.631505087124313,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 24075
    },
    {
      "epoch": 5.631738977897322,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 24076
    },
    {
      "epoch": 5.631972868670331,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.548,
      "step": 24077
    },
    {
      "epoch": 5.63220675944334,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 24078
    },
    {
      "epoch": 5.632440650216349,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 24079
    },
    {
      "epoch": 5.632674540989358,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 24080
    },
    {
      "epoch": 5.632908431762367,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 24081
    },
    {
      "epoch": 5.633142322535376,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 24082
    },
    {
      "epoch": 5.633376213308385,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 24083
    },
    {
      "epoch": 5.633610104081394,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7384,
      "step": 24084
    },
    {
      "epoch": 5.633843994854403,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7655,
      "step": 24085
    },
    {
      "epoch": 5.634077885627412,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 24086
    },
    {
      "epoch": 5.634311776400421,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 24087
    },
    {
      "epoch": 5.63454566717343,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 24088
    },
    {
      "epoch": 5.634779557946439,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5265,
      "step": 24089
    },
    {
      "epoch": 5.635013448719448,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 24090
    },
    {
      "epoch": 5.635247339492457,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 24091
    },
    {
      "epoch": 5.635481230265466,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 24092
    },
    {
      "epoch": 5.635715121038475,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.8354,
      "step": 24093
    },
    {
      "epoch": 5.635949011811484,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 24094
    },
    {
      "epoch": 5.636182902584493,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 24095
    },
    {
      "epoch": 5.636416793357502,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 24096
    },
    {
      "epoch": 5.636650684130511,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 24097
    },
    {
      "epoch": 5.63688457490352,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.3324,
      "step": 24098
    },
    {
      "epoch": 5.637118465676529,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7447,
      "step": 24099
    },
    {
      "epoch": 5.637352356449538,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 24100
    },
    {
      "epoch": 5.637352356449538,
      "eval_runtime": 4.6177,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 24100
    },
    {
      "epoch": 5.637586247222547,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6621,
      "step": 24101
    },
    {
      "epoch": 5.637820137995556,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 24102
    },
    {
      "epoch": 5.638054028768565,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 24103
    },
    {
      "epoch": 5.638287919541574,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5721,
      "step": 24104
    },
    {
      "epoch": 5.638521810314583,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6864,
      "step": 24105
    },
    {
      "epoch": 5.638755701087592,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 24106
    },
    {
      "epoch": 5.638989591860601,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6897,
      "step": 24107
    },
    {
      "epoch": 5.63922348263361,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.646,
      "step": 24108
    },
    {
      "epoch": 5.63945737340662,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 24109
    },
    {
      "epoch": 5.639691264179628,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 24110
    },
    {
      "epoch": 5.639925154952637,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 24111
    },
    {
      "epoch": 5.640159045725646,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 24112
    },
    {
      "epoch": 5.640392936498655,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6337,
      "step": 24113
    },
    {
      "epoch": 5.640626827271664,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 24114
    },
    {
      "epoch": 5.6408607180446735,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 24115
    },
    {
      "epoch": 5.641094608817682,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 24116
    },
    {
      "epoch": 5.641328499590691,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 24117
    },
    {
      "epoch": 5.6415623903637,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 24118
    },
    {
      "epoch": 5.641796281136709,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6545,
      "step": 24119
    },
    {
      "epoch": 5.642030171909719,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 24120
    },
    {
      "epoch": 5.642264062682727,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 24121
    },
    {
      "epoch": 5.642497953455736,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 24122
    },
    {
      "epoch": 5.642731844228745,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 24123
    },
    {
      "epoch": 5.642965735001754,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 24124
    },
    {
      "epoch": 5.643199625774763,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 24125
    },
    {
      "epoch": 5.6434335165477725,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 24126
    },
    {
      "epoch": 5.643667407320781,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 24127
    },
    {
      "epoch": 5.64390129809379,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7531,
      "step": 24128
    },
    {
      "epoch": 5.644135188866799,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 24129
    },
    {
      "epoch": 5.644369079639808,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 24130
    },
    {
      "epoch": 5.6446029704128176,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7377,
      "step": 24131
    },
    {
      "epoch": 5.644836861185826,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 24132
    },
    {
      "epoch": 5.645070751958835,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 24133
    },
    {
      "epoch": 5.645304642731844,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 24134
    },
    {
      "epoch": 5.645538533504853,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6654,
      "step": 24135
    },
    {
      "epoch": 5.645772424277862,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 24136
    },
    {
      "epoch": 5.6460063150508715,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6155,
      "step": 24137
    },
    {
      "epoch": 5.64624020582388,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 24138
    },
    {
      "epoch": 5.646474096596889,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 24139
    },
    {
      "epoch": 5.646707987369898,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 24140
    },
    {
      "epoch": 5.646941878142908,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 24141
    },
    {
      "epoch": 5.6471757689159165,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 24142
    },
    {
      "epoch": 5.647409659688925,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6935,
      "step": 24143
    },
    {
      "epoch": 5.647643550461934,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 24144
    },
    {
      "epoch": 5.647877441234943,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5626,
      "step": 24145
    },
    {
      "epoch": 5.648111332007952,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 24146
    },
    {
      "epoch": 5.648345222780962,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8508,
      "step": 24147
    },
    {
      "epoch": 5.6485791135539705,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 24148
    },
    {
      "epoch": 5.648813004326979,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6697,
      "step": 24149
    },
    {
      "epoch": 5.649046895099988,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 24150
    },
    {
      "epoch": 5.649280785872997,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 24151
    },
    {
      "epoch": 5.649514676646007,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5898,
      "step": 24152
    },
    {
      "epoch": 5.6497485674190155,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 24153
    },
    {
      "epoch": 5.649982458192024,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 24154
    },
    {
      "epoch": 5.650216348965033,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 24155
    },
    {
      "epoch": 5.650450239738042,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 24156
    },
    {
      "epoch": 5.650684130511051,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 24157
    },
    {
      "epoch": 5.650918021284061,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 24158
    },
    {
      "epoch": 5.6511519120570695,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 24159
    },
    {
      "epoch": 5.651385802830078,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.4979,
      "step": 24160
    },
    {
      "epoch": 5.651619693603087,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 24161
    },
    {
      "epoch": 5.651853584376097,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 24162
    },
    {
      "epoch": 5.652087475149106,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 24163
    },
    {
      "epoch": 5.6523213659221145,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.7275,
      "step": 24164
    },
    {
      "epoch": 5.652555256695123,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 24165
    },
    {
      "epoch": 5.652789147468132,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 24166
    },
    {
      "epoch": 5.653023038241141,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6609,
      "step": 24167
    },
    {
      "epoch": 5.653256929014151,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4417,
      "step": 24168
    },
    {
      "epoch": 5.65349081978716,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 24169
    },
    {
      "epoch": 5.6537247105601685,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 24170
    },
    {
      "epoch": 5.653958601333177,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6673,
      "step": 24171
    },
    {
      "epoch": 5.654192492106186,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 24172
    },
    {
      "epoch": 5.654426382879196,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7396,
      "step": 24173
    },
    {
      "epoch": 5.654660273652205,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 24174
    },
    {
      "epoch": 5.6548941644252135,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 24175
    },
    {
      "epoch": 5.655128055198222,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 24176
    },
    {
      "epoch": 5.655361945971231,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 24177
    },
    {
      "epoch": 5.65559583674424,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 24178
    },
    {
      "epoch": 5.65582972751725,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 24179
    },
    {
      "epoch": 5.656063618290259,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 24180
    },
    {
      "epoch": 5.6562975090632674,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 24181
    },
    {
      "epoch": 5.656531399836276,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 24182
    },
    {
      "epoch": 5.656765290609285,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7141,
      "step": 24183
    },
    {
      "epoch": 5.656999181382295,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 24184
    },
    {
      "epoch": 5.657233072155304,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.361,
      "step": 24185
    },
    {
      "epoch": 5.6574669629283125,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7074,
      "step": 24186
    },
    {
      "epoch": 5.657700853701321,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 24187
    },
    {
      "epoch": 5.65793474447433,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 24188
    },
    {
      "epoch": 5.658168635247339,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6556,
      "step": 24189
    },
    {
      "epoch": 5.658402526020349,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 24190
    },
    {
      "epoch": 5.658636416793358,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7713,
      "step": 24191
    },
    {
      "epoch": 5.658870307566366,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 24192
    },
    {
      "epoch": 5.659104198339375,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 24193
    },
    {
      "epoch": 5.659338089112385,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5539,
      "step": 24194
    },
    {
      "epoch": 5.659571979885394,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 24195
    },
    {
      "epoch": 5.659805870658403,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 24196
    },
    {
      "epoch": 5.6600397614314115,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 24197
    },
    {
      "epoch": 5.66027365220442,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 24198
    },
    {
      "epoch": 5.660507542977429,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 24199
    },
    {
      "epoch": 5.660741433750439,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 24200
    },
    {
      "epoch": 5.660741433750439,
      "eval_runtime": 4.6077,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 24200
    },
    {
      "epoch": 5.660975324523448,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 24201
    },
    {
      "epoch": 5.661209215296457,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 24202
    },
    {
      "epoch": 5.661443106069465,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 24203
    },
    {
      "epoch": 5.661676996842474,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 24204
    },
    {
      "epoch": 5.661910887615484,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 24205
    },
    {
      "epoch": 5.662144778388493,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 24206
    },
    {
      "epoch": 5.662378669161502,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 24207
    },
    {
      "epoch": 5.6626125599345105,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 24208
    },
    {
      "epoch": 5.662846450707519,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 24209
    },
    {
      "epoch": 5.663080341480528,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2327,
      "step": 24210
    },
    {
      "epoch": 5.663314232253538,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 24211
    },
    {
      "epoch": 5.663548123026547,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 24212
    },
    {
      "epoch": 5.663782013799556,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 24213
    },
    {
      "epoch": 5.664015904572564,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 24214
    },
    {
      "epoch": 5.664249795345573,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2876,
      "step": 24215
    },
    {
      "epoch": 5.664483686118583,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 24216
    },
    {
      "epoch": 5.664717576891592,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 24217
    },
    {
      "epoch": 5.664951467664601,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8011,
      "step": 24218
    },
    {
      "epoch": 5.6651853584376095,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.609,
      "step": 24219
    },
    {
      "epoch": 5.665419249210618,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5404,
      "step": 24220
    },
    {
      "epoch": 5.665653139983627,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6832,
      "step": 24221
    },
    {
      "epoch": 5.665887030756637,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6876,
      "step": 24222
    },
    {
      "epoch": 5.666120921529646,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 24223
    },
    {
      "epoch": 5.666354812302655,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2704,
      "step": 24224
    },
    {
      "epoch": 5.666588703075663,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 24225
    },
    {
      "epoch": 5.666822593848673,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 24226
    },
    {
      "epoch": 5.667056484621682,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.4494,
      "step": 24227
    },
    {
      "epoch": 5.667290375394691,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7803,
      "step": 24228
    },
    {
      "epoch": 5.6675242661677,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 24229
    },
    {
      "epoch": 5.6677581569407085,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5917,
      "step": 24230
    },
    {
      "epoch": 5.667992047713717,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 24231
    },
    {
      "epoch": 5.668225938486727,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 24232
    },
    {
      "epoch": 5.668459829259736,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 24233
    },
    {
      "epoch": 5.668693720032745,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 24234
    },
    {
      "epoch": 5.668927610805754,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 24235
    },
    {
      "epoch": 5.669161501578762,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6447,
      "step": 24236
    },
    {
      "epoch": 5.669395392351772,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 24237
    },
    {
      "epoch": 5.669629283124781,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 24238
    },
    {
      "epoch": 5.66986317389779,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 24239
    },
    {
      "epoch": 5.670097064670799,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6133,
      "step": 24240
    },
    {
      "epoch": 5.6703309554438075,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 24241
    },
    {
      "epoch": 5.670564846216816,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.8535,
      "step": 24242
    },
    {
      "epoch": 5.670798736989826,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6562,
      "step": 24243
    },
    {
      "epoch": 5.671032627762835,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 24244
    },
    {
      "epoch": 5.671266518535844,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 24245
    },
    {
      "epoch": 5.671500409308853,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 24246
    },
    {
      "epoch": 5.671734300081861,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 24247
    },
    {
      "epoch": 5.671968190854871,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 24248
    },
    {
      "epoch": 5.67220208162788,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 24249
    },
    {
      "epoch": 5.672435972400889,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 24250
    },
    {
      "epoch": 5.672669863173898,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5796,
      "step": 24251
    },
    {
      "epoch": 5.6729037539469065,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 24252
    },
    {
      "epoch": 5.673137644719915,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 24253
    },
    {
      "epoch": 5.673371535492925,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 24254
    },
    {
      "epoch": 5.673605426265934,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 24255
    },
    {
      "epoch": 5.673839317038943,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 24256
    },
    {
      "epoch": 5.674073207811952,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6398,
      "step": 24257
    },
    {
      "epoch": 5.674307098584961,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6411,
      "step": 24258
    },
    {
      "epoch": 5.67454098935797,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 24259
    },
    {
      "epoch": 5.674774880130979,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 24260
    },
    {
      "epoch": 5.675008770903988,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.838,
      "step": 24261
    },
    {
      "epoch": 5.675242661676997,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5434,
      "step": 24262
    },
    {
      "epoch": 5.6754765524500055,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 24263
    },
    {
      "epoch": 5.675710443223015,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 24264
    },
    {
      "epoch": 5.675944333996024,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 24265
    },
    {
      "epoch": 5.676178224769033,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 24266
    },
    {
      "epoch": 5.676412115542042,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5546,
      "step": 24267
    },
    {
      "epoch": 5.676646006315051,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 24268
    },
    {
      "epoch": 5.67687989708806,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 24269
    },
    {
      "epoch": 5.677113787861069,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 24270
    },
    {
      "epoch": 5.677347678634078,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 24271
    },
    {
      "epoch": 5.677581569407087,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6165,
      "step": 24272
    },
    {
      "epoch": 5.677815460180096,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 24273
    },
    {
      "epoch": 5.6780493509531045,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 24274
    },
    {
      "epoch": 5.678283241726114,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 24275
    },
    {
      "epoch": 5.678517132499123,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5797,
      "step": 24276
    },
    {
      "epoch": 5.678751023272132,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1611,
      "step": 24277
    },
    {
      "epoch": 5.678984914045141,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 24278
    },
    {
      "epoch": 5.67921880481815,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 24279
    },
    {
      "epoch": 5.679452695591159,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 24280
    },
    {
      "epoch": 5.679686586364168,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6606,
      "step": 24281
    },
    {
      "epoch": 5.679920477137177,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 24282
    },
    {
      "epoch": 5.680154367910186,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7858,
      "step": 24283
    },
    {
      "epoch": 5.680388258683195,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 24284
    },
    {
      "epoch": 5.6806221494562035,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 24285
    },
    {
      "epoch": 5.680856040229213,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 24286
    },
    {
      "epoch": 5.681089931002222,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 24287
    },
    {
      "epoch": 5.681323821775231,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 24288
    },
    {
      "epoch": 5.68155771254824,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 24289
    },
    {
      "epoch": 5.681791603321249,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6831,
      "step": 24290
    },
    {
      "epoch": 5.682025494094258,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4466,
      "step": 24291
    },
    {
      "epoch": 5.682259384867267,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 24292
    },
    {
      "epoch": 5.682493275640276,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 24293
    },
    {
      "epoch": 5.682727166413285,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 24294
    },
    {
      "epoch": 5.682961057186294,
      "grad_norm": 9.4375,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 24295
    },
    {
      "epoch": 5.683194947959303,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 24296
    },
    {
      "epoch": 5.683428838732312,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 24297
    },
    {
      "epoch": 5.683662729505321,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 24298
    },
    {
      "epoch": 5.68389662027833,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 24299
    },
    {
      "epoch": 5.684130511051339,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 24300
    },
    {
      "epoch": 5.684130511051339,
      "eval_runtime": 4.636,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 24300
    },
    {
      "epoch": 5.684364401824348,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7559,
      "step": 24301
    },
    {
      "epoch": 5.684598292597357,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7161,
      "step": 24302
    },
    {
      "epoch": 5.684832183370366,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 24303
    },
    {
      "epoch": 5.685066074143375,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 24304
    },
    {
      "epoch": 5.685299964916384,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 24305
    },
    {
      "epoch": 5.685533855689393,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 24306
    },
    {
      "epoch": 5.685767746462402,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 24307
    },
    {
      "epoch": 5.686001637235411,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9924,
      "step": 24308
    },
    {
      "epoch": 5.68623552800842,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 24309
    },
    {
      "epoch": 5.686469418781429,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 24310
    },
    {
      "epoch": 5.686703309554438,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 24311
    },
    {
      "epoch": 5.686937200327447,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 24312
    },
    {
      "epoch": 5.687171091100456,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6889,
      "step": 24313
    },
    {
      "epoch": 5.687404981873465,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 24314
    },
    {
      "epoch": 5.687638872646474,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 24315
    },
    {
      "epoch": 5.687872763419483,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 24316
    },
    {
      "epoch": 5.688106654192492,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 24317
    },
    {
      "epoch": 5.688340544965501,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 24318
    },
    {
      "epoch": 5.68857443573851,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 24319
    },
    {
      "epoch": 5.688808326511519,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.658,
      "step": 24320
    },
    {
      "epoch": 5.689042217284528,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 24321
    },
    {
      "epoch": 5.689276108057538,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 24322
    },
    {
      "epoch": 5.689509998830546,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 24323
    },
    {
      "epoch": 5.689743889603555,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 24324
    },
    {
      "epoch": 5.689977780376564,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 24325
    },
    {
      "epoch": 5.690211671149573,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 24326
    },
    {
      "epoch": 5.690445561922582,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 24327
    },
    {
      "epoch": 5.6906794526955915,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9563,
      "step": 24328
    },
    {
      "epoch": 5.6909133434686,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 24329
    },
    {
      "epoch": 5.691147234241609,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7267,
      "step": 24330
    },
    {
      "epoch": 5.691381125014618,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 24331
    },
    {
      "epoch": 5.691615015787627,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 24332
    },
    {
      "epoch": 5.691848906560637,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 24333
    },
    {
      "epoch": 5.692082797333645,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 24334
    },
    {
      "epoch": 5.692316688106654,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 24335
    },
    {
      "epoch": 5.692550578879663,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.736,
      "step": 24336
    },
    {
      "epoch": 5.692784469652672,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 24337
    },
    {
      "epoch": 5.693018360425681,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 24338
    },
    {
      "epoch": 5.6932522511986905,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 24339
    },
    {
      "epoch": 5.693486141971699,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8657,
      "step": 24340
    },
    {
      "epoch": 5.693720032744708,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 24341
    },
    {
      "epoch": 5.693953923517717,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 24342
    },
    {
      "epoch": 5.694187814290726,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6538,
      "step": 24343
    },
    {
      "epoch": 5.694421705063736,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 24344
    },
    {
      "epoch": 5.694655595836744,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 24345
    },
    {
      "epoch": 5.694889486609753,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 24346
    },
    {
      "epoch": 5.695123377382762,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 24347
    },
    {
      "epoch": 5.695357268155771,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6012,
      "step": 24348
    },
    {
      "epoch": 5.69559115892878,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 24349
    },
    {
      "epoch": 5.6958250497017895,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5505,
      "step": 24350
    },
    {
      "epoch": 5.696058940474798,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 24351
    },
    {
      "epoch": 5.696292831247807,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7758,
      "step": 24352
    },
    {
      "epoch": 5.696526722020816,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 24353
    },
    {
      "epoch": 5.696760612793826,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 24354
    },
    {
      "epoch": 5.696994503566835,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 24355
    },
    {
      "epoch": 5.697228394339843,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6913,
      "step": 24356
    },
    {
      "epoch": 5.697462285112852,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 24357
    },
    {
      "epoch": 5.697696175885861,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 24358
    },
    {
      "epoch": 5.69793006665887,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 24359
    },
    {
      "epoch": 5.69816395743188,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 24360
    },
    {
      "epoch": 5.6983978482048885,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 24361
    },
    {
      "epoch": 5.698631738977897,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 24362
    },
    {
      "epoch": 5.698865629750906,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 24363
    },
    {
      "epoch": 5.699099520523915,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 24364
    },
    {
      "epoch": 5.699333411296925,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7831,
      "step": 24365
    },
    {
      "epoch": 5.699567302069934,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7112,
      "step": 24366
    },
    {
      "epoch": 5.699801192842942,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 24367
    },
    {
      "epoch": 5.700035083615951,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 24368
    },
    {
      "epoch": 5.70026897438896,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 24369
    },
    {
      "epoch": 5.700502865161969,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 24370
    },
    {
      "epoch": 5.700736755934979,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 24371
    },
    {
      "epoch": 5.7009706467079875,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 24372
    },
    {
      "epoch": 5.701204537480996,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 24373
    },
    {
      "epoch": 5.701438428254005,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7946,
      "step": 24374
    },
    {
      "epoch": 5.701672319027015,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5172,
      "step": 24375
    },
    {
      "epoch": 5.701906209800024,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 24376
    },
    {
      "epoch": 5.7021401005730326,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 24377
    },
    {
      "epoch": 5.702373991346041,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 24378
    },
    {
      "epoch": 5.70260788211905,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 24379
    },
    {
      "epoch": 5.702841772892059,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 24380
    },
    {
      "epoch": 5.703075663665069,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 24381
    },
    {
      "epoch": 5.703309554438078,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 24382
    },
    {
      "epoch": 5.7035434452110865,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 24383
    },
    {
      "epoch": 5.703777335984095,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 24384
    },
    {
      "epoch": 5.704011226757104,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 24385
    },
    {
      "epoch": 5.704245117530114,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8601,
      "step": 24386
    },
    {
      "epoch": 5.704479008303123,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 24387
    },
    {
      "epoch": 5.7047128990761315,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 24388
    },
    {
      "epoch": 5.70494678984914,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 24389
    },
    {
      "epoch": 5.705180680622149,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 24390
    },
    {
      "epoch": 5.705414571395158,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 24391
    },
    {
      "epoch": 5.705648462168168,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6584,
      "step": 24392
    },
    {
      "epoch": 5.705882352941177,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 24393
    },
    {
      "epoch": 5.7061162437141855,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 24394
    },
    {
      "epoch": 5.706350134487194,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 24395
    },
    {
      "epoch": 5.706584025260203,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 24396
    },
    {
      "epoch": 5.706817916033213,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6203,
      "step": 24397
    },
    {
      "epoch": 5.707051806806222,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 24398
    },
    {
      "epoch": 5.7072856975792305,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 24399
    },
    {
      "epoch": 5.707519588352239,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 24400
    },
    {
      "epoch": 5.707519588352239,
      "eval_runtime": 4.5757,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 24400
    },
    {
      "epoch": 5.707753479125248,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9657,
      "step": 24401
    },
    {
      "epoch": 5.707987369898257,
      "grad_norm": 12.4375,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 24402
    },
    {
      "epoch": 5.708221260671267,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.3937,
      "step": 24403
    },
    {
      "epoch": 5.708455151444276,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 24404
    },
    {
      "epoch": 5.7086890422172845,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5265,
      "step": 24405
    },
    {
      "epoch": 5.708922932990293,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7469,
      "step": 24406
    },
    {
      "epoch": 5.709156823763303,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 24407
    },
    {
      "epoch": 5.709390714536312,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 24408
    },
    {
      "epoch": 5.709624605309321,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5724,
      "step": 24409
    },
    {
      "epoch": 5.7098584960823295,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 24410
    },
    {
      "epoch": 5.710092386855338,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7453,
      "step": 24411
    },
    {
      "epoch": 5.710326277628347,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 24412
    },
    {
      "epoch": 5.710560168401357,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 24413
    },
    {
      "epoch": 5.710794059174366,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8468,
      "step": 24414
    },
    {
      "epoch": 5.711027949947375,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 24415
    },
    {
      "epoch": 5.7112618407203835,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 24416
    },
    {
      "epoch": 5.711495731493392,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 24417
    },
    {
      "epoch": 5.711729622266402,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7364,
      "step": 24418
    },
    {
      "epoch": 5.711963513039411,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 24419
    },
    {
      "epoch": 5.71219740381242,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 24420
    },
    {
      "epoch": 5.7124312945854285,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 24421
    },
    {
      "epoch": 5.712665185358437,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 24422
    },
    {
      "epoch": 5.712899076131446,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 24423
    },
    {
      "epoch": 5.713132966904456,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7475,
      "step": 24424
    },
    {
      "epoch": 5.713366857677465,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 24425
    },
    {
      "epoch": 5.713600748450474,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 24426
    },
    {
      "epoch": 5.7138346392234824,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 24427
    },
    {
      "epoch": 5.714068529996491,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 24428
    },
    {
      "epoch": 5.714302420769501,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6262,
      "step": 24429
    },
    {
      "epoch": 5.71453631154251,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8178,
      "step": 24430
    },
    {
      "epoch": 5.714770202315519,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 24431
    },
    {
      "epoch": 5.7150040930885275,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 24432
    },
    {
      "epoch": 5.715237983861536,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 24433
    },
    {
      "epoch": 5.715471874634545,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 24434
    },
    {
      "epoch": 5.715705765407555,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 24435
    },
    {
      "epoch": 5.715939656180564,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 24436
    },
    {
      "epoch": 5.716173546953573,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 24437
    },
    {
      "epoch": 5.716407437726581,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5734,
      "step": 24438
    },
    {
      "epoch": 5.716641328499591,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 24439
    },
    {
      "epoch": 5.7168752192726,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 24440
    },
    {
      "epoch": 5.717109110045609,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.5598,
      "step": 24441
    },
    {
      "epoch": 5.717343000818618,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 24442
    },
    {
      "epoch": 5.7175768915916265,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 24443
    },
    {
      "epoch": 5.717810782364635,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 24444
    },
    {
      "epoch": 5.718044673137645,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 24445
    },
    {
      "epoch": 5.718278563910654,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 24446
    },
    {
      "epoch": 5.718512454683663,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 24447
    },
    {
      "epoch": 5.718746345456672,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5085,
      "step": 24448
    },
    {
      "epoch": 5.71898023622968,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 24449
    },
    {
      "epoch": 5.71921412700269,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 24450
    },
    {
      "epoch": 5.719448017775699,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 24451
    },
    {
      "epoch": 5.719681908548708,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 24452
    },
    {
      "epoch": 5.719915799321717,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.866,
      "step": 24453
    },
    {
      "epoch": 5.7201496900947255,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 24454
    },
    {
      "epoch": 5.720383580867734,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 24455
    },
    {
      "epoch": 5.720617471640744,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 24456
    },
    {
      "epoch": 5.720851362413753,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 24457
    },
    {
      "epoch": 5.721085253186762,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 24458
    },
    {
      "epoch": 5.721319143959771,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 24459
    },
    {
      "epoch": 5.721553034732779,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 24460
    },
    {
      "epoch": 5.721786925505789,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 24461
    },
    {
      "epoch": 5.722020816278798,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 24462
    },
    {
      "epoch": 5.722254707051807,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7267,
      "step": 24463
    },
    {
      "epoch": 5.722488597824816,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 24464
    },
    {
      "epoch": 5.7227224885978245,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 24465
    },
    {
      "epoch": 5.722956379370833,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 24466
    },
    {
      "epoch": 5.723190270143843,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 24467
    },
    {
      "epoch": 5.723424160916852,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 24468
    },
    {
      "epoch": 5.723658051689861,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 24469
    },
    {
      "epoch": 5.72389194246287,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 24470
    },
    {
      "epoch": 5.724125833235879,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 24471
    },
    {
      "epoch": 5.724359724008888,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5505,
      "step": 24472
    },
    {
      "epoch": 5.724593614781897,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 24473
    },
    {
      "epoch": 5.724827505554906,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 24474
    },
    {
      "epoch": 5.725061396327915,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 24475
    },
    {
      "epoch": 5.7252952871009235,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 24476
    },
    {
      "epoch": 5.725529177873933,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 24477
    },
    {
      "epoch": 5.725763068646942,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.788,
      "step": 24478
    },
    {
      "epoch": 5.725996959419951,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0503,
      "step": 24479
    },
    {
      "epoch": 5.72623085019296,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 24480
    },
    {
      "epoch": 5.726464740965969,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.4638,
      "step": 24481
    },
    {
      "epoch": 5.726698631738978,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 24482
    },
    {
      "epoch": 5.726932522511987,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 24483
    },
    {
      "epoch": 5.727166413284996,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.866,
      "step": 24484
    },
    {
      "epoch": 5.727400304058005,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7321,
      "step": 24485
    },
    {
      "epoch": 5.727634194831014,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 24486
    },
    {
      "epoch": 5.7278680856040225,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 24487
    },
    {
      "epoch": 5.728101976377032,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.4176,
      "step": 24488
    },
    {
      "epoch": 5.728335867150041,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 24489
    },
    {
      "epoch": 5.72856975792305,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 24490
    },
    {
      "epoch": 5.728803648696059,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 24491
    },
    {
      "epoch": 5.729037539469068,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 24492
    },
    {
      "epoch": 5.729271430242077,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 24493
    },
    {
      "epoch": 5.729505321015086,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7682,
      "step": 24494
    },
    {
      "epoch": 5.729739211788095,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 24495
    },
    {
      "epoch": 5.729973102561104,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 24496
    },
    {
      "epoch": 5.730206993334113,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 24497
    },
    {
      "epoch": 5.7304408841071215,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 24498
    },
    {
      "epoch": 5.730674774880131,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5598,
      "step": 24499
    },
    {
      "epoch": 5.73090866565314,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6653,
      "step": 24500
    },
    {
      "epoch": 5.73090866565314,
      "eval_runtime": 4.675,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 24500
    },
    {
      "epoch": 5.731142556426149,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3162,
      "step": 24501
    },
    {
      "epoch": 5.731376447199158,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 24502
    },
    {
      "epoch": 5.7316103379721675,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 24503
    },
    {
      "epoch": 5.731844228745176,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 24504
    },
    {
      "epoch": 5.732078119518185,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 24505
    },
    {
      "epoch": 5.732312010291194,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 24506
    },
    {
      "epoch": 5.732545901064203,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 24507
    },
    {
      "epoch": 5.732779791837212,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 24508
    },
    {
      "epoch": 5.733013682610221,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 24509
    },
    {
      "epoch": 5.73324757338323,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 24510
    },
    {
      "epoch": 5.733481464156239,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 24511
    },
    {
      "epoch": 5.733715354929248,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.6949,
      "step": 24512
    },
    {
      "epoch": 5.733949245702257,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 24513
    },
    {
      "epoch": 5.7341831364752665,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8034,
      "step": 24514
    },
    {
      "epoch": 5.734417027248275,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 24515
    },
    {
      "epoch": 5.734650918021284,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5242,
      "step": 24516
    },
    {
      "epoch": 5.734884808794293,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5896,
      "step": 24517
    },
    {
      "epoch": 5.735118699567302,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 24518
    },
    {
      "epoch": 5.735352590340311,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 24519
    },
    {
      "epoch": 5.73558648111332,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.6625,
      "step": 24520
    },
    {
      "epoch": 5.735820371886329,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7526,
      "step": 24521
    },
    {
      "epoch": 5.736054262659338,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 24522
    },
    {
      "epoch": 5.736288153432347,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5207,
      "step": 24523
    },
    {
      "epoch": 5.736522044205356,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 24524
    },
    {
      "epoch": 5.7367559349783654,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7475,
      "step": 24525
    },
    {
      "epoch": 5.736989825751374,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 24526
    },
    {
      "epoch": 5.737223716524383,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.3316,
      "step": 24527
    },
    {
      "epoch": 5.737457607297392,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9622,
      "step": 24528
    },
    {
      "epoch": 5.737691498070401,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 24529
    },
    {
      "epoch": 5.73792538884341,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.6048,
      "step": 24530
    },
    {
      "epoch": 5.738159279616419,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7713,
      "step": 24531
    },
    {
      "epoch": 5.738393170389428,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7367,
      "step": 24532
    },
    {
      "epoch": 5.738627061162437,
      "grad_norm": 8.1875,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 24533
    },
    {
      "epoch": 5.738860951935446,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 24534
    },
    {
      "epoch": 5.739094842708456,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 24535
    },
    {
      "epoch": 5.739328733481464,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 24536
    },
    {
      "epoch": 5.739562624254473,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5892,
      "step": 24537
    },
    {
      "epoch": 5.739796515027482,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 24538
    },
    {
      "epoch": 5.740030405800491,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 24539
    },
    {
      "epoch": 5.7402642965735,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 24540
    },
    {
      "epoch": 5.7404981873465095,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 24541
    },
    {
      "epoch": 5.740732078119518,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 24542
    },
    {
      "epoch": 5.740965968892527,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 24543
    },
    {
      "epoch": 5.741199859665536,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.78,
      "step": 24544
    },
    {
      "epoch": 5.741433750438545,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 24545
    },
    {
      "epoch": 5.741667641211555,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 24546
    },
    {
      "epoch": 5.741901531984563,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 24547
    },
    {
      "epoch": 5.742135422757572,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 24548
    },
    {
      "epoch": 5.742369313530581,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 24549
    },
    {
      "epoch": 5.74260320430359,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6339,
      "step": 24550
    },
    {
      "epoch": 5.742837095076599,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5739,
      "step": 24551
    },
    {
      "epoch": 5.7430709858496085,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6299,
      "step": 24552
    },
    {
      "epoch": 5.743304876622617,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 24553
    },
    {
      "epoch": 5.743538767395626,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 24554
    },
    {
      "epoch": 5.743772658168635,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 24555
    },
    {
      "epoch": 5.744006548941644,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7156,
      "step": 24556
    },
    {
      "epoch": 5.744240439714654,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7608,
      "step": 24557
    },
    {
      "epoch": 5.744474330487662,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 24558
    },
    {
      "epoch": 5.744708221260671,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7525,
      "step": 24559
    },
    {
      "epoch": 5.74494211203368,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 24560
    },
    {
      "epoch": 5.745176002806689,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5845,
      "step": 24561
    },
    {
      "epoch": 5.745409893579699,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6123,
      "step": 24562
    },
    {
      "epoch": 5.7456437843527075,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5409,
      "step": 24563
    },
    {
      "epoch": 5.745877675125716,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 24564
    },
    {
      "epoch": 5.746111565898725,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 24565
    },
    {
      "epoch": 5.746345456671734,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 24566
    },
    {
      "epoch": 5.746579347444744,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 24567
    },
    {
      "epoch": 5.746813238217753,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 24568
    },
    {
      "epoch": 5.747047128990761,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 24569
    },
    {
      "epoch": 5.74728101976377,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 24570
    },
    {
      "epoch": 5.747514910536779,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7296,
      "step": 24571
    },
    {
      "epoch": 5.747748801309788,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 24572
    },
    {
      "epoch": 5.747982692082798,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6885,
      "step": 24573
    },
    {
      "epoch": 5.7482165828558065,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5355,
      "step": 24574
    },
    {
      "epoch": 5.748450473628815,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 24575
    },
    {
      "epoch": 5.748684364401824,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 24576
    },
    {
      "epoch": 5.748918255174833,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 24577
    },
    {
      "epoch": 5.749152145947843,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.713,
      "step": 24578
    },
    {
      "epoch": 5.749386036720852,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 24579
    },
    {
      "epoch": 5.74961992749386,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 24580
    },
    {
      "epoch": 5.749853818266869,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 24581
    },
    {
      "epoch": 5.750087709039878,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 24582
    },
    {
      "epoch": 5.750321599812887,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 24583
    },
    {
      "epoch": 5.750555490585897,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6481,
      "step": 24584
    },
    {
      "epoch": 5.7507893813589055,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 24585
    },
    {
      "epoch": 5.751023272131914,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 24586
    },
    {
      "epoch": 5.751257162904923,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 24587
    },
    {
      "epoch": 5.751491053677933,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6048,
      "step": 24588
    },
    {
      "epoch": 5.751724944450942,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 24589
    },
    {
      "epoch": 5.751958835223951,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 24590
    },
    {
      "epoch": 5.752192725996959,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 24591
    },
    {
      "epoch": 5.752426616769968,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 24592
    },
    {
      "epoch": 5.752660507542977,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 24593
    },
    {
      "epoch": 5.752894398315987,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 24594
    },
    {
      "epoch": 5.753128289088996,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6827,
      "step": 24595
    },
    {
      "epoch": 5.7533621798620045,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4783,
      "step": 24596
    },
    {
      "epoch": 5.753596070635013,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6466,
      "step": 24597
    },
    {
      "epoch": 5.753829961408022,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 24598
    },
    {
      "epoch": 5.754063852181032,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 24599
    },
    {
      "epoch": 5.754297742954041,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 24600
    },
    {
      "epoch": 5.754297742954041,
      "eval_runtime": 4.6141,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 24600
    },
    {
      "epoch": 5.75453163372705,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 24601
    },
    {
      "epoch": 5.754765524500058,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6042,
      "step": 24602
    },
    {
      "epoch": 5.754999415273067,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 24603
    },
    {
      "epoch": 5.755233306046076,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 24604
    },
    {
      "epoch": 5.755467196819086,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 24605
    },
    {
      "epoch": 5.755701087592095,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 24606
    },
    {
      "epoch": 5.7559349783651035,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 24607
    },
    {
      "epoch": 5.756168869138112,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 24608
    },
    {
      "epoch": 5.756402759911121,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 24609
    },
    {
      "epoch": 5.756636650684131,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 24610
    },
    {
      "epoch": 5.75687054145714,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 24611
    },
    {
      "epoch": 5.757104432230149,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 24612
    },
    {
      "epoch": 5.757338323003157,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3195,
      "step": 24613
    },
    {
      "epoch": 5.757572213776166,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5334,
      "step": 24614
    },
    {
      "epoch": 5.757806104549175,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7359,
      "step": 24615
    },
    {
      "epoch": 5.758039995322185,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7931,
      "step": 24616
    },
    {
      "epoch": 5.758273886095194,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 24617
    },
    {
      "epoch": 5.7585077768682025,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 24618
    },
    {
      "epoch": 5.758741667641211,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6673,
      "step": 24619
    },
    {
      "epoch": 5.758975558414221,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 24620
    },
    {
      "epoch": 5.75920944918723,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6283,
      "step": 24621
    },
    {
      "epoch": 5.759443339960239,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 24622
    },
    {
      "epoch": 5.7596772307332476,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 24623
    },
    {
      "epoch": 5.759911121506256,
      "grad_norm": 8.0,
      "learning_rate": 3e-05,
      "loss": 2.6227,
      "step": 24624
    },
    {
      "epoch": 5.760145012279265,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 24625
    },
    {
      "epoch": 5.760378903052275,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 24626
    },
    {
      "epoch": 5.760612793825284,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 24627
    },
    {
      "epoch": 5.760846684598293,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 24628
    },
    {
      "epoch": 5.7610805753713015,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4115,
      "step": 24629
    },
    {
      "epoch": 5.76131446614431,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6921,
      "step": 24630
    },
    {
      "epoch": 5.76154835691732,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 24631
    },
    {
      "epoch": 5.761782247690329,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 24632
    },
    {
      "epoch": 5.762016138463338,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 24633
    },
    {
      "epoch": 5.7622500292363465,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 24634
    },
    {
      "epoch": 5.762483920009355,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 24635
    },
    {
      "epoch": 5.762717810782364,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 24636
    },
    {
      "epoch": 5.762951701555374,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.691,
      "step": 24637
    },
    {
      "epoch": 5.763185592328383,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 24638
    },
    {
      "epoch": 5.763419483101392,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 24639
    },
    {
      "epoch": 5.7636533738744005,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 24640
    },
    {
      "epoch": 5.763887264647409,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 24641
    },
    {
      "epoch": 5.764121155420419,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 24642
    },
    {
      "epoch": 5.764355046193428,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 24643
    },
    {
      "epoch": 5.764588936966437,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 24644
    },
    {
      "epoch": 5.7648228277394455,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 24645
    },
    {
      "epoch": 5.765056718512454,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 24646
    },
    {
      "epoch": 5.765290609285463,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.62,
      "step": 24647
    },
    {
      "epoch": 5.765524500058473,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 24648
    },
    {
      "epoch": 5.765758390831482,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 24649
    },
    {
      "epoch": 5.765992281604491,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 24650
    },
    {
      "epoch": 5.7662261723774995,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 24651
    },
    {
      "epoch": 5.766460063150509,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3187,
      "step": 24652
    },
    {
      "epoch": 5.766693953923518,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 24653
    },
    {
      "epoch": 5.766927844696527,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7936,
      "step": 24654
    },
    {
      "epoch": 5.767161735469536,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 24655
    },
    {
      "epoch": 5.7673956262425445,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 24656
    },
    {
      "epoch": 5.767629517015553,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 24657
    },
    {
      "epoch": 5.767863407788563,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 24658
    },
    {
      "epoch": 5.768097298561572,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 24659
    },
    {
      "epoch": 5.768331189334581,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 24660
    },
    {
      "epoch": 5.76856508010759,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 24661
    },
    {
      "epoch": 5.7687989708805985,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 24662
    },
    {
      "epoch": 5.769032861653608,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7782,
      "step": 24663
    },
    {
      "epoch": 5.769266752426617,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 24664
    },
    {
      "epoch": 5.769500643199626,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 24665
    },
    {
      "epoch": 5.769734533972635,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7275,
      "step": 24666
    },
    {
      "epoch": 5.7699684247456435,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 24667
    },
    {
      "epoch": 5.770202315518652,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.4341,
      "step": 24668
    },
    {
      "epoch": 5.770436206291662,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 24669
    },
    {
      "epoch": 5.770670097064671,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 24670
    },
    {
      "epoch": 5.77090398783768,
      "grad_norm": 9.6875,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 24671
    },
    {
      "epoch": 5.771137878610689,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8141,
      "step": 24672
    },
    {
      "epoch": 5.7713717693836974,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 24673
    },
    {
      "epoch": 5.771605660156707,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 24674
    },
    {
      "epoch": 5.771839550929716,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 24675
    },
    {
      "epoch": 5.772073441702725,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 24676
    },
    {
      "epoch": 5.772307332475734,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 24677
    },
    {
      "epoch": 5.7725412232487425,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 24678
    },
    {
      "epoch": 5.772775114021751,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 24679
    },
    {
      "epoch": 5.773009004794761,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3689,
      "step": 24680
    },
    {
      "epoch": 5.77324289556777,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 24681
    },
    {
      "epoch": 5.773476786340779,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 24682
    },
    {
      "epoch": 5.773710677113788,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 24683
    },
    {
      "epoch": 5.773944567886797,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 24684
    },
    {
      "epoch": 5.774178458659806,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 24685
    },
    {
      "epoch": 5.774412349432815,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.3376,
      "step": 24686
    },
    {
      "epoch": 5.774646240205824,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 24687
    },
    {
      "epoch": 5.774880130978833,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 24688
    },
    {
      "epoch": 5.7751140217518415,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 24689
    },
    {
      "epoch": 5.775347912524851,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0599,
      "step": 24690
    },
    {
      "epoch": 5.77558180329786,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 24691
    },
    {
      "epoch": 5.775815694070869,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8824,
      "step": 24692
    },
    {
      "epoch": 5.776049584843878,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 24693
    },
    {
      "epoch": 5.776283475616887,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.611,
      "step": 24694
    },
    {
      "epoch": 5.776517366389896,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 24695
    },
    {
      "epoch": 5.776751257162905,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 24696
    },
    {
      "epoch": 5.776985147935914,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 24697
    },
    {
      "epoch": 5.777219038708923,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 24698
    },
    {
      "epoch": 5.777452929481932,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7478,
      "step": 24699
    },
    {
      "epoch": 5.7776868202549405,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 24700
    },
    {
      "epoch": 5.7776868202549405,
      "eval_runtime": 4.5797,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 24700
    },
    {
      "epoch": 5.77792071102795,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6838,
      "step": 24701
    },
    {
      "epoch": 5.778154601800959,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 24702
    },
    {
      "epoch": 5.778388492573968,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.741,
      "step": 24703
    },
    {
      "epoch": 5.778622383346977,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6204,
      "step": 24704
    },
    {
      "epoch": 5.778856274119986,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 24705
    },
    {
      "epoch": 5.779090164892995,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 24706
    },
    {
      "epoch": 5.779324055666004,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 24707
    },
    {
      "epoch": 5.779557946439013,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 24708
    },
    {
      "epoch": 5.779791837212022,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 24709
    },
    {
      "epoch": 5.780025727985031,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 24710
    },
    {
      "epoch": 5.7802596187580395,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7474,
      "step": 24711
    },
    {
      "epoch": 5.780493509531049,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.6667,
      "step": 24712
    },
    {
      "epoch": 5.780727400304058,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 24713
    },
    {
      "epoch": 5.780961291077067,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 24714
    },
    {
      "epoch": 5.781195181850076,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 24715
    },
    {
      "epoch": 5.7814290726230855,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6137,
      "step": 24716
    },
    {
      "epoch": 5.781662963396094,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6987,
      "step": 24717
    },
    {
      "epoch": 5.781896854169103,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 24718
    },
    {
      "epoch": 5.782130744942112,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 24719
    },
    {
      "epoch": 5.782364635715121,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6635,
      "step": 24720
    },
    {
      "epoch": 5.78259852648813,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 24721
    },
    {
      "epoch": 5.782832417261139,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 24722
    },
    {
      "epoch": 5.783066308034148,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 24723
    },
    {
      "epoch": 5.783300198807157,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 24724
    },
    {
      "epoch": 5.783534089580166,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.4331,
      "step": 24725
    },
    {
      "epoch": 5.783767980353175,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8101,
      "step": 24726
    },
    {
      "epoch": 5.7840018711261845,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6527,
      "step": 24727
    },
    {
      "epoch": 5.784235761899193,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 24728
    },
    {
      "epoch": 5.784469652672202,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 24729
    },
    {
      "epoch": 5.784703543445211,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 24730
    },
    {
      "epoch": 5.78493743421822,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.372,
      "step": 24731
    },
    {
      "epoch": 5.785171324991229,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 24732
    },
    {
      "epoch": 5.785405215764238,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7499,
      "step": 24733
    },
    {
      "epoch": 5.785639106537247,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 24734
    },
    {
      "epoch": 5.785872997310256,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 24735
    },
    {
      "epoch": 5.786106888083265,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 24736
    },
    {
      "epoch": 5.786340778856274,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 24737
    },
    {
      "epoch": 5.7865746696292835,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 24738
    },
    {
      "epoch": 5.786808560402292,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8337,
      "step": 24739
    },
    {
      "epoch": 5.787042451175301,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 24740
    },
    {
      "epoch": 5.78727634194831,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 24741
    },
    {
      "epoch": 5.787510232721319,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 24742
    },
    {
      "epoch": 5.787744123494328,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 24743
    },
    {
      "epoch": 5.787978014267337,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7397,
      "step": 24744
    },
    {
      "epoch": 5.788211905040346,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4785,
      "step": 24745
    },
    {
      "epoch": 5.788445795813355,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 24746
    },
    {
      "epoch": 5.788679686586364,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 24747
    },
    {
      "epoch": 5.788913577359374,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 24748
    },
    {
      "epoch": 5.7891474681323825,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7524,
      "step": 24749
    },
    {
      "epoch": 5.789381358905391,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 24750
    },
    {
      "epoch": 5.7896152496784,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 24751
    },
    {
      "epoch": 5.789849140451409,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 24752
    },
    {
      "epoch": 5.790083031224418,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 24753
    },
    {
      "epoch": 5.7903169219974275,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 24754
    },
    {
      "epoch": 5.790550812770436,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 24755
    },
    {
      "epoch": 5.790784703543445,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 24756
    },
    {
      "epoch": 5.791018594316454,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 24757
    },
    {
      "epoch": 5.791252485089463,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 24758
    },
    {
      "epoch": 5.791486375862473,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 24759
    },
    {
      "epoch": 5.7917202666354815,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 24760
    },
    {
      "epoch": 5.79195415740849,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 24761
    },
    {
      "epoch": 5.792188048181499,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5363,
      "step": 24762
    },
    {
      "epoch": 5.792421938954508,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 24763
    },
    {
      "epoch": 5.792655829727517,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 24764
    },
    {
      "epoch": 5.7928897205005265,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 24765
    },
    {
      "epoch": 5.793123611273535,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 24766
    },
    {
      "epoch": 5.793357502046544,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 24767
    },
    {
      "epoch": 5.793591392819553,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.755,
      "step": 24768
    },
    {
      "epoch": 5.793825283592563,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5008,
      "step": 24769
    },
    {
      "epoch": 5.794059174365572,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 24770
    },
    {
      "epoch": 5.7942930651385804,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 24771
    },
    {
      "epoch": 5.794526955911589,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5458,
      "step": 24772
    },
    {
      "epoch": 5.794760846684598,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6152,
      "step": 24773
    },
    {
      "epoch": 5.794994737457607,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 24774
    },
    {
      "epoch": 5.795228628230617,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 24775
    },
    {
      "epoch": 5.7954625190036255,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 24776
    },
    {
      "epoch": 5.795696409776634,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 24777
    },
    {
      "epoch": 5.795930300549643,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6225,
      "step": 24778
    },
    {
      "epoch": 5.796164191322652,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5118,
      "step": 24779
    },
    {
      "epoch": 5.796398082095662,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 24780
    },
    {
      "epoch": 5.796631972868671,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 24781
    },
    {
      "epoch": 5.796865863641679,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 24782
    },
    {
      "epoch": 5.797099754414688,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 24783
    },
    {
      "epoch": 5.797333645187697,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7666,
      "step": 24784
    },
    {
      "epoch": 5.797567535960706,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.4327,
      "step": 24785
    },
    {
      "epoch": 5.797801426733716,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8195,
      "step": 24786
    },
    {
      "epoch": 5.7980353175067245,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 24787
    },
    {
      "epoch": 5.798269208279733,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 24788
    },
    {
      "epoch": 5.798503099052742,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6552,
      "step": 24789
    },
    {
      "epoch": 5.798736989825751,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 24790
    },
    {
      "epoch": 5.798970880598761,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 24791
    },
    {
      "epoch": 5.79920477137177,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 24792
    },
    {
      "epoch": 5.799438662144778,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 24793
    },
    {
      "epoch": 5.799672552917787,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 24794
    },
    {
      "epoch": 5.799906443690796,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 24795
    },
    {
      "epoch": 5.800140334463805,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8108,
      "step": 24796
    },
    {
      "epoch": 5.800374225236815,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 24797
    },
    {
      "epoch": 5.8006081160098235,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6935,
      "step": 24798
    },
    {
      "epoch": 5.800842006782832,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 24799
    },
    {
      "epoch": 5.801075897555841,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 24800
    },
    {
      "epoch": 5.801075897555841,
      "eval_runtime": 4.6459,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 24800
    },
    {
      "epoch": 5.801309788328851,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6334,
      "step": 24801
    },
    {
      "epoch": 5.80154367910186,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5714,
      "step": 24802
    },
    {
      "epoch": 5.801777569874869,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 24803
    },
    {
      "epoch": 5.802011460647877,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 24804
    },
    {
      "epoch": 5.802245351420886,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 24805
    },
    {
      "epoch": 5.802479242193895,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 24806
    },
    {
      "epoch": 5.802713132966905,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 24807
    },
    {
      "epoch": 5.802947023739914,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 24808
    },
    {
      "epoch": 5.8031809145129225,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 24809
    },
    {
      "epoch": 5.803414805285931,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 24810
    },
    {
      "epoch": 5.80364869605894,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 24811
    },
    {
      "epoch": 5.80388258683195,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5391,
      "step": 24812
    },
    {
      "epoch": 5.804116477604959,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 24813
    },
    {
      "epoch": 5.804350368377968,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 24814
    },
    {
      "epoch": 5.804584259150976,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 24815
    },
    {
      "epoch": 5.804818149923985,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 24816
    },
    {
      "epoch": 5.805052040696994,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 24817
    },
    {
      "epoch": 5.805285931470004,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 24818
    },
    {
      "epoch": 5.805519822243013,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 24819
    },
    {
      "epoch": 5.8057537130160215,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5558,
      "step": 24820
    },
    {
      "epoch": 5.80598760378903,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7166,
      "step": 24821
    },
    {
      "epoch": 5.806221494562039,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 24822
    },
    {
      "epoch": 5.806455385335049,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 24823
    },
    {
      "epoch": 5.806689276108058,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5964,
      "step": 24824
    },
    {
      "epoch": 5.806923166881067,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 24825
    },
    {
      "epoch": 5.807157057654075,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 24826
    },
    {
      "epoch": 5.807390948427084,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 24827
    },
    {
      "epoch": 5.807624839200093,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 24828
    },
    {
      "epoch": 5.807858729973103,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.702,
      "step": 24829
    },
    {
      "epoch": 5.808092620746112,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 24830
    },
    {
      "epoch": 5.8083265115191205,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 24831
    },
    {
      "epoch": 5.808560402292129,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 24832
    },
    {
      "epoch": 5.808794293065139,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 24833
    },
    {
      "epoch": 5.809028183838148,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 24834
    },
    {
      "epoch": 5.809262074611157,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 24835
    },
    {
      "epoch": 5.809495965384166,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5543,
      "step": 24836
    },
    {
      "epoch": 5.809729856157174,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 24837
    },
    {
      "epoch": 5.809963746930183,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 24838
    },
    {
      "epoch": 5.810197637703193,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 24839
    },
    {
      "epoch": 5.810431528476202,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6719,
      "step": 24840
    },
    {
      "epoch": 5.810665419249211,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5901,
      "step": 24841
    },
    {
      "epoch": 5.8108993100222195,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 24842
    },
    {
      "epoch": 5.811133200795228,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 24843
    },
    {
      "epoch": 5.811367091568238,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 24844
    },
    {
      "epoch": 5.811600982341247,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.66,
      "step": 24845
    },
    {
      "epoch": 5.811834873114256,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 24846
    },
    {
      "epoch": 5.812068763887265,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7854,
      "step": 24847
    },
    {
      "epoch": 5.812302654660273,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 24848
    },
    {
      "epoch": 5.812536545433282,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 24849
    },
    {
      "epoch": 5.812770436206292,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 24850
    },
    {
      "epoch": 5.813004326979301,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 24851
    },
    {
      "epoch": 5.81323821775231,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 24852
    },
    {
      "epoch": 5.8134721085253185,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 24853
    },
    {
      "epoch": 5.813705999298327,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 24854
    },
    {
      "epoch": 5.813939890071337,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 24855
    },
    {
      "epoch": 5.814173780844346,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 24856
    },
    {
      "epoch": 5.814407671617355,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 24857
    },
    {
      "epoch": 5.814641562390364,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 24858
    },
    {
      "epoch": 5.814875453163372,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 24859
    },
    {
      "epoch": 5.815109343936381,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 24860
    },
    {
      "epoch": 5.815343234709391,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 24861
    },
    {
      "epoch": 5.8155771254824,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 24862
    },
    {
      "epoch": 5.815811016255409,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 24863
    },
    {
      "epoch": 5.8160449070284175,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 24864
    },
    {
      "epoch": 5.816278797801427,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 24865
    },
    {
      "epoch": 5.816512688574436,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 24866
    },
    {
      "epoch": 5.816746579347445,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5176,
      "step": 24867
    },
    {
      "epoch": 5.816980470120454,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 24868
    },
    {
      "epoch": 5.8172143608934626,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 24869
    },
    {
      "epoch": 5.817448251666471,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 24870
    },
    {
      "epoch": 5.817682142439481,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 24871
    },
    {
      "epoch": 5.81791603321249,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 24872
    },
    {
      "epoch": 5.818149923985499,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7059,
      "step": 24873
    },
    {
      "epoch": 5.818383814758508,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 24874
    },
    {
      "epoch": 5.8186177055315165,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5485,
      "step": 24875
    },
    {
      "epoch": 5.818851596304526,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 24876
    },
    {
      "epoch": 5.819085487077535,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 24877
    },
    {
      "epoch": 5.819319377850544,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 24878
    },
    {
      "epoch": 5.819553268623553,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 24879
    },
    {
      "epoch": 5.8197871593965615,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 24880
    },
    {
      "epoch": 5.82002105016957,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 24881
    },
    {
      "epoch": 5.82025494094258,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 24882
    },
    {
      "epoch": 5.820488831715589,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 24883
    },
    {
      "epoch": 5.820722722488598,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 24884
    },
    {
      "epoch": 5.820956613261607,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7627,
      "step": 24885
    },
    {
      "epoch": 5.8211905040346155,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 24886
    },
    {
      "epoch": 5.821424394807625,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 24887
    },
    {
      "epoch": 5.821658285580634,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 24888
    },
    {
      "epoch": 5.821892176353643,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.669,
      "step": 24889
    },
    {
      "epoch": 5.822126067126652,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 24890
    },
    {
      "epoch": 5.8223599578996605,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 24891
    },
    {
      "epoch": 5.822593848672669,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 24892
    },
    {
      "epoch": 5.822827739445679,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6852,
      "step": 24893
    },
    {
      "epoch": 5.823061630218688,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6936,
      "step": 24894
    },
    {
      "epoch": 5.823295520991697,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 24895
    },
    {
      "epoch": 5.823529411764706,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4675,
      "step": 24896
    },
    {
      "epoch": 5.823763302537715,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 24897
    },
    {
      "epoch": 5.823997193310724,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.719,
      "step": 24898
    },
    {
      "epoch": 5.824231084083733,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 24899
    },
    {
      "epoch": 5.824464974856742,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6646,
      "step": 24900
    },
    {
      "epoch": 5.824464974856742,
      "eval_runtime": 4.5968,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 24900
    },
    {
      "epoch": 5.824698865629751,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 24901
    },
    {
      "epoch": 5.8249327564027595,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 24902
    },
    {
      "epoch": 5.825166647175769,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 24903
    },
    {
      "epoch": 5.825400537948778,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 24904
    },
    {
      "epoch": 5.825634428721787,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5812,
      "step": 24905
    },
    {
      "epoch": 5.825868319494796,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7947,
      "step": 24906
    },
    {
      "epoch": 5.826102210267805,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 24907
    },
    {
      "epoch": 5.826336101040814,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 24908
    },
    {
      "epoch": 5.826569991813823,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 24909
    },
    {
      "epoch": 5.826803882586832,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 24910
    },
    {
      "epoch": 5.827037773359841,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 24911
    },
    {
      "epoch": 5.82727166413285,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 24912
    },
    {
      "epoch": 5.8275055549058585,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 24913
    },
    {
      "epoch": 5.827739445678868,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6061,
      "step": 24914
    },
    {
      "epoch": 5.827973336451877,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 24915
    },
    {
      "epoch": 5.828207227224886,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7317,
      "step": 24916
    },
    {
      "epoch": 5.828441117997895,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 24917
    },
    {
      "epoch": 5.828675008770904,
      "grad_norm": 2.390625,
      "learning_rate": 3e-05,
      "loss": 1.4546,
      "step": 24918
    },
    {
      "epoch": 5.828908899543913,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 24919
    },
    {
      "epoch": 5.829142790316922,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8289,
      "step": 24920
    },
    {
      "epoch": 5.829376681089931,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7532,
      "step": 24921
    },
    {
      "epoch": 5.82961057186294,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6498,
      "step": 24922
    },
    {
      "epoch": 5.829844462635949,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 24923
    },
    {
      "epoch": 5.8300783534089575,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 24924
    },
    {
      "epoch": 5.830312244181967,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8487,
      "step": 24925
    },
    {
      "epoch": 5.830546134954976,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 24926
    },
    {
      "epoch": 5.830780025727985,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 24927
    },
    {
      "epoch": 5.831013916500994,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6568,
      "step": 24928
    },
    {
      "epoch": 5.8312478072740035,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8653,
      "step": 24929
    },
    {
      "epoch": 5.831481698047012,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 24930
    },
    {
      "epoch": 5.831715588820021,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 24931
    },
    {
      "epoch": 5.83194947959303,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6904,
      "step": 24932
    },
    {
      "epoch": 5.832183370366039,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 24933
    },
    {
      "epoch": 5.832417261139048,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 24934
    },
    {
      "epoch": 5.832651151912057,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 24935
    },
    {
      "epoch": 5.832885042685066,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5207,
      "step": 24936
    },
    {
      "epoch": 5.833118933458075,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 24937
    },
    {
      "epoch": 5.833352824231084,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6232,
      "step": 24938
    },
    {
      "epoch": 5.833586715004093,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 24939
    },
    {
      "epoch": 5.8338206057771025,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 24940
    },
    {
      "epoch": 5.834054496550111,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 24941
    },
    {
      "epoch": 5.83428838732312,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 24942
    },
    {
      "epoch": 5.834522278096129,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 24943
    },
    {
      "epoch": 5.834756168869138,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 24944
    },
    {
      "epoch": 5.834990059642147,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5942,
      "step": 24945
    },
    {
      "epoch": 5.835223950415156,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6961,
      "step": 24946
    },
    {
      "epoch": 5.835457841188165,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 24947
    },
    {
      "epoch": 5.835691731961174,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 24948
    },
    {
      "epoch": 5.835925622734183,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 24949
    },
    {
      "epoch": 5.836159513507192,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 24950
    },
    {
      "epoch": 5.8363934042802015,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8501,
      "step": 24951
    },
    {
      "epoch": 5.83662729505321,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 24952
    },
    {
      "epoch": 5.836861185826219,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 24953
    },
    {
      "epoch": 5.837095076599228,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 24954
    },
    {
      "epoch": 5.837328967372237,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 24955
    },
    {
      "epoch": 5.837562858145247,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 24956
    },
    {
      "epoch": 5.837796748918255,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 24957
    },
    {
      "epoch": 5.838030639691264,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 24958
    },
    {
      "epoch": 5.838264530464273,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 24959
    },
    {
      "epoch": 5.838498421237282,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 24960
    },
    {
      "epoch": 5.838732312010292,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 24961
    },
    {
      "epoch": 5.8389662027833005,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 24962
    },
    {
      "epoch": 5.839200093556309,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 2.3095,
      "step": 24963
    },
    {
      "epoch": 5.839433984329318,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 24964
    },
    {
      "epoch": 5.839667875102327,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.607,
      "step": 24965
    },
    {
      "epoch": 5.839901765875336,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 24966
    },
    {
      "epoch": 5.8401356566483456,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 24967
    },
    {
      "epoch": 5.840369547421354,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4,
      "step": 24968
    },
    {
      "epoch": 5.840603438194363,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7701,
      "step": 24969
    },
    {
      "epoch": 5.840837328967372,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 24970
    },
    {
      "epoch": 5.841071219740381,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6188,
      "step": 24971
    },
    {
      "epoch": 5.841305110513391,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 24972
    },
    {
      "epoch": 5.8415390012863995,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6364,
      "step": 24973
    },
    {
      "epoch": 5.841772892059408,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5724,
      "step": 24974
    },
    {
      "epoch": 5.842006782832417,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 24975
    },
    {
      "epoch": 5.842240673605426,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6049,
      "step": 24976
    },
    {
      "epoch": 5.842474564378435,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 24977
    },
    {
      "epoch": 5.8427084551514445,
      "grad_norm": 10.3125,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 24978
    },
    {
      "epoch": 5.842942345924453,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 24979
    },
    {
      "epoch": 5.843176236697462,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 24980
    },
    {
      "epoch": 5.843410127470471,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6092,
      "step": 24981
    },
    {
      "epoch": 5.843644018243481,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 24982
    },
    {
      "epoch": 5.84387790901649,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 24983
    },
    {
      "epoch": 5.8441117997894985,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 24984
    },
    {
      "epoch": 5.844345690562507,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 24985
    },
    {
      "epoch": 5.844579581335516,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 24986
    },
    {
      "epoch": 5.844813472108525,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.62,
      "step": 24987
    },
    {
      "epoch": 5.845047362881535,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 24988
    },
    {
      "epoch": 5.8452812536545435,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6396,
      "step": 24989
    },
    {
      "epoch": 5.845515144427552,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 24990
    },
    {
      "epoch": 5.845749035200561,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 24991
    },
    {
      "epoch": 5.84598292597357,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 24992
    },
    {
      "epoch": 5.84621681674658,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 24993
    },
    {
      "epoch": 5.846450707519589,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6672,
      "step": 24994
    },
    {
      "epoch": 5.8466845982925975,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 24995
    },
    {
      "epoch": 5.846918489065606,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 24996
    },
    {
      "epoch": 5.847152379838615,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 24997
    },
    {
      "epoch": 5.847386270611624,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6428,
      "step": 24998
    },
    {
      "epoch": 5.847620161384634,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1838,
      "step": 24999
    },
    {
      "epoch": 5.8478540521576425,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 25000
    },
    {
      "epoch": 5.8478540521576425,
      "eval_runtime": 4.6284,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 25000
    },
    {
      "epoch": 5.848087942930651,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5436,
      "step": 25001
    },
    {
      "epoch": 5.84832183370366,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 25002
    },
    {
      "epoch": 5.848555724476669,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 25003
    },
    {
      "epoch": 5.848789615249679,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 25004
    },
    {
      "epoch": 5.849023506022688,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 25005
    },
    {
      "epoch": 5.8492573967956965,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 25006
    },
    {
      "epoch": 5.849491287568705,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 25007
    },
    {
      "epoch": 5.849725178341714,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.655,
      "step": 25008
    },
    {
      "epoch": 5.849959069114723,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 25009
    },
    {
      "epoch": 5.850192959887733,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 25010
    },
    {
      "epoch": 5.8504268506607415,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 25011
    },
    {
      "epoch": 5.85066074143375,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6395,
      "step": 25012
    },
    {
      "epoch": 5.850894632206759,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 25013
    },
    {
      "epoch": 5.851128522979769,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 25014
    },
    {
      "epoch": 5.851362413752778,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 25015
    },
    {
      "epoch": 5.851596304525787,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 25016
    },
    {
      "epoch": 5.8518301952987954,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 25017
    },
    {
      "epoch": 5.852064086071804,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 25018
    },
    {
      "epoch": 5.852297976844813,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 25019
    },
    {
      "epoch": 5.852531867617823,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 25020
    },
    {
      "epoch": 5.852765758390832,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 25021
    },
    {
      "epoch": 5.8529996491638405,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 25022
    },
    {
      "epoch": 5.853233539936849,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.3886,
      "step": 25023
    },
    {
      "epoch": 5.853467430709858,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 25024
    },
    {
      "epoch": 5.853701321482868,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 25025
    },
    {
      "epoch": 5.853935212255877,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4692,
      "step": 25026
    },
    {
      "epoch": 5.854169103028886,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.2981,
      "step": 25027
    },
    {
      "epoch": 5.854402993801894,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 25028
    },
    {
      "epoch": 5.854636884574903,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 25029
    },
    {
      "epoch": 5.854870775347912,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.5222,
      "step": 25030
    },
    {
      "epoch": 5.855104666120922,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5571,
      "step": 25031
    },
    {
      "epoch": 5.855338556893931,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 25032
    },
    {
      "epoch": 5.8555724476669395,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.57,
      "step": 25033
    },
    {
      "epoch": 5.855806338439948,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5957,
      "step": 25034
    },
    {
      "epoch": 5.856040229212957,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 25035
    },
    {
      "epoch": 5.856274119985967,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2589,
      "step": 25036
    },
    {
      "epoch": 5.856508010758976,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 25037
    },
    {
      "epoch": 5.856741901531985,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 25038
    },
    {
      "epoch": 5.856975792304993,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.951,
      "step": 25039
    },
    {
      "epoch": 5.857209683078002,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 25040
    },
    {
      "epoch": 5.857443573851011,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 25041
    },
    {
      "epoch": 5.857677464624021,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8234,
      "step": 25042
    },
    {
      "epoch": 5.85791135539703,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 25043
    },
    {
      "epoch": 5.8581452461700385,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 25044
    },
    {
      "epoch": 5.858379136943047,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 25045
    },
    {
      "epoch": 5.858613027716057,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 25046
    },
    {
      "epoch": 5.858846918489066,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 25047
    },
    {
      "epoch": 5.859080809262075,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 25048
    },
    {
      "epoch": 5.859314700035084,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6051,
      "step": 25049
    },
    {
      "epoch": 5.859548590808092,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 25050
    },
    {
      "epoch": 5.859782481581101,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 25051
    },
    {
      "epoch": 5.860016372354111,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 25052
    },
    {
      "epoch": 5.86025026312712,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 25053
    },
    {
      "epoch": 5.860484153900129,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 25054
    },
    {
      "epoch": 5.8607180446731375,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.922,
      "step": 25055
    },
    {
      "epoch": 5.860951935446146,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 25056
    },
    {
      "epoch": 5.861185826219156,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 25057
    },
    {
      "epoch": 5.861419716992165,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4571,
      "step": 25058
    },
    {
      "epoch": 5.861653607765174,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7401,
      "step": 25059
    },
    {
      "epoch": 5.861887498538183,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 25060
    },
    {
      "epoch": 5.862121389311191,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 25061
    },
    {
      "epoch": 5.8623552800842,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 25062
    },
    {
      "epoch": 5.86258917085721,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2586,
      "step": 25063
    },
    {
      "epoch": 5.862823061630219,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 25064
    },
    {
      "epoch": 5.863056952403228,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5554,
      "step": 25065
    },
    {
      "epoch": 5.8632908431762365,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7148,
      "step": 25066
    },
    {
      "epoch": 5.863524733949245,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 25067
    },
    {
      "epoch": 5.863758624722255,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6586,
      "step": 25068
    },
    {
      "epoch": 5.863992515495264,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 25069
    },
    {
      "epoch": 5.864226406268273,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 25070
    },
    {
      "epoch": 5.864460297041282,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 25071
    },
    {
      "epoch": 5.86469418781429,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 25072
    },
    {
      "epoch": 5.864928078587299,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 25073
    },
    {
      "epoch": 5.865161969360309,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4719,
      "step": 25074
    },
    {
      "epoch": 5.865395860133318,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 25075
    },
    {
      "epoch": 5.865629750906327,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 25076
    },
    {
      "epoch": 5.8658636416793355,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6809,
      "step": 25077
    },
    {
      "epoch": 5.866097532452345,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6619,
      "step": 25078
    },
    {
      "epoch": 5.866331423225354,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.3332,
      "step": 25079
    },
    {
      "epoch": 5.866565313998363,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 25080
    },
    {
      "epoch": 5.866799204771372,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7455,
      "step": 25081
    },
    {
      "epoch": 5.867033095544381,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 25082
    },
    {
      "epoch": 5.867266986317389,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0032,
      "step": 25083
    },
    {
      "epoch": 5.867500877090399,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 25084
    },
    {
      "epoch": 5.867734767863408,
      "grad_norm": 8.3125,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 25085
    },
    {
      "epoch": 5.867968658636417,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5104,
      "step": 25086
    },
    {
      "epoch": 5.868202549409426,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 25087
    },
    {
      "epoch": 5.8684364401824345,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 25088
    },
    {
      "epoch": 5.868670330955444,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 25089
    },
    {
      "epoch": 5.868904221728453,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7424,
      "step": 25090
    },
    {
      "epoch": 5.869138112501462,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 25091
    },
    {
      "epoch": 5.869372003274471,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6305,
      "step": 25092
    },
    {
      "epoch": 5.86960589404748,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 25093
    },
    {
      "epoch": 5.869839784820488,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 25094
    },
    {
      "epoch": 5.870073675593498,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 25095
    },
    {
      "epoch": 5.870307566366507,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7355,
      "step": 25096
    },
    {
      "epoch": 5.870541457139516,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 25097
    },
    {
      "epoch": 5.870775347912525,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 25098
    },
    {
      "epoch": 5.8710092386855335,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.5833,
      "step": 25099
    },
    {
      "epoch": 5.871243129458543,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5166,
      "step": 25100
    },
    {
      "epoch": 5.871243129458543,
      "eval_runtime": 4.6231,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 25100
    },
    {
      "epoch": 5.871477020231552,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 25101
    },
    {
      "epoch": 5.871710911004561,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 25102
    },
    {
      "epoch": 5.87194480177757,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 25103
    },
    {
      "epoch": 5.872178692550579,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5528,
      "step": 25104
    },
    {
      "epoch": 5.872412583323587,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5977,
      "step": 25105
    },
    {
      "epoch": 5.872646474096597,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 25106
    },
    {
      "epoch": 5.872880364869606,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 25107
    },
    {
      "epoch": 5.873114255642615,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5711,
      "step": 25108
    },
    {
      "epoch": 5.873348146415624,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 25109
    },
    {
      "epoch": 5.873582037188633,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 25110
    },
    {
      "epoch": 5.873815927961642,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 25111
    },
    {
      "epoch": 5.874049818734651,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 25112
    },
    {
      "epoch": 5.87428370950766,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 25113
    },
    {
      "epoch": 5.874517600280669,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 25114
    },
    {
      "epoch": 5.8747514910536776,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 25115
    },
    {
      "epoch": 5.874985381826687,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 25116
    },
    {
      "epoch": 5.875219272599696,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 25117
    },
    {
      "epoch": 5.875453163372705,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 25118
    },
    {
      "epoch": 5.875687054145714,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 25119
    },
    {
      "epoch": 5.875920944918723,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.498,
      "step": 25120
    },
    {
      "epoch": 5.876154835691732,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 25121
    },
    {
      "epoch": 5.876388726464741,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6688,
      "step": 25122
    },
    {
      "epoch": 5.87662261723775,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2009,
      "step": 25123
    },
    {
      "epoch": 5.876856508010759,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 25124
    },
    {
      "epoch": 5.877090398783768,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 25125
    },
    {
      "epoch": 5.8773242895567765,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 25126
    },
    {
      "epoch": 5.877558180329786,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 25127
    },
    {
      "epoch": 5.877792071102795,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.1958,
      "step": 25128
    },
    {
      "epoch": 5.878025961875804,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6741,
      "step": 25129
    },
    {
      "epoch": 5.878259852648813,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 25130
    },
    {
      "epoch": 5.878493743421822,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 25131
    },
    {
      "epoch": 5.878727634194831,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 25132
    },
    {
      "epoch": 5.87896152496784,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4983,
      "step": 25133
    },
    {
      "epoch": 5.879195415740849,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 25134
    },
    {
      "epoch": 5.879429306513858,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 25135
    },
    {
      "epoch": 5.879663197286867,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 1.6018,
      "step": 25136
    },
    {
      "epoch": 5.8798970880598755,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 25137
    },
    {
      "epoch": 5.880130978832885,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 25138
    },
    {
      "epoch": 5.880364869605894,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 25139
    },
    {
      "epoch": 5.880598760378903,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 25140
    },
    {
      "epoch": 5.880832651151912,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6764,
      "step": 25141
    },
    {
      "epoch": 5.8810665419249215,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.798,
      "step": 25142
    },
    {
      "epoch": 5.88130043269793,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 25143
    },
    {
      "epoch": 5.881534323470939,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5986,
      "step": 25144
    },
    {
      "epoch": 5.881768214243948,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.678,
      "step": 25145
    },
    {
      "epoch": 5.882002105016957,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 25146
    },
    {
      "epoch": 5.882235995789966,
      "grad_norm": 10.0,
      "learning_rate": 3e-05,
      "loss": 1.8001,
      "step": 25147
    },
    {
      "epoch": 5.882469886562975,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 25148
    },
    {
      "epoch": 5.882703777335984,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7352,
      "step": 25149
    },
    {
      "epoch": 5.882937668108993,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 25150
    },
    {
      "epoch": 5.883171558882002,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 25151
    },
    {
      "epoch": 5.883405449655011,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7282,
      "step": 25152
    },
    {
      "epoch": 5.8836393404280205,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 25153
    },
    {
      "epoch": 5.883873231201029,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 25154
    },
    {
      "epoch": 5.884107121974038,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 25155
    },
    {
      "epoch": 5.884341012747047,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 25156
    },
    {
      "epoch": 5.884574903520056,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 25157
    },
    {
      "epoch": 5.884808794293065,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 25158
    },
    {
      "epoch": 5.885042685066074,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 25159
    },
    {
      "epoch": 5.885276575839083,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7395,
      "step": 25160
    },
    {
      "epoch": 5.885510466612092,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6373,
      "step": 25161
    },
    {
      "epoch": 5.885744357385101,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 25162
    },
    {
      "epoch": 5.885978248158111,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 25163
    },
    {
      "epoch": 5.8862121389311195,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.3294,
      "step": 25164
    },
    {
      "epoch": 5.886446029704128,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 25165
    },
    {
      "epoch": 5.886679920477137,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6857,
      "step": 25166
    },
    {
      "epoch": 5.886913811250146,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 25167
    },
    {
      "epoch": 5.887147702023155,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 25168
    },
    {
      "epoch": 5.887381592796165,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8528,
      "step": 25169
    },
    {
      "epoch": 5.887615483569173,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6824,
      "step": 25170
    },
    {
      "epoch": 5.887849374342182,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.465,
      "step": 25171
    },
    {
      "epoch": 5.888083265115191,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 25172
    },
    {
      "epoch": 5.8883171558882,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 25173
    },
    {
      "epoch": 5.88855104666121,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 25174
    },
    {
      "epoch": 5.8887849374342185,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 25175
    },
    {
      "epoch": 5.889018828207227,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 25176
    },
    {
      "epoch": 5.889252718980236,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 25177
    },
    {
      "epoch": 5.889486609753245,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 25178
    },
    {
      "epoch": 5.889720500526254,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7481,
      "step": 25179
    },
    {
      "epoch": 5.889954391299264,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 25180
    },
    {
      "epoch": 5.890188282072272,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8038,
      "step": 25181
    },
    {
      "epoch": 5.890422172845281,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6402,
      "step": 25182
    },
    {
      "epoch": 5.89065606361829,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 25183
    },
    {
      "epoch": 5.890889954391299,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 25184
    },
    {
      "epoch": 5.891123845164309,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 25185
    },
    {
      "epoch": 5.8913577359373175,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 25186
    },
    {
      "epoch": 5.891591626710326,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 25187
    },
    {
      "epoch": 5.891825517483335,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 25188
    },
    {
      "epoch": 5.892059408256344,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6448,
      "step": 25189
    },
    {
      "epoch": 5.892293299029353,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 25190
    },
    {
      "epoch": 5.892527189802363,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 25191
    },
    {
      "epoch": 5.892761080575371,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 25192
    },
    {
      "epoch": 5.89299497134838,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 25193
    },
    {
      "epoch": 5.893228862121389,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 25194
    },
    {
      "epoch": 5.893462752894399,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 25195
    },
    {
      "epoch": 5.893696643667408,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 25196
    },
    {
      "epoch": 5.8939305344404165,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 25197
    },
    {
      "epoch": 5.894164425213425,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 25198
    },
    {
      "epoch": 5.894398315986434,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0492,
      "step": 25199
    },
    {
      "epoch": 5.894632206759443,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 25200
    },
    {
      "epoch": 5.894632206759443,
      "eval_runtime": 4.6015,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 25200
    },
    {
      "epoch": 5.894866097532453,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 25201
    },
    {
      "epoch": 5.895099988305462,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 25202
    },
    {
      "epoch": 5.89533387907847,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5583,
      "step": 25203
    },
    {
      "epoch": 5.895567769851479,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 25204
    },
    {
      "epoch": 5.895801660624488,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.697,
      "step": 25205
    },
    {
      "epoch": 5.896035551397498,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.717,
      "step": 25206
    },
    {
      "epoch": 5.896269442170507,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5013,
      "step": 25207
    },
    {
      "epoch": 5.8965033329435155,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5919,
      "step": 25208
    },
    {
      "epoch": 5.896737223716524,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 25209
    },
    {
      "epoch": 5.896971114489533,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 25210
    },
    {
      "epoch": 5.897205005262542,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5619,
      "step": 25211
    },
    {
      "epoch": 5.897438896035552,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 25212
    },
    {
      "epoch": 5.8976727868085606,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 25213
    },
    {
      "epoch": 5.897906677581569,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 25214
    },
    {
      "epoch": 5.898140568354578,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6451,
      "step": 25215
    },
    {
      "epoch": 5.898374459127587,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6448,
      "step": 25216
    },
    {
      "epoch": 5.898608349900597,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5244,
      "step": 25217
    },
    {
      "epoch": 5.898842240673606,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 25218
    },
    {
      "epoch": 5.8990761314466145,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 25219
    },
    {
      "epoch": 5.899310022219623,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 25220
    },
    {
      "epoch": 5.899543912992632,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 25221
    },
    {
      "epoch": 5.899777803765641,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 25222
    },
    {
      "epoch": 5.900011694538651,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 25223
    },
    {
      "epoch": 5.9002455853116595,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 25224
    },
    {
      "epoch": 5.900479476084668,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 25225
    },
    {
      "epoch": 5.900713366857677,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2681,
      "step": 25226
    },
    {
      "epoch": 5.900947257630687,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1757,
      "step": 25227
    },
    {
      "epoch": 5.901181148403696,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.647,
      "step": 25228
    },
    {
      "epoch": 5.901415039176705,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 2.3275,
      "step": 25229
    },
    {
      "epoch": 5.9016489299497135,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 25230
    },
    {
      "epoch": 5.901882820722722,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7549,
      "step": 25231
    },
    {
      "epoch": 5.902116711495731,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 25232
    },
    {
      "epoch": 5.902350602268741,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 25233
    },
    {
      "epoch": 5.90258449304175,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 25234
    },
    {
      "epoch": 5.9028183838147585,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 25235
    },
    {
      "epoch": 5.903052274587767,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 25236
    },
    {
      "epoch": 5.903286165360776,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 25237
    },
    {
      "epoch": 5.903520056133786,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5967,
      "step": 25238
    },
    {
      "epoch": 5.903753946906795,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 25239
    },
    {
      "epoch": 5.903987837679804,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 25240
    },
    {
      "epoch": 5.9042217284528125,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6671,
      "step": 25241
    },
    {
      "epoch": 5.904455619225821,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 25242
    },
    {
      "epoch": 5.90468950999883,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 25243
    },
    {
      "epoch": 5.90492340077184,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 25244
    },
    {
      "epoch": 5.905157291544849,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.5653,
      "step": 25245
    },
    {
      "epoch": 5.9053911823178575,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7831,
      "step": 25246
    },
    {
      "epoch": 5.905625073090866,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 25247
    },
    {
      "epoch": 5.905858963863875,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 25248
    },
    {
      "epoch": 5.906092854636885,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 25249
    },
    {
      "epoch": 5.906326745409894,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 25250
    },
    {
      "epoch": 5.906560636182903,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6686,
      "step": 25251
    },
    {
      "epoch": 5.9067945269559115,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 25252
    },
    {
      "epoch": 5.90702841772892,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8409,
      "step": 25253
    },
    {
      "epoch": 5.907262308501929,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8409,
      "step": 25254
    },
    {
      "epoch": 5.907496199274939,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 25255
    },
    {
      "epoch": 5.907730090047948,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 25256
    },
    {
      "epoch": 5.9079639808209565,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.922,
      "step": 25257
    },
    {
      "epoch": 5.908197871593965,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7632,
      "step": 25258
    },
    {
      "epoch": 5.908431762366975,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6936,
      "step": 25259
    },
    {
      "epoch": 5.908665653139984,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 25260
    },
    {
      "epoch": 5.908899543912993,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 25261
    },
    {
      "epoch": 5.909133434686002,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5897,
      "step": 25262
    },
    {
      "epoch": 5.9093673254590104,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 25263
    },
    {
      "epoch": 5.909601216232019,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.4465,
      "step": 25264
    },
    {
      "epoch": 5.909835107005029,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 25265
    },
    {
      "epoch": 5.910068997778038,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 25266
    },
    {
      "epoch": 5.910302888551047,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 25267
    },
    {
      "epoch": 5.9105367793240555,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 25268
    },
    {
      "epoch": 5.910770670097064,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 25269
    },
    {
      "epoch": 5.911004560870074,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 25270
    },
    {
      "epoch": 5.911238451643083,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 25271
    },
    {
      "epoch": 5.911472342416092,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 25272
    },
    {
      "epoch": 5.911706233189101,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5692,
      "step": 25273
    },
    {
      "epoch": 5.911940123962109,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6309,
      "step": 25274
    },
    {
      "epoch": 5.912174014735118,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3607,
      "step": 25275
    },
    {
      "epoch": 5.912407905508128,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8645,
      "step": 25276
    },
    {
      "epoch": 5.912641796281137,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 25277
    },
    {
      "epoch": 5.912875687054146,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 25278
    },
    {
      "epoch": 5.9131095778271545,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 25279
    },
    {
      "epoch": 5.913343468600163,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 25280
    },
    {
      "epoch": 5.913577359373173,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 25281
    },
    {
      "epoch": 5.913811250146182,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1483,
      "step": 25282
    },
    {
      "epoch": 5.914045140919191,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8131,
      "step": 25283
    },
    {
      "epoch": 5.9142790316922,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 25284
    },
    {
      "epoch": 5.914512922465208,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 25285
    },
    {
      "epoch": 5.914746813238217,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7244,
      "step": 25286
    },
    {
      "epoch": 5.914980704011227,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6761,
      "step": 25287
    },
    {
      "epoch": 5.915214594784236,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 25288
    },
    {
      "epoch": 5.915448485557245,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 25289
    },
    {
      "epoch": 5.9156823763302535,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 25290
    },
    {
      "epoch": 5.915916267103263,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 25291
    },
    {
      "epoch": 5.916150157876272,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 25292
    },
    {
      "epoch": 5.916384048649281,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 25293
    },
    {
      "epoch": 5.91661793942229,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5625,
      "step": 25294
    },
    {
      "epoch": 5.916851830195299,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 25295
    },
    {
      "epoch": 5.917085720968307,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5753,
      "step": 25296
    },
    {
      "epoch": 5.917319611741317,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 25297
    },
    {
      "epoch": 5.917553502514326,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5729,
      "step": 25298
    },
    {
      "epoch": 5.917787393287335,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5457,
      "step": 25299
    },
    {
      "epoch": 5.918021284060344,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6402,
      "step": 25300
    },
    {
      "epoch": 5.918021284060344,
      "eval_runtime": 4.6491,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 25300
    },
    {
      "epoch": 5.9182551748333525,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 25301
    },
    {
      "epoch": 5.918489065606362,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 25302
    },
    {
      "epoch": 5.918722956379371,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 25303
    },
    {
      "epoch": 5.91895684715238,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 25304
    },
    {
      "epoch": 5.919190737925389,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 25305
    },
    {
      "epoch": 5.919424628698398,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9322,
      "step": 25306
    },
    {
      "epoch": 5.919658519471406,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.436,
      "step": 25307
    },
    {
      "epoch": 5.919892410244416,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 25308
    },
    {
      "epoch": 5.920126301017425,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 25309
    },
    {
      "epoch": 5.920360191790434,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 25310
    },
    {
      "epoch": 5.920594082563443,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 25311
    },
    {
      "epoch": 5.9208279733364515,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 25312
    },
    {
      "epoch": 5.921061864109461,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 25313
    },
    {
      "epoch": 5.92129575488247,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 25314
    },
    {
      "epoch": 5.921529645655479,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 25315
    },
    {
      "epoch": 5.921763536428488,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5887,
      "step": 25316
    },
    {
      "epoch": 5.921997427201497,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 25317
    },
    {
      "epoch": 5.922231317974505,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 25318
    },
    {
      "epoch": 5.922465208747515,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6678,
      "step": 25319
    },
    {
      "epoch": 5.922699099520524,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 25320
    },
    {
      "epoch": 5.922932990293533,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 25321
    },
    {
      "epoch": 5.923166881066542,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 25322
    },
    {
      "epoch": 5.923400771839551,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5243,
      "step": 25323
    },
    {
      "epoch": 5.92363466261256,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4219,
      "step": 25324
    },
    {
      "epoch": 5.923868553385569,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 25325
    },
    {
      "epoch": 5.924102444158578,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 25326
    },
    {
      "epoch": 5.924336334931587,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7202,
      "step": 25327
    },
    {
      "epoch": 5.924570225704596,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.366,
      "step": 25328
    },
    {
      "epoch": 5.924804116477605,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.844,
      "step": 25329
    },
    {
      "epoch": 5.925038007250614,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 25330
    },
    {
      "epoch": 5.925271898023623,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7205,
      "step": 25331
    },
    {
      "epoch": 5.925505788796632,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 25332
    },
    {
      "epoch": 5.925739679569641,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 25333
    },
    {
      "epoch": 5.92597357034265,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 25334
    },
    {
      "epoch": 5.926207461115659,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5574,
      "step": 25335
    },
    {
      "epoch": 5.926441351888668,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7831,
      "step": 25336
    },
    {
      "epoch": 5.926675242661677,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6714,
      "step": 25337
    },
    {
      "epoch": 5.926909133434686,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 25338
    },
    {
      "epoch": 5.927143024207695,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.528,
      "step": 25339
    },
    {
      "epoch": 5.927376914980704,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 25340
    },
    {
      "epoch": 5.927610805753713,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 25341
    },
    {
      "epoch": 5.927844696526722,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 25342
    },
    {
      "epoch": 5.928078587299731,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 25343
    },
    {
      "epoch": 5.92831247807274,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 25344
    },
    {
      "epoch": 5.928546368845749,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.563,
      "step": 25345
    },
    {
      "epoch": 5.928780259618758,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 25346
    },
    {
      "epoch": 5.929014150391767,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 25347
    },
    {
      "epoch": 5.929248041164776,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 25348
    },
    {
      "epoch": 5.929481931937785,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 25349
    },
    {
      "epoch": 5.929715822710794,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6241,
      "step": 25350
    },
    {
      "epoch": 5.929949713483803,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 25351
    },
    {
      "epoch": 5.930183604256812,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.4993,
      "step": 25352
    },
    {
      "epoch": 5.930417495029821,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7356,
      "step": 25353
    },
    {
      "epoch": 5.93065138580283,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4187,
      "step": 25354
    },
    {
      "epoch": 5.9308852765758395,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5721,
      "step": 25355
    },
    {
      "epoch": 5.931119167348848,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.6071,
      "step": 25356
    },
    {
      "epoch": 5.931353058121857,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 25357
    },
    {
      "epoch": 5.931586948894866,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.716,
      "step": 25358
    },
    {
      "epoch": 5.931820839667875,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5265,
      "step": 25359
    },
    {
      "epoch": 5.932054730440884,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 25360
    },
    {
      "epoch": 5.932288621213893,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 25361
    },
    {
      "epoch": 5.932522511986902,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6908,
      "step": 25362
    },
    {
      "epoch": 5.932756402759911,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 25363
    },
    {
      "epoch": 5.93299029353292,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 25364
    },
    {
      "epoch": 5.933224184305929,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 25365
    },
    {
      "epoch": 5.9334580750789385,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7052,
      "step": 25366
    },
    {
      "epoch": 5.933691965851947,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 25367
    },
    {
      "epoch": 5.933925856624956,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6475,
      "step": 25368
    },
    {
      "epoch": 5.934159747397965,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.729,
      "step": 25369
    },
    {
      "epoch": 5.934393638170974,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 25370
    },
    {
      "epoch": 5.934627528943983,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 25371
    },
    {
      "epoch": 5.934861419716992,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 25372
    },
    {
      "epoch": 5.935095310490001,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 25373
    },
    {
      "epoch": 5.93532920126301,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 25374
    },
    {
      "epoch": 5.935563092036019,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 25375
    },
    {
      "epoch": 5.935796982809029,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.96,
      "step": 25376
    },
    {
      "epoch": 5.9360308735820375,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 25377
    },
    {
      "epoch": 5.936264764355046,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5987,
      "step": 25378
    },
    {
      "epoch": 5.936498655128055,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 25379
    },
    {
      "epoch": 5.936732545901064,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 25380
    },
    {
      "epoch": 5.936966436674073,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 25381
    },
    {
      "epoch": 5.937200327447083,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 25382
    },
    {
      "epoch": 5.937434218220091,
      "grad_norm": 2.484375,
      "learning_rate": 3e-05,
      "loss": 1.4493,
      "step": 25383
    },
    {
      "epoch": 5.9376681089931,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 25384
    },
    {
      "epoch": 5.937901999766109,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 25385
    },
    {
      "epoch": 5.938135890539118,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5834,
      "step": 25386
    },
    {
      "epoch": 5.938369781312128,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 25387
    },
    {
      "epoch": 5.9386036720851365,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 25388
    },
    {
      "epoch": 5.938837562858145,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 25389
    },
    {
      "epoch": 5.939071453631154,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 25390
    },
    {
      "epoch": 5.939305344404163,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 25391
    },
    {
      "epoch": 5.939539235177172,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 25392
    },
    {
      "epoch": 5.939773125950182,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 25393
    },
    {
      "epoch": 5.94000701672319,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 25394
    },
    {
      "epoch": 5.940240907496199,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 25395
    },
    {
      "epoch": 5.940474798269208,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5439,
      "step": 25396
    },
    {
      "epoch": 5.940708689042217,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 25397
    },
    {
      "epoch": 5.940942579815227,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5851,
      "step": 25398
    },
    {
      "epoch": 5.9411764705882355,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 25399
    },
    {
      "epoch": 5.941410361361244,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.3545,
      "step": 25400
    },
    {
      "epoch": 5.941410361361244,
      "eval_runtime": 4.6041,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 25400
    },
    {
      "epoch": 5.941644252134253,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 25401
    },
    {
      "epoch": 5.941878142907262,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.5988,
      "step": 25402
    },
    {
      "epoch": 5.942112033680271,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 25403
    },
    {
      "epoch": 5.942345924453281,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 25404
    },
    {
      "epoch": 5.942579815226289,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.718,
      "step": 25405
    },
    {
      "epoch": 5.942813705999298,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6342,
      "step": 25406
    },
    {
      "epoch": 5.943047596772307,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.3668,
      "step": 25407
    },
    {
      "epoch": 5.943281487545317,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6768,
      "step": 25408
    },
    {
      "epoch": 5.943515378318326,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 25409
    },
    {
      "epoch": 5.9437492690913345,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6964,
      "step": 25410
    },
    {
      "epoch": 5.943983159864343,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5933,
      "step": 25411
    },
    {
      "epoch": 5.944217050637352,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.4388,
      "step": 25412
    },
    {
      "epoch": 5.944450941410361,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 25413
    },
    {
      "epoch": 5.944684832183371,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6594,
      "step": 25414
    },
    {
      "epoch": 5.94491872295638,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4299,
      "step": 25415
    },
    {
      "epoch": 5.945152613729388,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 25416
    },
    {
      "epoch": 5.945386504502397,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 25417
    },
    {
      "epoch": 5.945620395275406,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 25418
    },
    {
      "epoch": 5.945854286048416,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 25419
    },
    {
      "epoch": 5.946088176821425,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 25420
    },
    {
      "epoch": 5.9463220675944335,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7381,
      "step": 25421
    },
    {
      "epoch": 5.946555958367442,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 25422
    },
    {
      "epoch": 5.946789849140451,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 25423
    },
    {
      "epoch": 5.94702373991346,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 25424
    },
    {
      "epoch": 5.94725763068647,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 25425
    },
    {
      "epoch": 5.947491521459479,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 25426
    },
    {
      "epoch": 5.947725412232487,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 25427
    },
    {
      "epoch": 5.947959303005496,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0252,
      "step": 25428
    },
    {
      "epoch": 5.948193193778505,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 25429
    },
    {
      "epoch": 5.948427084551515,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 25430
    },
    {
      "epoch": 5.948660975324524,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 25431
    },
    {
      "epoch": 5.9488948660975325,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2435,
      "step": 25432
    },
    {
      "epoch": 5.949128756870541,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8535,
      "step": 25433
    },
    {
      "epoch": 5.94936264764355,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 25434
    },
    {
      "epoch": 5.949596538416559,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 25435
    },
    {
      "epoch": 5.949830429189569,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 25436
    },
    {
      "epoch": 5.950064319962578,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 25437
    },
    {
      "epoch": 5.950298210735586,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 25438
    },
    {
      "epoch": 5.950532101508595,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 25439
    },
    {
      "epoch": 5.950765992281605,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 25440
    },
    {
      "epoch": 5.950999883054614,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 25441
    },
    {
      "epoch": 5.951233773827623,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 25442
    },
    {
      "epoch": 5.9514676646006315,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 25443
    },
    {
      "epoch": 5.95170155537364,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6726,
      "step": 25444
    },
    {
      "epoch": 5.951935446146649,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 25445
    },
    {
      "epoch": 5.952169336919659,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8528,
      "step": 25446
    },
    {
      "epoch": 5.952403227692668,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 25447
    },
    {
      "epoch": 5.952637118465677,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 25448
    },
    {
      "epoch": 5.952871009238685,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 25449
    },
    {
      "epoch": 5.953104900011694,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 25450
    },
    {
      "epoch": 5.953338790784704,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 25451
    },
    {
      "epoch": 5.953572681557713,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8178,
      "step": 25452
    },
    {
      "epoch": 5.953806572330722,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 25453
    },
    {
      "epoch": 5.9540404631037305,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 25454
    },
    {
      "epoch": 5.954274353876739,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 25455
    },
    {
      "epoch": 5.954508244649748,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 25456
    },
    {
      "epoch": 5.954742135422758,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5403,
      "step": 25457
    },
    {
      "epoch": 5.954976026195767,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 25458
    },
    {
      "epoch": 5.9552099169687756,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7011,
      "step": 25459
    },
    {
      "epoch": 5.955443807741784,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 25460
    },
    {
      "epoch": 5.955677698514793,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 25461
    },
    {
      "epoch": 5.955911589287803,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 25462
    },
    {
      "epoch": 5.956145480060812,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7595,
      "step": 25463
    },
    {
      "epoch": 5.956379370833821,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 25464
    },
    {
      "epoch": 5.9566132616068295,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4747,
      "step": 25465
    },
    {
      "epoch": 5.956847152379838,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7297,
      "step": 25466
    },
    {
      "epoch": 5.957081043152847,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 25467
    },
    {
      "epoch": 5.957314933925857,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 25468
    },
    {
      "epoch": 5.957548824698866,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6865,
      "step": 25469
    },
    {
      "epoch": 5.9577827154718745,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.686,
      "step": 25470
    },
    {
      "epoch": 5.958016606244883,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 25471
    },
    {
      "epoch": 5.958250497017893,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7074,
      "step": 25472
    },
    {
      "epoch": 5.958484387790902,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6413,
      "step": 25473
    },
    {
      "epoch": 5.958718278563911,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 25474
    },
    {
      "epoch": 5.95895216933692,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6512,
      "step": 25475
    },
    {
      "epoch": 5.9591860601099285,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 25476
    },
    {
      "epoch": 5.959419950882937,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 25477
    },
    {
      "epoch": 5.959653841655947,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6235,
      "step": 25478
    },
    {
      "epoch": 5.959887732428956,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 25479
    },
    {
      "epoch": 5.960121623201965,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 25480
    },
    {
      "epoch": 5.9603555139749735,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 25481
    },
    {
      "epoch": 5.960589404747982,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 25482
    },
    {
      "epoch": 5.960823295520992,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8165,
      "step": 25483
    },
    {
      "epoch": 5.961057186294001,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 25484
    },
    {
      "epoch": 5.96129107706701,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 25485
    },
    {
      "epoch": 5.961524967840019,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 25486
    },
    {
      "epoch": 5.9617588586130275,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 25487
    },
    {
      "epoch": 5.961992749386036,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 25488
    },
    {
      "epoch": 5.962226640159046,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 25489
    },
    {
      "epoch": 5.962460530932055,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 25490
    },
    {
      "epoch": 5.962694421705064,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9184,
      "step": 25491
    },
    {
      "epoch": 5.9629283124780725,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 25492
    },
    {
      "epoch": 5.963162203251081,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6759,
      "step": 25493
    },
    {
      "epoch": 5.963396094024091,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 25494
    },
    {
      "epoch": 5.9636299847971,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6641,
      "step": 25495
    },
    {
      "epoch": 5.963863875570109,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 25496
    },
    {
      "epoch": 5.964097766343118,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 25497
    },
    {
      "epoch": 5.9643316571161265,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 25498
    },
    {
      "epoch": 5.964565547889135,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 25499
    },
    {
      "epoch": 5.964799438662145,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 25500
    },
    {
      "epoch": 5.964799438662145,
      "eval_runtime": 4.6336,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 25500
    },
    {
      "epoch": 5.965033329435154,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 25501
    },
    {
      "epoch": 5.965267220208163,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 25502
    },
    {
      "epoch": 5.9655011109811715,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 25503
    },
    {
      "epoch": 5.965735001754181,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7479,
      "step": 25504
    },
    {
      "epoch": 5.96596889252719,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 25505
    },
    {
      "epoch": 5.966202783300199,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 25506
    },
    {
      "epoch": 5.966436674073208,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 25507
    },
    {
      "epoch": 5.966670564846217,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 25508
    },
    {
      "epoch": 5.9669044556192254,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 25509
    },
    {
      "epoch": 5.967138346392235,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 25510
    },
    {
      "epoch": 5.967372237165244,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7061,
      "step": 25511
    },
    {
      "epoch": 5.967606127938253,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 25512
    },
    {
      "epoch": 5.967840018711262,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6596,
      "step": 25513
    },
    {
      "epoch": 5.9680739094842705,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 25514
    },
    {
      "epoch": 5.96830780025728,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.681,
      "step": 25515
    },
    {
      "epoch": 5.968541691030289,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6008,
      "step": 25516
    },
    {
      "epoch": 5.968775581803298,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 25517
    },
    {
      "epoch": 5.969009472576307,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 25518
    },
    {
      "epoch": 5.969243363349316,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 25519
    },
    {
      "epoch": 5.969477254122324,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 25520
    },
    {
      "epoch": 5.969711144895334,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5805,
      "step": 25521
    },
    {
      "epoch": 5.969945035668343,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 25522
    },
    {
      "epoch": 5.970178926441352,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 25523
    },
    {
      "epoch": 5.970412817214361,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5887,
      "step": 25524
    },
    {
      "epoch": 5.9706467079873695,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 25525
    },
    {
      "epoch": 5.970880598760379,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 25526
    },
    {
      "epoch": 5.971114489533388,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6021,
      "step": 25527
    },
    {
      "epoch": 5.971348380306397,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 25528
    },
    {
      "epoch": 5.971582271079406,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 25529
    },
    {
      "epoch": 5.971816161852415,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 25530
    },
    {
      "epoch": 5.972050052625423,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 25531
    },
    {
      "epoch": 5.972283943398433,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8837,
      "step": 25532
    },
    {
      "epoch": 5.972517834171442,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 25533
    },
    {
      "epoch": 5.972751724944451,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 25534
    },
    {
      "epoch": 5.97298561571746,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 25535
    },
    {
      "epoch": 5.973219506490469,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 25536
    },
    {
      "epoch": 5.973453397263478,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 25537
    },
    {
      "epoch": 5.973687288036487,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 25538
    },
    {
      "epoch": 5.973921178809496,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 25539
    },
    {
      "epoch": 5.974155069582505,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 25540
    },
    {
      "epoch": 5.974388960355514,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6755,
      "step": 25541
    },
    {
      "epoch": 5.974622851128523,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 25542
    },
    {
      "epoch": 5.974856741901532,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6552,
      "step": 25543
    },
    {
      "epoch": 5.975090632674541,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 25544
    },
    {
      "epoch": 5.97532452344755,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 25545
    },
    {
      "epoch": 5.975558414220559,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 25546
    },
    {
      "epoch": 5.975792304993568,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 25547
    },
    {
      "epoch": 5.976026195766577,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6548,
      "step": 25548
    },
    {
      "epoch": 5.976260086539586,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 25549
    },
    {
      "epoch": 5.976493977312595,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 25550
    },
    {
      "epoch": 5.976727868085604,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 25551
    },
    {
      "epoch": 5.976961758858613,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7814,
      "step": 25552
    },
    {
      "epoch": 5.977195649631622,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 25553
    },
    {
      "epoch": 5.977429540404631,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7543,
      "step": 25554
    },
    {
      "epoch": 5.97766343117764,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 25555
    },
    {
      "epoch": 5.977897321950649,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 25556
    },
    {
      "epoch": 5.9781312127236585,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 25557
    },
    {
      "epoch": 5.978365103496667,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 25558
    },
    {
      "epoch": 5.978598994269676,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5694,
      "step": 25559
    },
    {
      "epoch": 5.978832885042685,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 25560
    },
    {
      "epoch": 5.979066775815694,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 25561
    },
    {
      "epoch": 5.979300666588703,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 25562
    },
    {
      "epoch": 5.9795345573617125,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7256,
      "step": 25563
    },
    {
      "epoch": 5.979768448134721,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.679,
      "step": 25564
    },
    {
      "epoch": 5.98000233890773,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 25565
    },
    {
      "epoch": 5.980236229680739,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7267,
      "step": 25566
    },
    {
      "epoch": 5.980470120453748,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6553,
      "step": 25567
    },
    {
      "epoch": 5.9807040112267575,
      "grad_norm": 11.5625,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 25568
    },
    {
      "epoch": 5.980937901999766,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 25569
    },
    {
      "epoch": 5.981171792772775,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 25570
    },
    {
      "epoch": 5.981405683545784,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 25571
    },
    {
      "epoch": 5.981639574318793,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 25572
    },
    {
      "epoch": 5.981873465091802,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 25573
    },
    {
      "epoch": 5.9821073558648115,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 25574
    },
    {
      "epoch": 5.98234124663782,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 25575
    },
    {
      "epoch": 5.982575137410829,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 25576
    },
    {
      "epoch": 5.982809028183838,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 25577
    },
    {
      "epoch": 5.983042918956847,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 25578
    },
    {
      "epoch": 5.9832768097298565,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 25579
    },
    {
      "epoch": 5.983510700502865,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 25580
    },
    {
      "epoch": 5.983744591275874,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8308,
      "step": 25581
    },
    {
      "epoch": 5.983978482048883,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 25582
    },
    {
      "epoch": 5.984212372821892,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 25583
    },
    {
      "epoch": 5.984446263594901,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7381,
      "step": 25584
    },
    {
      "epoch": 5.9846801543679105,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 25585
    },
    {
      "epoch": 5.984914045140919,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 25586
    },
    {
      "epoch": 5.985147935913928,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 25587
    },
    {
      "epoch": 5.985381826686937,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 25588
    },
    {
      "epoch": 5.985615717459947,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 25589
    },
    {
      "epoch": 5.9858496082329555,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 25590
    },
    {
      "epoch": 5.986083499005964,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 25591
    },
    {
      "epoch": 5.986317389778973,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7356,
      "step": 25592
    },
    {
      "epoch": 5.986551280551982,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 25593
    },
    {
      "epoch": 5.986785171324991,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 25594
    },
    {
      "epoch": 5.987019062098001,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 25595
    },
    {
      "epoch": 5.9872529528710094,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5921,
      "step": 25596
    },
    {
      "epoch": 5.987486843644018,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6596,
      "step": 25597
    },
    {
      "epoch": 5.987720734417027,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7049,
      "step": 25598
    },
    {
      "epoch": 5.987954625190036,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5163,
      "step": 25599
    },
    {
      "epoch": 5.988188515963046,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 25600
    },
    {
      "epoch": 5.988188515963046,
      "eval_runtime": 4.6214,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 25600
    },
    {
      "epoch": 5.9884224067360545,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9642,
      "step": 25601
    },
    {
      "epoch": 5.988656297509063,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7876,
      "step": 25602
    },
    {
      "epoch": 5.988890188282072,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 25603
    },
    {
      "epoch": 5.989124079055081,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 25604
    },
    {
      "epoch": 5.98935796982809,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 25605
    },
    {
      "epoch": 5.9895918606011,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 25606
    },
    {
      "epoch": 5.989825751374108,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 25607
    },
    {
      "epoch": 5.990059642147117,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 25608
    },
    {
      "epoch": 5.990293532920126,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5695,
      "step": 25609
    },
    {
      "epoch": 5.990527423693135,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 25610
    },
    {
      "epoch": 5.990761314466145,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 25611
    },
    {
      "epoch": 5.9909952052391535,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 25612
    },
    {
      "epoch": 5.991229096012162,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 25613
    },
    {
      "epoch": 5.991462986785171,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 25614
    },
    {
      "epoch": 5.99169687755818,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 25615
    },
    {
      "epoch": 5.991930768331189,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 25616
    },
    {
      "epoch": 5.992164659104199,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7914,
      "step": 25617
    },
    {
      "epoch": 5.992398549877207,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 25618
    },
    {
      "epoch": 5.992632440650216,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 25619
    },
    {
      "epoch": 5.992866331423225,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 25620
    },
    {
      "epoch": 5.993100222196235,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 25621
    },
    {
      "epoch": 5.993334112969244,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 25622
    },
    {
      "epoch": 5.9935680037422525,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 25623
    },
    {
      "epoch": 5.993801894515261,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 25624
    },
    {
      "epoch": 5.99403578528827,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7617,
      "step": 25625
    },
    {
      "epoch": 5.994269676061279,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 25626
    },
    {
      "epoch": 5.994503566834289,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 25627
    },
    {
      "epoch": 5.994737457607298,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5846,
      "step": 25628
    },
    {
      "epoch": 5.994971348380306,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 25629
    },
    {
      "epoch": 5.995205239153315,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 25630
    },
    {
      "epoch": 5.995439129926324,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 25631
    },
    {
      "epoch": 5.995673020699334,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6462,
      "step": 25632
    },
    {
      "epoch": 5.995906911472343,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 25633
    },
    {
      "epoch": 5.9961408022453515,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 25634
    },
    {
      "epoch": 5.99637469301836,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 25635
    },
    {
      "epoch": 5.996608583791369,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 25636
    },
    {
      "epoch": 5.996842474564378,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 25637
    },
    {
      "epoch": 5.997076365337388,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 25638
    },
    {
      "epoch": 5.997310256110397,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 25639
    },
    {
      "epoch": 5.997544146883405,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 25640
    },
    {
      "epoch": 5.997778037656414,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 25641
    },
    {
      "epoch": 5.998011928429423,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 25642
    },
    {
      "epoch": 5.998245819202433,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 25643
    },
    {
      "epoch": 5.998479709975442,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.677,
      "step": 25644
    },
    {
      "epoch": 5.9987136007484505,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 25645
    },
    {
      "epoch": 5.998947491521459,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5854,
      "step": 25646
    },
    {
      "epoch": 5.999181382294468,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8366,
      "step": 25647
    },
    {
      "epoch": 5.999415273067477,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9255,
      "step": 25648
    },
    {
      "epoch": 5.999649163840487,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 25649
    },
    {
      "epoch": 5.999883054613496,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 25650
    },
    {
      "epoch": 6.000233890773009,
      "grad_norm": 10.125,
      "learning_rate": 3e-05,
      "loss": 3.507,
      "step": 25651
    },
    {
      "epoch": 6.000467781546018,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5577,
      "step": 25652
    },
    {
      "epoch": 6.000701672319027,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 25653
    },
    {
      "epoch": 6.000935563092036,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6804,
      "step": 25654
    },
    {
      "epoch": 6.001169453865045,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6566,
      "step": 25655
    },
    {
      "epoch": 6.001403344638054,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5993,
      "step": 25656
    },
    {
      "epoch": 6.001637235411063,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.729,
      "step": 25657
    },
    {
      "epoch": 6.0018711261840725,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 25658
    },
    {
      "epoch": 6.002105016957081,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 25659
    },
    {
      "epoch": 6.00233890773009,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 25660
    },
    {
      "epoch": 6.002572798503099,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 25661
    },
    {
      "epoch": 6.002806689276108,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 25662
    },
    {
      "epoch": 6.003040580049117,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 25663
    },
    {
      "epoch": 6.003274470822126,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 25664
    },
    {
      "epoch": 6.003508361595135,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 25665
    },
    {
      "epoch": 6.003742252368144,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 25666
    },
    {
      "epoch": 6.003976143141153,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 25667
    },
    {
      "epoch": 6.004210033914162,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6689,
      "step": 25668
    },
    {
      "epoch": 6.0044439246871715,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8156,
      "step": 25669
    },
    {
      "epoch": 6.00467781546018,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7204,
      "step": 25670
    },
    {
      "epoch": 6.004911706233189,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 25671
    },
    {
      "epoch": 6.005145597006198,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 25672
    },
    {
      "epoch": 6.005379487779207,
      "grad_norm": 7.75,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 25673
    },
    {
      "epoch": 6.0056133785522166,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 25674
    },
    {
      "epoch": 6.005847269325225,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 25675
    },
    {
      "epoch": 6.006081160098234,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7617,
      "step": 25676
    },
    {
      "epoch": 6.006315050871243,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 25677
    },
    {
      "epoch": 6.006548941644252,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 25678
    },
    {
      "epoch": 6.006782832417261,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 25679
    },
    {
      "epoch": 6.0070167231902705,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6679,
      "step": 25680
    },
    {
      "epoch": 6.007250613963279,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 25681
    },
    {
      "epoch": 6.007484504736288,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 25682
    },
    {
      "epoch": 6.007718395509297,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 25683
    },
    {
      "epoch": 6.007952286282306,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.538,
      "step": 25684
    },
    {
      "epoch": 6.0081861770553155,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 25685
    },
    {
      "epoch": 6.008420067828324,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4243,
      "step": 25686
    },
    {
      "epoch": 6.008653958601333,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 25687
    },
    {
      "epoch": 6.008887849374342,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 25688
    },
    {
      "epoch": 6.009121740147351,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 25689
    },
    {
      "epoch": 6.009355630920361,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5655,
      "step": 25690
    },
    {
      "epoch": 6.0095895216933695,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 25691
    },
    {
      "epoch": 6.009823412466378,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 25692
    },
    {
      "epoch": 6.010057303239387,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.4844,
      "step": 25693
    },
    {
      "epoch": 6.010291194012396,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 25694
    },
    {
      "epoch": 6.010525084785405,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5838,
      "step": 25695
    },
    {
      "epoch": 6.0107589755584145,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 25696
    },
    {
      "epoch": 6.010992866331423,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 25697
    },
    {
      "epoch": 6.011226757104432,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5406,
      "step": 25698
    },
    {
      "epoch": 6.011460647877441,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6913,
      "step": 25699
    },
    {
      "epoch": 6.01169453865045,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 25700
    },
    {
      "epoch": 6.01169453865045,
      "eval_runtime": 4.6277,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 25700
    },
    {
      "epoch": 6.01192842942346,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6673,
      "step": 25701
    },
    {
      "epoch": 6.0121623201964685,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.6636,
      "step": 25702
    },
    {
      "epoch": 6.012396210969477,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 25703
    },
    {
      "epoch": 6.012630101742486,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 25704
    },
    {
      "epoch": 6.012863992515495,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 25705
    },
    {
      "epoch": 6.013097883288505,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 25706
    },
    {
      "epoch": 6.0133317740615135,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 25707
    },
    {
      "epoch": 6.013565664834522,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 25708
    },
    {
      "epoch": 6.013799555607531,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8366,
      "step": 25709
    },
    {
      "epoch": 6.01403344638054,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 25710
    },
    {
      "epoch": 6.014267337153549,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 25711
    },
    {
      "epoch": 6.014501227926559,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 25712
    },
    {
      "epoch": 6.0147351186995675,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 25713
    },
    {
      "epoch": 6.014969009472576,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 25714
    },
    {
      "epoch": 6.015202900245585,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5886,
      "step": 25715
    },
    {
      "epoch": 6.015436791018594,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 25716
    },
    {
      "epoch": 6.015670681791604,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.661,
      "step": 25717
    },
    {
      "epoch": 6.0159045725646125,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 25718
    },
    {
      "epoch": 6.016138463337621,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 25719
    },
    {
      "epoch": 6.01637235411063,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7858,
      "step": 25720
    },
    {
      "epoch": 6.016606244883639,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 25721
    },
    {
      "epoch": 6.016840135656649,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 25722
    },
    {
      "epoch": 6.017074026429658,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 25723
    },
    {
      "epoch": 6.0173079172026664,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 25724
    },
    {
      "epoch": 6.017541807975675,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7624,
      "step": 25725
    },
    {
      "epoch": 6.017775698748684,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6667,
      "step": 25726
    },
    {
      "epoch": 6.018009589521693,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 25727
    },
    {
      "epoch": 6.018243480294703,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 25728
    },
    {
      "epoch": 6.0184773710677115,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 25729
    },
    {
      "epoch": 6.01871126184072,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 25730
    },
    {
      "epoch": 6.018945152613729,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4778,
      "step": 25731
    },
    {
      "epoch": 6.019179043386738,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5866,
      "step": 25732
    },
    {
      "epoch": 6.019412934159748,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4925,
      "step": 25733
    },
    {
      "epoch": 6.019646824932757,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 25734
    },
    {
      "epoch": 6.019880715705765,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 25735
    },
    {
      "epoch": 6.020114606478774,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 25736
    },
    {
      "epoch": 6.020348497251783,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 25737
    },
    {
      "epoch": 6.020582388024793,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 25738
    },
    {
      "epoch": 6.020816278797802,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 25739
    },
    {
      "epoch": 6.0210501695708105,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 25740
    },
    {
      "epoch": 6.021284060343819,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 25741
    },
    {
      "epoch": 6.021517951116828,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7474,
      "step": 25742
    },
    {
      "epoch": 6.021751841889837,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5226,
      "step": 25743
    },
    {
      "epoch": 6.021985732662847,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6213,
      "step": 25744
    },
    {
      "epoch": 6.022219623435856,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 25745
    },
    {
      "epoch": 6.022453514208864,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 25746
    },
    {
      "epoch": 6.022687404981873,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 25747
    },
    {
      "epoch": 6.022921295754882,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6106,
      "step": 25748
    },
    {
      "epoch": 6.023155186527892,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5028,
      "step": 25749
    },
    {
      "epoch": 6.023389077300901,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6938,
      "step": 25750
    },
    {
      "epoch": 6.0236229680739095,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 25751
    },
    {
      "epoch": 6.023856858846918,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 25752
    },
    {
      "epoch": 6.024090749619927,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 25753
    },
    {
      "epoch": 6.024324640392937,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 25754
    },
    {
      "epoch": 6.024558531165946,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 25755
    },
    {
      "epoch": 6.024792421938955,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 25756
    },
    {
      "epoch": 6.025026312711963,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5793,
      "step": 25757
    },
    {
      "epoch": 6.025260203484972,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 25758
    },
    {
      "epoch": 6.025494094257981,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5801,
      "step": 25759
    },
    {
      "epoch": 6.025727985030991,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 25760
    },
    {
      "epoch": 6.025961875804,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8047,
      "step": 25761
    },
    {
      "epoch": 6.0261957665770085,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5018,
      "step": 25762
    },
    {
      "epoch": 6.026429657350017,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.857,
      "step": 25763
    },
    {
      "epoch": 6.026663548123026,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 25764
    },
    {
      "epoch": 6.026897438896036,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5658,
      "step": 25765
    },
    {
      "epoch": 6.027131329669045,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 25766
    },
    {
      "epoch": 6.027365220442054,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 25767
    },
    {
      "epoch": 6.027599111215062,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 25768
    },
    {
      "epoch": 6.027833001988071,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 25769
    },
    {
      "epoch": 6.028066892761081,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5568,
      "step": 25770
    },
    {
      "epoch": 6.02830078353409,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 25771
    },
    {
      "epoch": 6.028534674307099,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 25772
    },
    {
      "epoch": 6.0287685650801075,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 25773
    },
    {
      "epoch": 6.029002455853116,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 25774
    },
    {
      "epoch": 6.029236346626125,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 25775
    },
    {
      "epoch": 6.029470237399135,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5587,
      "step": 25776
    },
    {
      "epoch": 6.029704128172144,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7277,
      "step": 25777
    },
    {
      "epoch": 6.029938018945153,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 25778
    },
    {
      "epoch": 6.030171909718161,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.4435,
      "step": 25779
    },
    {
      "epoch": 6.03040580049117,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5319,
      "step": 25780
    },
    {
      "epoch": 6.03063969126418,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 25781
    },
    {
      "epoch": 6.030873582037189,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 25782
    },
    {
      "epoch": 6.031107472810198,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 25783
    },
    {
      "epoch": 6.0313413635832065,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8257,
      "step": 25784
    },
    {
      "epoch": 6.031575254356215,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7366,
      "step": 25785
    },
    {
      "epoch": 6.031809145129225,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 25786
    },
    {
      "epoch": 6.032043035902234,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.762,
      "step": 25787
    },
    {
      "epoch": 6.032276926675243,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.656,
      "step": 25788
    },
    {
      "epoch": 6.032510817448252,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4061,
      "step": 25789
    },
    {
      "epoch": 6.03274470822126,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 25790
    },
    {
      "epoch": 6.032978598994269,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 25791
    },
    {
      "epoch": 6.033212489767279,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 25792
    },
    {
      "epoch": 6.033446380540288,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 25793
    },
    {
      "epoch": 6.033680271313297,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 25794
    },
    {
      "epoch": 6.0339141620863055,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 25795
    },
    {
      "epoch": 6.034148052859314,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 25796
    },
    {
      "epoch": 6.034381943632324,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 25797
    },
    {
      "epoch": 6.034615834405333,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 25798
    },
    {
      "epoch": 6.034849725178342,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 25799
    },
    {
      "epoch": 6.035083615951351,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 25800
    },
    {
      "epoch": 6.035083615951351,
      "eval_runtime": 4.65,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 25800
    },
    {
      "epoch": 6.035317506724359,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5839,
      "step": 25801
    },
    {
      "epoch": 6.035551397497369,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 25802
    },
    {
      "epoch": 6.035785288270378,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 25803
    },
    {
      "epoch": 6.036019179043387,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3173,
      "step": 25804
    },
    {
      "epoch": 6.036253069816396,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7008,
      "step": 25805
    },
    {
      "epoch": 6.0364869605894045,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6086,
      "step": 25806
    },
    {
      "epoch": 6.036720851362413,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8473,
      "step": 25807
    },
    {
      "epoch": 6.036954742135423,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 25808
    },
    {
      "epoch": 6.037188632908432,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 25809
    },
    {
      "epoch": 6.037422523681441,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6608,
      "step": 25810
    },
    {
      "epoch": 6.03765641445445,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7883,
      "step": 25811
    },
    {
      "epoch": 6.037890305227458,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 25812
    },
    {
      "epoch": 6.038124196000468,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6938,
      "step": 25813
    },
    {
      "epoch": 6.038358086773477,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5366,
      "step": 25814
    },
    {
      "epoch": 6.038591977546486,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6246,
      "step": 25815
    },
    {
      "epoch": 6.038825868319495,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.3367,
      "step": 25816
    },
    {
      "epoch": 6.0390597590925035,
      "grad_norm": 7.40625,
      "learning_rate": 3e-05,
      "loss": 2.3465,
      "step": 25817
    },
    {
      "epoch": 6.039293649865513,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 25818
    },
    {
      "epoch": 6.039527540638522,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 25819
    },
    {
      "epoch": 6.039761431411531,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 25820
    },
    {
      "epoch": 6.03999532218454,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6021,
      "step": 25821
    },
    {
      "epoch": 6.0402292129575486,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7769,
      "step": 25822
    },
    {
      "epoch": 6.040463103730557,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 25823
    },
    {
      "epoch": 6.040696994503567,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.759,
      "step": 25824
    },
    {
      "epoch": 6.040930885276576,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 25825
    },
    {
      "epoch": 6.041164776049585,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7056,
      "step": 25826
    },
    {
      "epoch": 6.041398666822594,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 25827
    },
    {
      "epoch": 6.0416325575956025,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 25828
    },
    {
      "epoch": 6.041866448368612,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 25829
    },
    {
      "epoch": 6.042100339141621,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 25830
    },
    {
      "epoch": 6.04233422991463,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8684,
      "step": 25831
    },
    {
      "epoch": 6.042568120687639,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6334,
      "step": 25832
    },
    {
      "epoch": 6.0428020114606475,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 25833
    },
    {
      "epoch": 6.043035902233657,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.588,
      "step": 25834
    },
    {
      "epoch": 6.043269793006666,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 25835
    },
    {
      "epoch": 6.043503683779675,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 25836
    },
    {
      "epoch": 6.043737574552684,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 25837
    },
    {
      "epoch": 6.043971465325693,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 25838
    },
    {
      "epoch": 6.044205356098702,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.96,
      "step": 25839
    },
    {
      "epoch": 6.044439246871711,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.677,
      "step": 25840
    },
    {
      "epoch": 6.04467313764472,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 25841
    },
    {
      "epoch": 6.044907028417729,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 25842
    },
    {
      "epoch": 6.045140919190738,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6817,
      "step": 25843
    },
    {
      "epoch": 6.0453748099637465,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 25844
    },
    {
      "epoch": 6.045608700736756,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 25845
    },
    {
      "epoch": 6.045842591509765,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6296,
      "step": 25846
    },
    {
      "epoch": 6.046076482282774,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6487,
      "step": 25847
    },
    {
      "epoch": 6.046310373055783,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 25848
    },
    {
      "epoch": 6.046544263828792,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 25849
    },
    {
      "epoch": 6.046778154601801,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7449,
      "step": 25850
    },
    {
      "epoch": 6.04701204537481,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 25851
    },
    {
      "epoch": 6.047245936147819,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 25852
    },
    {
      "epoch": 6.047479826920828,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 25853
    },
    {
      "epoch": 6.047713717693837,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 25854
    },
    {
      "epoch": 6.0479476084668455,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 25855
    },
    {
      "epoch": 6.048181499239855,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7687,
      "step": 25856
    },
    {
      "epoch": 6.048415390012864,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 25857
    },
    {
      "epoch": 6.048649280785873,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6249,
      "step": 25858
    },
    {
      "epoch": 6.048883171558882,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.415,
      "step": 25859
    },
    {
      "epoch": 6.049117062331891,
      "grad_norm": 12.125,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 25860
    },
    {
      "epoch": 6.0493509531049,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7714,
      "step": 25861
    },
    {
      "epoch": 6.049584843877909,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 25862
    },
    {
      "epoch": 6.049818734650918,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7694,
      "step": 25863
    },
    {
      "epoch": 6.050052625423927,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6729,
      "step": 25864
    },
    {
      "epoch": 6.050286516196936,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6975,
      "step": 25865
    },
    {
      "epoch": 6.050520406969945,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 25866
    },
    {
      "epoch": 6.050754297742954,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7054,
      "step": 25867
    },
    {
      "epoch": 6.050988188515963,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.4949,
      "step": 25868
    },
    {
      "epoch": 6.051222079288972,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5666,
      "step": 25869
    },
    {
      "epoch": 6.051455970061981,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 25870
    },
    {
      "epoch": 6.0516898608349905,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7282,
      "step": 25871
    },
    {
      "epoch": 6.051923751607999,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5218,
      "step": 25872
    },
    {
      "epoch": 6.052157642381008,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 25873
    },
    {
      "epoch": 6.052391533154017,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5686,
      "step": 25874
    },
    {
      "epoch": 6.052625423927026,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 25875
    },
    {
      "epoch": 6.052859314700035,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.625,
      "step": 25876
    },
    {
      "epoch": 6.053093205473044,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 25877
    },
    {
      "epoch": 6.053327096246053,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 25878
    },
    {
      "epoch": 6.053560987019062,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 25879
    },
    {
      "epoch": 6.053794877792071,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 25880
    },
    {
      "epoch": 6.05402876856508,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 25881
    },
    {
      "epoch": 6.0542626593380895,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 25882
    },
    {
      "epoch": 6.054496550111098,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4519,
      "step": 25883
    },
    {
      "epoch": 6.054730440884107,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 25884
    },
    {
      "epoch": 6.054964331657116,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 25885
    },
    {
      "epoch": 6.055198222430125,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.5279,
      "step": 25886
    },
    {
      "epoch": 6.055432113203135,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 25887
    },
    {
      "epoch": 6.055666003976143,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6051,
      "step": 25888
    },
    {
      "epoch": 6.055899894749152,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.4489,
      "step": 25889
    },
    {
      "epoch": 6.056133785522161,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 25890
    },
    {
      "epoch": 6.05636767629517,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5727,
      "step": 25891
    },
    {
      "epoch": 6.056601567068179,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 25892
    },
    {
      "epoch": 6.0568354578411885,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.4559,
      "step": 25893
    },
    {
      "epoch": 6.057069348614197,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 25894
    },
    {
      "epoch": 6.057303239387206,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6005,
      "step": 25895
    },
    {
      "epoch": 6.057537130160215,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 25896
    },
    {
      "epoch": 6.057771020933224,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 25897
    },
    {
      "epoch": 6.058004911706234,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 25898
    },
    {
      "epoch": 6.058238802479242,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7481,
      "step": 25899
    },
    {
      "epoch": 6.058472693252251,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 25900
    },
    {
      "epoch": 6.058472693252251,
      "eval_runtime": 4.6079,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 25900
    },
    {
      "epoch": 6.05870658402526,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 25901
    },
    {
      "epoch": 6.058940474798269,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 25902
    },
    {
      "epoch": 6.059174365571279,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 25903
    },
    {
      "epoch": 6.0594082563442875,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6889,
      "step": 25904
    },
    {
      "epoch": 6.059642147117296,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 25905
    },
    {
      "epoch": 6.059876037890305,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 25906
    },
    {
      "epoch": 6.060109928663314,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6656,
      "step": 25907
    },
    {
      "epoch": 6.060343819436323,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8693,
      "step": 25908
    },
    {
      "epoch": 6.060577710209333,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 25909
    },
    {
      "epoch": 6.060811600982341,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 25910
    },
    {
      "epoch": 6.06104549175535,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1511,
      "step": 25911
    },
    {
      "epoch": 6.061279382528359,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 25912
    },
    {
      "epoch": 6.061513273301368,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 25913
    },
    {
      "epoch": 6.061747164074378,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6429,
      "step": 25914
    },
    {
      "epoch": 6.0619810548473865,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5605,
      "step": 25915
    },
    {
      "epoch": 6.062214945620395,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5966,
      "step": 25916
    },
    {
      "epoch": 6.062448836393404,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 25917
    },
    {
      "epoch": 6.062682727166413,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7296,
      "step": 25918
    },
    {
      "epoch": 6.062916617939423,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 25919
    },
    {
      "epoch": 6.0631505087124316,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 25920
    },
    {
      "epoch": 6.06338439948544,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.637,
      "step": 25921
    },
    {
      "epoch": 6.063618290258449,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 25922
    },
    {
      "epoch": 6.063852181031458,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6033,
      "step": 25923
    },
    {
      "epoch": 6.064086071804467,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 25924
    },
    {
      "epoch": 6.064319962577477,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6783,
      "step": 25925
    },
    {
      "epoch": 6.0645538533504855,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.696,
      "step": 25926
    },
    {
      "epoch": 6.064787744123494,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7022,
      "step": 25927
    },
    {
      "epoch": 6.065021634896503,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6763,
      "step": 25928
    },
    {
      "epoch": 6.065255525669512,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 25929
    },
    {
      "epoch": 6.065489416442522,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.525,
      "step": 25930
    },
    {
      "epoch": 6.0657233072155305,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 25931
    },
    {
      "epoch": 6.065957197988539,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 25932
    },
    {
      "epoch": 6.066191088761548,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 25933
    },
    {
      "epoch": 6.066424979534557,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 25934
    },
    {
      "epoch": 6.066658870307567,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 25935
    },
    {
      "epoch": 6.066892761080576,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 25936
    },
    {
      "epoch": 6.0671266518535845,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 25937
    },
    {
      "epoch": 6.067360542626593,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 25938
    },
    {
      "epoch": 6.067594433399602,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.716,
      "step": 25939
    },
    {
      "epoch": 6.067828324172611,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 25940
    },
    {
      "epoch": 6.068062214945621,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 25941
    },
    {
      "epoch": 6.0682961057186295,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 25942
    },
    {
      "epoch": 6.068529996491638,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 25943
    },
    {
      "epoch": 6.068763887264647,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 25944
    },
    {
      "epoch": 6.068997778037656,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6578,
      "step": 25945
    },
    {
      "epoch": 6.069231668810666,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 25946
    },
    {
      "epoch": 6.069465559583675,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 25947
    },
    {
      "epoch": 6.0696994503566835,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 25948
    },
    {
      "epoch": 6.069933341129692,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 25949
    },
    {
      "epoch": 6.070167231902701,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 25950
    },
    {
      "epoch": 6.070401122675711,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5583,
      "step": 25951
    },
    {
      "epoch": 6.07063501344872,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6797,
      "step": 25952
    },
    {
      "epoch": 6.0708689042217285,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7666,
      "step": 25953
    },
    {
      "epoch": 6.071102794994737,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 25954
    },
    {
      "epoch": 6.071336685767746,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 25955
    },
    {
      "epoch": 6.071570576540755,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 25956
    },
    {
      "epoch": 6.071804467313765,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7608,
      "step": 25957
    },
    {
      "epoch": 6.072038358086774,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 25958
    },
    {
      "epoch": 6.0722722488597825,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6663,
      "step": 25959
    },
    {
      "epoch": 6.072506139632791,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 25960
    },
    {
      "epoch": 6.0727400304058,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 25961
    },
    {
      "epoch": 6.07297392117881,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 25962
    },
    {
      "epoch": 6.073207811951819,
      "grad_norm": 7.75,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 25963
    },
    {
      "epoch": 6.0734417027248275,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 25964
    },
    {
      "epoch": 6.073675593497836,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.762,
      "step": 25965
    },
    {
      "epoch": 6.073909484270845,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 25966
    },
    {
      "epoch": 6.074143375043855,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 25967
    },
    {
      "epoch": 6.074377265816864,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 25968
    },
    {
      "epoch": 6.074611156589873,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6455,
      "step": 25969
    },
    {
      "epoch": 6.0748450473628814,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 25970
    },
    {
      "epoch": 6.07507893813589,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 25971
    },
    {
      "epoch": 6.075312828908899,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 25972
    },
    {
      "epoch": 6.075546719681909,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 25973
    },
    {
      "epoch": 6.075780610454918,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 25974
    },
    {
      "epoch": 6.0760145012279265,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 25975
    },
    {
      "epoch": 6.076248392000935,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 25976
    },
    {
      "epoch": 6.076482282773944,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 25977
    },
    {
      "epoch": 6.076716173546954,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 25978
    },
    {
      "epoch": 6.076950064319963,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6096,
      "step": 25979
    },
    {
      "epoch": 6.077183955092972,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.4616,
      "step": 25980
    },
    {
      "epoch": 6.07741784586598,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 25981
    },
    {
      "epoch": 6.077651736638989,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6226,
      "step": 25982
    },
    {
      "epoch": 6.077885627411999,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 25983
    },
    {
      "epoch": 6.078119518185008,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5272,
      "step": 25984
    },
    {
      "epoch": 6.078353408958017,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 25985
    },
    {
      "epoch": 6.0785872997310255,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6044,
      "step": 25986
    },
    {
      "epoch": 6.078821190504034,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.5089,
      "step": 25987
    },
    {
      "epoch": 6.079055081277043,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7337,
      "step": 25988
    },
    {
      "epoch": 6.079288972050053,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 25989
    },
    {
      "epoch": 6.079522862823062,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.736,
      "step": 25990
    },
    {
      "epoch": 6.079756753596071,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5059,
      "step": 25991
    },
    {
      "epoch": 6.079990644369079,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 25992
    },
    {
      "epoch": 6.080224535142088,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5331,
      "step": 25993
    },
    {
      "epoch": 6.080458425915098,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7049,
      "step": 25994
    },
    {
      "epoch": 6.080692316688107,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5935,
      "step": 25995
    },
    {
      "epoch": 6.080926207461116,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 25996
    },
    {
      "epoch": 6.0811600982341245,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5211,
      "step": 25997
    },
    {
      "epoch": 6.081393989007133,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 25998
    },
    {
      "epoch": 6.081627879780143,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6261,
      "step": 25999
    },
    {
      "epoch": 6.081861770553152,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7064,
      "step": 26000
    },
    {
      "epoch": 6.081861770553152,
      "eval_runtime": 4.614,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 26000
    },
    {
      "epoch": 6.082095661326161,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.4995,
      "step": 26001
    },
    {
      "epoch": 6.08232955209917,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 26002
    },
    {
      "epoch": 6.082563442872178,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 26003
    },
    {
      "epoch": 6.082797333645187,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6903,
      "step": 26004
    },
    {
      "epoch": 6.083031224418197,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6081,
      "step": 26005
    },
    {
      "epoch": 6.083265115191206,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 26006
    },
    {
      "epoch": 6.083499005964215,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 26007
    },
    {
      "epoch": 6.0837328967372235,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6886,
      "step": 26008
    },
    {
      "epoch": 6.083966787510232,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6167,
      "step": 26009
    },
    {
      "epoch": 6.084200678283242,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 26010
    },
    {
      "epoch": 6.084434569056251,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 26011
    },
    {
      "epoch": 6.08466845982926,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.4801,
      "step": 26012
    },
    {
      "epoch": 6.084902350602269,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 26013
    },
    {
      "epoch": 6.085136241375277,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2826,
      "step": 26014
    },
    {
      "epoch": 6.085370132148287,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 26015
    },
    {
      "epoch": 6.085604022921296,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 26016
    },
    {
      "epoch": 6.085837913694305,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 26017
    },
    {
      "epoch": 6.086071804467314,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 26018
    },
    {
      "epoch": 6.0863056952403225,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6135,
      "step": 26019
    },
    {
      "epoch": 6.086539586013331,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 26020
    },
    {
      "epoch": 6.086773476786341,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6804,
      "step": 26021
    },
    {
      "epoch": 6.08700736755935,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 26022
    },
    {
      "epoch": 6.087241258332359,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 26023
    },
    {
      "epoch": 6.087475149105368,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.65,
      "step": 26024
    },
    {
      "epoch": 6.087709039878376,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9432,
      "step": 26025
    },
    {
      "epoch": 6.087942930651386,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 26026
    },
    {
      "epoch": 6.088176821424395,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 26027
    },
    {
      "epoch": 6.088410712197404,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 26028
    },
    {
      "epoch": 6.088644602970413,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 26029
    },
    {
      "epoch": 6.0888784937434215,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 26030
    },
    {
      "epoch": 6.089112384516431,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 26031
    },
    {
      "epoch": 6.08934627528944,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 26032
    },
    {
      "epoch": 6.089580166062449,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 26033
    },
    {
      "epoch": 6.089814056835458,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6674,
      "step": 26034
    },
    {
      "epoch": 6.090047947608467,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7396,
      "step": 26035
    },
    {
      "epoch": 6.090281838381475,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 26036
    },
    {
      "epoch": 6.090515729154485,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 26037
    },
    {
      "epoch": 6.090749619927494,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 26038
    },
    {
      "epoch": 6.090983510700503,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4338,
      "step": 26039
    },
    {
      "epoch": 6.091217401473512,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 26040
    },
    {
      "epoch": 6.0914512922465205,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 26041
    },
    {
      "epoch": 6.09168518301953,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 26042
    },
    {
      "epoch": 6.091919073792539,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 26043
    },
    {
      "epoch": 6.092152964565548,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 26044
    },
    {
      "epoch": 6.092386855338557,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 26045
    },
    {
      "epoch": 6.092620746111566,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 26046
    },
    {
      "epoch": 6.092854636884575,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.4601,
      "step": 26047
    },
    {
      "epoch": 6.093088527657584,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6901,
      "step": 26048
    },
    {
      "epoch": 6.093322418430593,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 26049
    },
    {
      "epoch": 6.093556309203602,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6847,
      "step": 26050
    },
    {
      "epoch": 6.093790199976611,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 26051
    },
    {
      "epoch": 6.09402409074962,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 26052
    },
    {
      "epoch": 6.094257981522629,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6381,
      "step": 26053
    },
    {
      "epoch": 6.094491872295638,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 26054
    },
    {
      "epoch": 6.094725763068647,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 26055
    },
    {
      "epoch": 6.094959653841656,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6067,
      "step": 26056
    },
    {
      "epoch": 6.095193544614665,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6474,
      "step": 26057
    },
    {
      "epoch": 6.095427435387674,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 26058
    },
    {
      "epoch": 6.095661326160683,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5889,
      "step": 26059
    },
    {
      "epoch": 6.095895216933692,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 26060
    },
    {
      "epoch": 6.096129107706701,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 26061
    },
    {
      "epoch": 6.09636299847971,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6295,
      "step": 26062
    },
    {
      "epoch": 6.096596889252719,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6812,
      "step": 26063
    },
    {
      "epoch": 6.096830780025728,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6427,
      "step": 26064
    },
    {
      "epoch": 6.097064670798737,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 26065
    },
    {
      "epoch": 6.097298561571746,
      "grad_norm": 11.3125,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 26066
    },
    {
      "epoch": 6.097532452344755,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6669,
      "step": 26067
    },
    {
      "epoch": 6.0977663431177636,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 26068
    },
    {
      "epoch": 6.098000233890773,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 26069
    },
    {
      "epoch": 6.098234124663782,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 26070
    },
    {
      "epoch": 6.098468015436791,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 26071
    },
    {
      "epoch": 6.0987019062098,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 26072
    },
    {
      "epoch": 6.098935796982809,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6442,
      "step": 26073
    },
    {
      "epoch": 6.099169687755818,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6046,
      "step": 26074
    },
    {
      "epoch": 6.099403578528827,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 26075
    },
    {
      "epoch": 6.099637469301836,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7851,
      "step": 26076
    },
    {
      "epoch": 6.099871360074845,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 26077
    },
    {
      "epoch": 6.100105250847854,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 26078
    },
    {
      "epoch": 6.100339141620863,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 26079
    },
    {
      "epoch": 6.100573032393872,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 26080
    },
    {
      "epoch": 6.100806923166881,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5267,
      "step": 26081
    },
    {
      "epoch": 6.10104081393989,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 26082
    },
    {
      "epoch": 6.101274704712899,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 26083
    },
    {
      "epoch": 6.1015085954859085,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.729,
      "step": 26084
    },
    {
      "epoch": 6.101742486258917,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 26085
    },
    {
      "epoch": 6.101976377031926,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5988,
      "step": 26086
    },
    {
      "epoch": 6.102210267804935,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 26087
    },
    {
      "epoch": 6.102444158577944,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.632,
      "step": 26088
    },
    {
      "epoch": 6.102678049350953,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6449,
      "step": 26089
    },
    {
      "epoch": 6.102911940123962,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 26090
    },
    {
      "epoch": 6.103145830896971,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 26091
    },
    {
      "epoch": 6.10337972166998,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 26092
    },
    {
      "epoch": 6.103613612442989,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 26093
    },
    {
      "epoch": 6.103847503215998,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 26094
    },
    {
      "epoch": 6.1040813939890075,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 26095
    },
    {
      "epoch": 6.104315284762016,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 26096
    },
    {
      "epoch": 6.104549175535025,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 26097
    },
    {
      "epoch": 6.104783066308034,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 26098
    },
    {
      "epoch": 6.105016957081043,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 26099
    },
    {
      "epoch": 6.105250847854053,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 26100
    },
    {
      "epoch": 6.105250847854053,
      "eval_runtime": 4.5843,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 26100
    },
    {
      "epoch": 6.105484738627061,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5925,
      "step": 26101
    },
    {
      "epoch": 6.10571862940007,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 26102
    },
    {
      "epoch": 6.105952520173079,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 26103
    },
    {
      "epoch": 6.106186410946088,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6262,
      "step": 26104
    },
    {
      "epoch": 6.106420301719097,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 26105
    },
    {
      "epoch": 6.1066541924921065,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 26106
    },
    {
      "epoch": 6.106888083265115,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 26107
    },
    {
      "epoch": 6.107121974038124,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.4803,
      "step": 26108
    },
    {
      "epoch": 6.107355864811133,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 26109
    },
    {
      "epoch": 6.107589755584142,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 26110
    },
    {
      "epoch": 6.107823646357152,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6295,
      "step": 26111
    },
    {
      "epoch": 6.10805753713016,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5562,
      "step": 26112
    },
    {
      "epoch": 6.108291427903169,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6867,
      "step": 26113
    },
    {
      "epoch": 6.108525318676178,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 26114
    },
    {
      "epoch": 6.108759209449187,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.4379,
      "step": 26115
    },
    {
      "epoch": 6.108993100222197,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 26116
    },
    {
      "epoch": 6.1092269909952055,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 26117
    },
    {
      "epoch": 6.109460881768214,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 26118
    },
    {
      "epoch": 6.109694772541223,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 26119
    },
    {
      "epoch": 6.109928663314232,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 26120
    },
    {
      "epoch": 6.110162554087241,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5797,
      "step": 26121
    },
    {
      "epoch": 6.110396444860251,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 26122
    },
    {
      "epoch": 6.110630335633259,
      "grad_norm": 2.546875,
      "learning_rate": 3e-05,
      "loss": 1.3984,
      "step": 26123
    },
    {
      "epoch": 6.110864226406268,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4452,
      "step": 26124
    },
    {
      "epoch": 6.111098117179277,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 26125
    },
    {
      "epoch": 6.111332007952286,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 26126
    },
    {
      "epoch": 6.111565898725296,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 26127
    },
    {
      "epoch": 6.1117997894983045,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.538,
      "step": 26128
    },
    {
      "epoch": 6.112033680271313,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4839,
      "step": 26129
    },
    {
      "epoch": 6.112267571044322,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 26130
    },
    {
      "epoch": 6.112501461817331,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 26131
    },
    {
      "epoch": 6.112735352590341,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 26132
    },
    {
      "epoch": 6.11296924336335,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 26133
    },
    {
      "epoch": 6.113203134136358,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 26134
    },
    {
      "epoch": 6.113437024909367,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.3574,
      "step": 26135
    },
    {
      "epoch": 6.113670915682376,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.445,
      "step": 26136
    },
    {
      "epoch": 6.113904806455385,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5869,
      "step": 26137
    },
    {
      "epoch": 6.114138697228395,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 26138
    },
    {
      "epoch": 6.1143725880014035,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8242,
      "step": 26139
    },
    {
      "epoch": 6.114606478774412,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7643,
      "step": 26140
    },
    {
      "epoch": 6.114840369547421,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 26141
    },
    {
      "epoch": 6.11507426032043,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 26142
    },
    {
      "epoch": 6.11530815109344,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.4528,
      "step": 26143
    },
    {
      "epoch": 6.115542041866449,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8608,
      "step": 26144
    },
    {
      "epoch": 6.115775932639457,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 26145
    },
    {
      "epoch": 6.116009823412466,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7265,
      "step": 26146
    },
    {
      "epoch": 6.116243714185475,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 26147
    },
    {
      "epoch": 6.116477604958485,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 26148
    },
    {
      "epoch": 6.116711495731494,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 26149
    },
    {
      "epoch": 6.1169453865045025,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 26150
    },
    {
      "epoch": 6.117179277277511,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 26151
    },
    {
      "epoch": 6.11741316805052,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 26152
    },
    {
      "epoch": 6.117647058823529,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 26153
    },
    {
      "epoch": 6.117880949596539,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 26154
    },
    {
      "epoch": 6.118114840369548,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 26155
    },
    {
      "epoch": 6.118348731142556,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 26156
    },
    {
      "epoch": 6.118582621915565,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 26157
    },
    {
      "epoch": 6.118816512688574,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 26158
    },
    {
      "epoch": 6.119050403461584,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 26159
    },
    {
      "epoch": 6.119284294234593,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 26160
    },
    {
      "epoch": 6.1195181850076015,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 26161
    },
    {
      "epoch": 6.11975207578061,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8346,
      "step": 26162
    },
    {
      "epoch": 6.119985966553619,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 26163
    },
    {
      "epoch": 6.120219857326629,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 26164
    },
    {
      "epoch": 6.120453748099638,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4449,
      "step": 26165
    },
    {
      "epoch": 6.1206876388726466,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6803,
      "step": 26166
    },
    {
      "epoch": 6.120921529645655,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 26167
    },
    {
      "epoch": 6.121155420418664,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 26168
    },
    {
      "epoch": 6.121389311191673,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 26169
    },
    {
      "epoch": 6.121623201964683,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 26170
    },
    {
      "epoch": 6.121857092737692,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8306,
      "step": 26171
    },
    {
      "epoch": 6.1220909835107005,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 26172
    },
    {
      "epoch": 6.122324874283709,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5628,
      "step": 26173
    },
    {
      "epoch": 6.122558765056718,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.587,
      "step": 26174
    },
    {
      "epoch": 6.122792655829728,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 26175
    },
    {
      "epoch": 6.123026546602737,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 26176
    },
    {
      "epoch": 6.1232604373757455,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 26177
    },
    {
      "epoch": 6.123494328148754,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7271,
      "step": 26178
    },
    {
      "epoch": 6.123728218921763,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.6679,
      "step": 26179
    },
    {
      "epoch": 6.123962109694773,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 26180
    },
    {
      "epoch": 6.124196000467782,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.674,
      "step": 26181
    },
    {
      "epoch": 6.124429891240791,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 26182
    },
    {
      "epoch": 6.1246637820137995,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 26183
    },
    {
      "epoch": 6.124897672786808,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6305,
      "step": 26184
    },
    {
      "epoch": 6.125131563559817,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6475,
      "step": 26185
    },
    {
      "epoch": 6.125365454332827,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6479,
      "step": 26186
    },
    {
      "epoch": 6.125599345105836,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 26187
    },
    {
      "epoch": 6.1258332358788445,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 26188
    },
    {
      "epoch": 6.126067126651853,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 26189
    },
    {
      "epoch": 6.126301017424862,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 26190
    },
    {
      "epoch": 6.126534908197872,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7278,
      "step": 26191
    },
    {
      "epoch": 6.126768798970881,
      "grad_norm": 8.875,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 26192
    },
    {
      "epoch": 6.12700268974389,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 26193
    },
    {
      "epoch": 6.1272365805168985,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 26194
    },
    {
      "epoch": 6.127470471289907,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 26195
    },
    {
      "epoch": 6.127704362062917,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 26196
    },
    {
      "epoch": 6.127938252835926,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7642,
      "step": 26197
    },
    {
      "epoch": 6.128172143608935,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 26198
    },
    {
      "epoch": 6.1284060343819435,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5487,
      "step": 26199
    },
    {
      "epoch": 6.128639925154952,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 26200
    },
    {
      "epoch": 6.128639925154952,
      "eval_runtime": 4.6501,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 26200
    },
    {
      "epoch": 6.128873815927962,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 26201
    },
    {
      "epoch": 6.129107706700971,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 26202
    },
    {
      "epoch": 6.12934159747398,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5685,
      "step": 26203
    },
    {
      "epoch": 6.129575488246989,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5899,
      "step": 26204
    },
    {
      "epoch": 6.1298093790199975,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 26205
    },
    {
      "epoch": 6.130043269793006,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 26206
    },
    {
      "epoch": 6.130277160566016,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 26207
    },
    {
      "epoch": 6.130511051339025,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6434,
      "step": 26208
    },
    {
      "epoch": 6.130744942112034,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 26209
    },
    {
      "epoch": 6.1309788328850425,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 26210
    },
    {
      "epoch": 6.131212723658051,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 26211
    },
    {
      "epoch": 6.131446614431061,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 26212
    },
    {
      "epoch": 6.13168050520407,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 26213
    },
    {
      "epoch": 6.131914395977079,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 26214
    },
    {
      "epoch": 6.132148286750088,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 26215
    },
    {
      "epoch": 6.1323821775230964,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 26216
    },
    {
      "epoch": 6.132616068296105,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 26217
    },
    {
      "epoch": 6.132849959069115,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 26218
    },
    {
      "epoch": 6.133083849842124,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8929,
      "step": 26219
    },
    {
      "epoch": 6.133317740615133,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6408,
      "step": 26220
    },
    {
      "epoch": 6.1335516313881415,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5599,
      "step": 26221
    },
    {
      "epoch": 6.13378552216115,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 26222
    },
    {
      "epoch": 6.13401941293416,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8676,
      "step": 26223
    },
    {
      "epoch": 6.134253303707169,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 26224
    },
    {
      "epoch": 6.134487194480178,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 26225
    },
    {
      "epoch": 6.134721085253187,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.7764,
      "step": 26226
    },
    {
      "epoch": 6.134954976026195,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.651,
      "step": 26227
    },
    {
      "epoch": 6.135188866799205,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 26228
    },
    {
      "epoch": 6.135422757572214,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5419,
      "step": 26229
    },
    {
      "epoch": 6.135656648345223,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 26230
    },
    {
      "epoch": 6.135890539118232,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6699,
      "step": 26231
    },
    {
      "epoch": 6.1361244298912405,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 26232
    },
    {
      "epoch": 6.13635832066425,
      "grad_norm": 7.625,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 26233
    },
    {
      "epoch": 6.136592211437259,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 26234
    },
    {
      "epoch": 6.136826102210268,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 26235
    },
    {
      "epoch": 6.137059992983277,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 26236
    },
    {
      "epoch": 6.137293883756286,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 26237
    },
    {
      "epoch": 6.137527774529294,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6257,
      "step": 26238
    },
    {
      "epoch": 6.137761665302304,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6976,
      "step": 26239
    },
    {
      "epoch": 6.137995556075313,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.4098,
      "step": 26240
    },
    {
      "epoch": 6.138229446848322,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 26241
    },
    {
      "epoch": 6.138463337621331,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 26242
    },
    {
      "epoch": 6.1386972283943395,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6441,
      "step": 26243
    },
    {
      "epoch": 6.138931119167349,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 26244
    },
    {
      "epoch": 6.139165009940358,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6489,
      "step": 26245
    },
    {
      "epoch": 6.139398900713367,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 26246
    },
    {
      "epoch": 6.139632791486376,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 26247
    },
    {
      "epoch": 6.139866682259385,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 26248
    },
    {
      "epoch": 6.140100573032393,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 26249
    },
    {
      "epoch": 6.140334463805403,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7245,
      "step": 26250
    },
    {
      "epoch": 6.140568354578412,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 26251
    },
    {
      "epoch": 6.140802245351421,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 26252
    },
    {
      "epoch": 6.14103613612443,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 26253
    },
    {
      "epoch": 6.1412700268974385,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 26254
    },
    {
      "epoch": 6.141503917670448,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7416,
      "step": 26255
    },
    {
      "epoch": 6.141737808443457,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 26256
    },
    {
      "epoch": 6.141971699216466,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 26257
    },
    {
      "epoch": 6.142205589989475,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 26258
    },
    {
      "epoch": 6.142439480762484,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6902,
      "step": 26259
    },
    {
      "epoch": 6.142673371535493,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 26260
    },
    {
      "epoch": 6.142907262308502,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7013,
      "step": 26261
    },
    {
      "epoch": 6.143141153081511,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 26262
    },
    {
      "epoch": 6.14337504385452,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 26263
    },
    {
      "epoch": 6.143608934627529,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 26264
    },
    {
      "epoch": 6.143842825400538,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5959,
      "step": 26265
    },
    {
      "epoch": 6.144076716173547,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 26266
    },
    {
      "epoch": 6.144310606946556,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9759,
      "step": 26267
    },
    {
      "epoch": 6.144544497719565,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 26268
    },
    {
      "epoch": 6.144778388492574,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 26269
    },
    {
      "epoch": 6.145012279265583,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.702,
      "step": 26270
    },
    {
      "epoch": 6.145246170038592,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.61,
      "step": 26271
    },
    {
      "epoch": 6.145480060811601,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 26272
    },
    {
      "epoch": 6.14571395158461,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 26273
    },
    {
      "epoch": 6.145947842357619,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7209,
      "step": 26274
    },
    {
      "epoch": 6.146181733130628,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5861,
      "step": 26275
    },
    {
      "epoch": 6.146415623903637,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 26276
    },
    {
      "epoch": 6.146649514676646,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 26277
    },
    {
      "epoch": 6.146883405449655,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6208,
      "step": 26278
    },
    {
      "epoch": 6.147117296222664,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 26279
    },
    {
      "epoch": 6.147351186995673,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 26280
    },
    {
      "epoch": 6.147585077768682,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 26281
    },
    {
      "epoch": 6.147818968541691,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 26282
    },
    {
      "epoch": 6.1480528593147,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 26283
    },
    {
      "epoch": 6.148286750087709,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 26284
    },
    {
      "epoch": 6.148520640860718,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 26285
    },
    {
      "epoch": 6.148754531633727,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 26286
    },
    {
      "epoch": 6.148988422406736,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 26287
    },
    {
      "epoch": 6.149222313179745,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7372,
      "step": 26288
    },
    {
      "epoch": 6.149456203952754,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 26289
    },
    {
      "epoch": 6.149690094725763,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6971,
      "step": 26290
    },
    {
      "epoch": 6.149923985498772,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8363,
      "step": 26291
    },
    {
      "epoch": 6.1501578762717815,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6836,
      "step": 26292
    },
    {
      "epoch": 6.15039176704479,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.662,
      "step": 26293
    },
    {
      "epoch": 6.150625657817799,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.4496,
      "step": 26294
    },
    {
      "epoch": 6.150859548590808,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6604,
      "step": 26295
    },
    {
      "epoch": 6.151093439363817,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 26296
    },
    {
      "epoch": 6.1513273301368265,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 26297
    },
    {
      "epoch": 6.151561220909835,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6668,
      "step": 26298
    },
    {
      "epoch": 6.151795111682844,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 26299
    },
    {
      "epoch": 6.152029002455853,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 26300
    },
    {
      "epoch": 6.152029002455853,
      "eval_runtime": 4.5885,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 26300
    },
    {
      "epoch": 6.152262893228862,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.3792,
      "step": 26301
    },
    {
      "epoch": 6.152496784001871,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7424,
      "step": 26302
    },
    {
      "epoch": 6.1527306747748804,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5763,
      "step": 26303
    },
    {
      "epoch": 6.152964565547889,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 26304
    },
    {
      "epoch": 6.153198456320898,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1719,
      "step": 26305
    },
    {
      "epoch": 6.153432347093907,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 26306
    },
    {
      "epoch": 6.153666237866916,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 26307
    },
    {
      "epoch": 6.1539001286399255,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.5641,
      "step": 26308
    },
    {
      "epoch": 6.154134019412934,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 26309
    },
    {
      "epoch": 6.154367910185943,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.439,
      "step": 26310
    },
    {
      "epoch": 6.154601800958952,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8101,
      "step": 26311
    },
    {
      "epoch": 6.154835691731961,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 26312
    },
    {
      "epoch": 6.155069582504971,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 26313
    },
    {
      "epoch": 6.155303473277979,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 26314
    },
    {
      "epoch": 6.155537364050988,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5747,
      "step": 26315
    },
    {
      "epoch": 6.155771254823997,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8751,
      "step": 26316
    },
    {
      "epoch": 6.156005145597006,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 26317
    },
    {
      "epoch": 6.156239036370015,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 26318
    },
    {
      "epoch": 6.1564729271430245,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 26319
    },
    {
      "epoch": 6.156706817916033,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 26320
    },
    {
      "epoch": 6.156940708689042,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 26321
    },
    {
      "epoch": 6.157174599462051,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 26322
    },
    {
      "epoch": 6.15740849023506,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 26323
    },
    {
      "epoch": 6.15764238100807,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7166,
      "step": 26324
    },
    {
      "epoch": 6.157876271781078,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6862,
      "step": 26325
    },
    {
      "epoch": 6.158110162554087,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 26326
    },
    {
      "epoch": 6.158344053327096,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 26327
    },
    {
      "epoch": 6.158577944100105,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 26328
    },
    {
      "epoch": 6.158811834873115,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6461,
      "step": 26329
    },
    {
      "epoch": 6.1590457256461235,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7528,
      "step": 26330
    },
    {
      "epoch": 6.159279616419132,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 26331
    },
    {
      "epoch": 6.159513507192141,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 26332
    },
    {
      "epoch": 6.15974739796515,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5655,
      "step": 26333
    },
    {
      "epoch": 6.159981288738159,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.3507,
      "step": 26334
    },
    {
      "epoch": 6.160215179511169,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 26335
    },
    {
      "epoch": 6.160449070284177,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 26336
    },
    {
      "epoch": 6.160682961057186,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4783,
      "step": 26337
    },
    {
      "epoch": 6.160916851830195,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 26338
    },
    {
      "epoch": 6.161150742603204,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 26339
    },
    {
      "epoch": 6.161384633376214,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 26340
    },
    {
      "epoch": 6.1616185241492225,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 26341
    },
    {
      "epoch": 6.161852414922231,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 26342
    },
    {
      "epoch": 6.16208630569524,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 26343
    },
    {
      "epoch": 6.162320196468249,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5758,
      "step": 26344
    },
    {
      "epoch": 6.162554087241259,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6616,
      "step": 26345
    },
    {
      "epoch": 6.162787978014268,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 26346
    },
    {
      "epoch": 6.163021868787276,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7172,
      "step": 26347
    },
    {
      "epoch": 6.163255759560285,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4949,
      "step": 26348
    },
    {
      "epoch": 6.163489650333294,
      "grad_norm": 11.0,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 26349
    },
    {
      "epoch": 6.163723541106303,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 26350
    },
    {
      "epoch": 6.163957431879313,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 26351
    },
    {
      "epoch": 6.1641913226523215,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 26352
    },
    {
      "epoch": 6.16442521342533,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 1.8576,
      "step": 26353
    },
    {
      "epoch": 6.164659104198339,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3235,
      "step": 26354
    },
    {
      "epoch": 6.164892994971348,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 26355
    },
    {
      "epoch": 6.165126885744358,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 26356
    },
    {
      "epoch": 6.165360776517367,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.655,
      "step": 26357
    },
    {
      "epoch": 6.165594667290375,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 26358
    },
    {
      "epoch": 6.165828558063384,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 26359
    },
    {
      "epoch": 6.166062448836393,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6465,
      "step": 26360
    },
    {
      "epoch": 6.166296339609403,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6574,
      "step": 26361
    },
    {
      "epoch": 6.166530230382412,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 26362
    },
    {
      "epoch": 6.1667641211554205,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 26363
    },
    {
      "epoch": 6.166998011928429,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 26364
    },
    {
      "epoch": 6.167231902701438,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 26365
    },
    {
      "epoch": 6.167465793474447,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 26366
    },
    {
      "epoch": 6.167699684247457,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 26367
    },
    {
      "epoch": 6.167933575020466,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7573,
      "step": 26368
    },
    {
      "epoch": 6.168167465793474,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 26369
    },
    {
      "epoch": 6.168401356566483,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 26370
    },
    {
      "epoch": 6.168635247339492,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 26371
    },
    {
      "epoch": 6.168869138112502,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5824,
      "step": 26372
    },
    {
      "epoch": 6.169103028885511,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 26373
    },
    {
      "epoch": 6.1693369196585195,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7327,
      "step": 26374
    },
    {
      "epoch": 6.169570810431528,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 26375
    },
    {
      "epoch": 6.169804701204537,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6507,
      "step": 26376
    },
    {
      "epoch": 6.170038591977547,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9563,
      "step": 26377
    },
    {
      "epoch": 6.170272482750556,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 26378
    },
    {
      "epoch": 6.170506373523565,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 26379
    },
    {
      "epoch": 6.170740264296573,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 26380
    },
    {
      "epoch": 6.170974155069582,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 26381
    },
    {
      "epoch": 6.171208045842591,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6052,
      "step": 26382
    },
    {
      "epoch": 6.171441936615601,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6593,
      "step": 26383
    },
    {
      "epoch": 6.17167582738861,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 26384
    },
    {
      "epoch": 6.1719097181616185,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 26385
    },
    {
      "epoch": 6.172143608934627,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 26386
    },
    {
      "epoch": 6.172377499707636,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 26387
    },
    {
      "epoch": 6.172611390480646,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8089,
      "step": 26388
    },
    {
      "epoch": 6.172845281253655,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 26389
    },
    {
      "epoch": 6.173079172026664,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 26390
    },
    {
      "epoch": 6.173313062799672,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 26391
    },
    {
      "epoch": 6.173546953572681,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 26392
    },
    {
      "epoch": 6.173780844345691,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.4145,
      "step": 26393
    },
    {
      "epoch": 6.1740147351187,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.598,
      "step": 26394
    },
    {
      "epoch": 6.174248625891709,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4447,
      "step": 26395
    },
    {
      "epoch": 6.1744825166647175,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 26396
    },
    {
      "epoch": 6.174716407437726,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 26397
    },
    {
      "epoch": 6.174950298210735,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 26398
    },
    {
      "epoch": 6.175184188983745,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 26399
    },
    {
      "epoch": 6.175418079756754,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 26400
    },
    {
      "epoch": 6.175418079756754,
      "eval_runtime": 4.626,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 26400
    },
    {
      "epoch": 6.175651970529763,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 26401
    },
    {
      "epoch": 6.175885861302771,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 26402
    },
    {
      "epoch": 6.17611975207578,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 26403
    },
    {
      "epoch": 6.17635364284879,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.609,
      "step": 26404
    },
    {
      "epoch": 6.176587533621799,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7781,
      "step": 26405
    },
    {
      "epoch": 6.176821424394808,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6185,
      "step": 26406
    },
    {
      "epoch": 6.1770553151678165,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 26407
    },
    {
      "epoch": 6.177289205940825,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 26408
    },
    {
      "epoch": 6.177523096713835,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 26409
    },
    {
      "epoch": 6.177756987486844,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6129,
      "step": 26410
    },
    {
      "epoch": 6.177990878259853,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 26411
    },
    {
      "epoch": 6.1782247690328616,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 26412
    },
    {
      "epoch": 6.17845865980587,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 26413
    },
    {
      "epoch": 6.17869255057888,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5399,
      "step": 26414
    },
    {
      "epoch": 6.178926441351889,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 26415
    },
    {
      "epoch": 6.179160332124898,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.622,
      "step": 26416
    },
    {
      "epoch": 6.179394222897907,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 26417
    },
    {
      "epoch": 6.1796281136709155,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 26418
    },
    {
      "epoch": 6.179862004443924,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8905,
      "step": 26419
    },
    {
      "epoch": 6.180095895216934,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 26420
    },
    {
      "epoch": 6.180329785989943,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6219,
      "step": 26421
    },
    {
      "epoch": 6.180563676762952,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 26422
    },
    {
      "epoch": 6.1807975675359605,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 26423
    },
    {
      "epoch": 6.181031458308969,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 26424
    },
    {
      "epoch": 6.181265349081979,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 26425
    },
    {
      "epoch": 6.181499239854988,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6051,
      "step": 26426
    },
    {
      "epoch": 6.181733130627997,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 26427
    },
    {
      "epoch": 6.181967021401006,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 26428
    },
    {
      "epoch": 6.1822009121740145,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 1.8601,
      "step": 26429
    },
    {
      "epoch": 6.182434802947023,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.3295,
      "step": 26430
    },
    {
      "epoch": 6.182668693720033,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 26431
    },
    {
      "epoch": 6.182902584493042,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5942,
      "step": 26432
    },
    {
      "epoch": 6.183136475266051,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 26433
    },
    {
      "epoch": 6.1833703660390595,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6779,
      "step": 26434
    },
    {
      "epoch": 6.183604256812068,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 26435
    },
    {
      "epoch": 6.183838147585078,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4568,
      "step": 26436
    },
    {
      "epoch": 6.184072038358087,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7355,
      "step": 26437
    },
    {
      "epoch": 6.184305929131096,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5629,
      "step": 26438
    },
    {
      "epoch": 6.184539819904105,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6702,
      "step": 26439
    },
    {
      "epoch": 6.1847737106771135,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 26440
    },
    {
      "epoch": 6.185007601450123,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 26441
    },
    {
      "epoch": 6.185241492223132,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 26442
    },
    {
      "epoch": 6.185475382996141,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6178,
      "step": 26443
    },
    {
      "epoch": 6.18570927376915,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 26444
    },
    {
      "epoch": 6.1859431645421585,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 26445
    },
    {
      "epoch": 6.186177055315168,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 26446
    },
    {
      "epoch": 6.186410946088177,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7194,
      "step": 26447
    },
    {
      "epoch": 6.186644836861186,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 26448
    },
    {
      "epoch": 6.186878727634195,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 26449
    },
    {
      "epoch": 6.187112618407204,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 26450
    },
    {
      "epoch": 6.1873465091802125,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 26451
    },
    {
      "epoch": 6.187580399953222,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 26452
    },
    {
      "epoch": 6.187814290726231,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7718,
      "step": 26453
    },
    {
      "epoch": 6.18804818149924,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 26454
    },
    {
      "epoch": 6.188282072272249,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 26455
    },
    {
      "epoch": 6.1885159630452575,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 26456
    },
    {
      "epoch": 6.188749853818267,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 26457
    },
    {
      "epoch": 6.188983744591276,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.3473,
      "step": 26458
    },
    {
      "epoch": 6.189217635364285,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6205,
      "step": 26459
    },
    {
      "epoch": 6.189451526137294,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 26460
    },
    {
      "epoch": 6.189685416910303,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 26461
    },
    {
      "epoch": 6.1899193076833114,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 26462
    },
    {
      "epoch": 6.190153198456321,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 26463
    },
    {
      "epoch": 6.19038708922933,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 26464
    },
    {
      "epoch": 6.190620980002339,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5649,
      "step": 26465
    },
    {
      "epoch": 6.190854870775348,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6094,
      "step": 26466
    },
    {
      "epoch": 6.1910887615483565,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5778,
      "step": 26467
    },
    {
      "epoch": 6.191322652321366,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7265,
      "step": 26468
    },
    {
      "epoch": 6.191556543094375,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 26469
    },
    {
      "epoch": 6.191790433867384,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 26470
    },
    {
      "epoch": 6.192024324640393,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4677,
      "step": 26471
    },
    {
      "epoch": 6.192258215413402,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 26472
    },
    {
      "epoch": 6.192492106186411,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6878,
      "step": 26473
    },
    {
      "epoch": 6.19272599695942,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 26474
    },
    {
      "epoch": 6.192959887732429,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 26475
    },
    {
      "epoch": 6.193193778505438,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 26476
    },
    {
      "epoch": 6.193427669278447,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8002,
      "step": 26477
    },
    {
      "epoch": 6.193661560051456,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 26478
    },
    {
      "epoch": 6.193895450824465,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6174,
      "step": 26479
    },
    {
      "epoch": 6.194129341597474,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7335,
      "step": 26480
    },
    {
      "epoch": 6.194363232370483,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 26481
    },
    {
      "epoch": 6.194597123143492,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 26482
    },
    {
      "epoch": 6.194831013916501,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6796,
      "step": 26483
    },
    {
      "epoch": 6.19506490468951,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 26484
    },
    {
      "epoch": 6.195298795462519,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 26485
    },
    {
      "epoch": 6.195532686235528,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7008,
      "step": 26486
    },
    {
      "epoch": 6.195766577008537,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 26487
    },
    {
      "epoch": 6.196000467781546,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5535,
      "step": 26488
    },
    {
      "epoch": 6.196234358554555,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 26489
    },
    {
      "epoch": 6.196468249327564,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 26490
    },
    {
      "epoch": 6.196702140100573,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 26491
    },
    {
      "epoch": 6.196936030873582,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 26492
    },
    {
      "epoch": 6.197169921646591,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 26493
    },
    {
      "epoch": 6.1974038124196,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 26494
    },
    {
      "epoch": 6.197637703192609,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.3885,
      "step": 26495
    },
    {
      "epoch": 6.197871593965618,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 26496
    },
    {
      "epoch": 6.198105484738627,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5975,
      "step": 26497
    },
    {
      "epoch": 6.198339375511636,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 26498
    },
    {
      "epoch": 6.198573266284645,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 26499
    },
    {
      "epoch": 6.198807157057654,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8034,
      "step": 26500
    },
    {
      "epoch": 6.198807157057654,
      "eval_runtime": 4.5939,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 26500
    },
    {
      "epoch": 6.199041047830663,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6682,
      "step": 26501
    },
    {
      "epoch": 6.199274938603672,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5634,
      "step": 26502
    },
    {
      "epoch": 6.199508829376681,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 26503
    },
    {
      "epoch": 6.19974272014969,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7297,
      "step": 26504
    },
    {
      "epoch": 6.1999766109226995,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 26505
    },
    {
      "epoch": 6.200210501695708,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 26506
    },
    {
      "epoch": 6.200444392468717,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 26507
    },
    {
      "epoch": 6.200678283241726,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6608,
      "step": 26508
    },
    {
      "epoch": 6.200912174014735,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 26509
    },
    {
      "epoch": 6.2011460647877445,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 26510
    },
    {
      "epoch": 6.201379955560753,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.628,
      "step": 26511
    },
    {
      "epoch": 6.201613846333762,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 26512
    },
    {
      "epoch": 6.201847737106771,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 26513
    },
    {
      "epoch": 6.20208162787978,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 26514
    },
    {
      "epoch": 6.202315518652789,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 26515
    },
    {
      "epoch": 6.2025494094257985,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 26516
    },
    {
      "epoch": 6.202783300198807,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.727,
      "step": 26517
    },
    {
      "epoch": 6.203017190971816,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6986,
      "step": 26518
    },
    {
      "epoch": 6.203251081744825,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 26519
    },
    {
      "epoch": 6.203484972517834,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 26520
    },
    {
      "epoch": 6.2037188632908435,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 26521
    },
    {
      "epoch": 6.203952754063852,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5589,
      "step": 26522
    },
    {
      "epoch": 6.204186644836861,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5536,
      "step": 26523
    },
    {
      "epoch": 6.20442053560987,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6161,
      "step": 26524
    },
    {
      "epoch": 6.204654426382879,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 26525
    },
    {
      "epoch": 6.204888317155889,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 26526
    },
    {
      "epoch": 6.2051222079288975,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 26527
    },
    {
      "epoch": 6.205356098701906,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5876,
      "step": 26528
    },
    {
      "epoch": 6.205589989474915,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4386,
      "step": 26529
    },
    {
      "epoch": 6.205823880247924,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6806,
      "step": 26530
    },
    {
      "epoch": 6.206057771020933,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 26531
    },
    {
      "epoch": 6.2062916617939425,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 26532
    },
    {
      "epoch": 6.206525552566951,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 26533
    },
    {
      "epoch": 6.20675944333996,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 26534
    },
    {
      "epoch": 6.206993334112969,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 26535
    },
    {
      "epoch": 6.207227224885978,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6886,
      "step": 26536
    },
    {
      "epoch": 6.207461115658988,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 26537
    },
    {
      "epoch": 6.2076950064319965,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 26538
    },
    {
      "epoch": 6.207928897205005,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.532,
      "step": 26539
    },
    {
      "epoch": 6.208162787978014,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6604,
      "step": 26540
    },
    {
      "epoch": 6.208396678751023,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6111,
      "step": 26541
    },
    {
      "epoch": 6.208630569524033,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 26542
    },
    {
      "epoch": 6.2088644602970415,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 26543
    },
    {
      "epoch": 6.20909835107005,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7223,
      "step": 26544
    },
    {
      "epoch": 6.209332241843059,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6233,
      "step": 26545
    },
    {
      "epoch": 6.209566132616068,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 26546
    },
    {
      "epoch": 6.209800023389077,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5893,
      "step": 26547
    },
    {
      "epoch": 6.210033914162087,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 26548
    },
    {
      "epoch": 6.2102678049350954,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5462,
      "step": 26549
    },
    {
      "epoch": 6.210501695708104,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 26550
    },
    {
      "epoch": 6.210735586481113,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 26551
    },
    {
      "epoch": 6.210969477254122,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 26552
    },
    {
      "epoch": 6.211203368027132,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 26553
    },
    {
      "epoch": 6.2114372588001405,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 26554
    },
    {
      "epoch": 6.211671149573149,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6705,
      "step": 26555
    },
    {
      "epoch": 6.211905040346158,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 26556
    },
    {
      "epoch": 6.212138931119167,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 26557
    },
    {
      "epoch": 6.212372821892177,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7306,
      "step": 26558
    },
    {
      "epoch": 6.212606712665186,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 26559
    },
    {
      "epoch": 6.212840603438194,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 26560
    },
    {
      "epoch": 6.213074494211203,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 26561
    },
    {
      "epoch": 6.213308384984212,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6172,
      "step": 26562
    },
    {
      "epoch": 6.213542275757221,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 26563
    },
    {
      "epoch": 6.213776166530231,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 26564
    },
    {
      "epoch": 6.2140100573032395,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 26565
    },
    {
      "epoch": 6.214243948076248,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5218,
      "step": 26566
    },
    {
      "epoch": 6.214477838849257,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 26567
    },
    {
      "epoch": 6.214711729622266,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 26568
    },
    {
      "epoch": 6.214945620395276,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7731,
      "step": 26569
    },
    {
      "epoch": 6.215179511168285,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 26570
    },
    {
      "epoch": 6.215413401941293,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5637,
      "step": 26571
    },
    {
      "epoch": 6.215647292714302,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4747,
      "step": 26572
    },
    {
      "epoch": 6.215881183487311,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 26573
    },
    {
      "epoch": 6.216115074260321,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 26574
    },
    {
      "epoch": 6.21634896503333,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 26575
    },
    {
      "epoch": 6.2165828558063385,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6262,
      "step": 26576
    },
    {
      "epoch": 6.216816746579347,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 26577
    },
    {
      "epoch": 6.217050637352356,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5774,
      "step": 26578
    },
    {
      "epoch": 6.217284528125365,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.586,
      "step": 26579
    },
    {
      "epoch": 6.217518418898375,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4821,
      "step": 26580
    },
    {
      "epoch": 6.217752309671384,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 26581
    },
    {
      "epoch": 6.217986200444392,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 26582
    },
    {
      "epoch": 6.218220091217401,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 26583
    },
    {
      "epoch": 6.21845398199041,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 26584
    },
    {
      "epoch": 6.21868787276342,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 26585
    },
    {
      "epoch": 6.218921763536429,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 26586
    },
    {
      "epoch": 6.2191556543094375,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.2471,
      "step": 26587
    },
    {
      "epoch": 6.219389545082446,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4597,
      "step": 26588
    },
    {
      "epoch": 6.219623435855455,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 26589
    },
    {
      "epoch": 6.219857326628465,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 26590
    },
    {
      "epoch": 6.220091217401474,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6383,
      "step": 26591
    },
    {
      "epoch": 6.220325108174483,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 26592
    },
    {
      "epoch": 6.220558998947491,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7515,
      "step": 26593
    },
    {
      "epoch": 6.2207928897205,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 26594
    },
    {
      "epoch": 6.22102678049351,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6533,
      "step": 26595
    },
    {
      "epoch": 6.221260671266519,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 26596
    },
    {
      "epoch": 6.221494562039528,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 26597
    },
    {
      "epoch": 6.2217284528125365,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4567,
      "step": 26598
    },
    {
      "epoch": 6.221962343585545,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 26599
    },
    {
      "epoch": 6.222196234358554,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 26600
    },
    {
      "epoch": 6.222196234358554,
      "eval_runtime": 4.6765,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 26600
    },
    {
      "epoch": 6.222430125131564,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 26601
    },
    {
      "epoch": 6.222664015904573,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 26602
    },
    {
      "epoch": 6.222897906677582,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 26603
    },
    {
      "epoch": 6.22313179745059,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 26604
    },
    {
      "epoch": 6.223365688223599,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 26605
    },
    {
      "epoch": 6.223599578996609,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 26606
    },
    {
      "epoch": 6.223833469769618,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.5576,
      "step": 26607
    },
    {
      "epoch": 6.224067360542627,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 26608
    },
    {
      "epoch": 6.2243012513156355,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 26609
    },
    {
      "epoch": 6.224535142088644,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 26610
    },
    {
      "epoch": 6.224769032861653,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 26611
    },
    {
      "epoch": 6.225002923634663,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6556,
      "step": 26612
    },
    {
      "epoch": 6.225236814407672,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 26613
    },
    {
      "epoch": 6.225470705180681,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 26614
    },
    {
      "epoch": 6.225704595953689,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 26615
    },
    {
      "epoch": 6.225938486726698,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 26616
    },
    {
      "epoch": 6.226172377499708,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 26617
    },
    {
      "epoch": 6.226406268272717,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 26618
    },
    {
      "epoch": 6.226640159045726,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 26619
    },
    {
      "epoch": 6.2268740498187345,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 26620
    },
    {
      "epoch": 6.227107940591743,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 26621
    },
    {
      "epoch": 6.227341831364753,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 26622
    },
    {
      "epoch": 6.227575722137762,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 26623
    },
    {
      "epoch": 6.227809612910771,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6928,
      "step": 26624
    },
    {
      "epoch": 6.22804350368378,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 26625
    },
    {
      "epoch": 6.228277394456788,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 26626
    },
    {
      "epoch": 6.228511285229798,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6518,
      "step": 26627
    },
    {
      "epoch": 6.228745176002807,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 26628
    },
    {
      "epoch": 6.228979066775816,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.619,
      "step": 26629
    },
    {
      "epoch": 6.229212957548825,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 26630
    },
    {
      "epoch": 6.2294468483218335,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5286,
      "step": 26631
    },
    {
      "epoch": 6.229680739094842,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4911,
      "step": 26632
    },
    {
      "epoch": 6.229914629867852,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5281,
      "step": 26633
    },
    {
      "epoch": 6.230148520640861,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5845,
      "step": 26634
    },
    {
      "epoch": 6.23038241141387,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 26635
    },
    {
      "epoch": 6.230616302186879,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 26636
    },
    {
      "epoch": 6.230850192959887,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6188,
      "step": 26637
    },
    {
      "epoch": 6.231084083732897,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 26638
    },
    {
      "epoch": 6.231317974505906,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 26639
    },
    {
      "epoch": 6.231551865278915,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5024,
      "step": 26640
    },
    {
      "epoch": 6.231785756051924,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 26641
    },
    {
      "epoch": 6.2320196468249325,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 26642
    },
    {
      "epoch": 6.232253537597941,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6819,
      "step": 26643
    },
    {
      "epoch": 6.232487428370951,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5817,
      "step": 26644
    },
    {
      "epoch": 6.23272131914396,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6545,
      "step": 26645
    },
    {
      "epoch": 6.232955209916969,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9209,
      "step": 26646
    },
    {
      "epoch": 6.233189100689978,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 26647
    },
    {
      "epoch": 6.233422991462986,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6256,
      "step": 26648
    },
    {
      "epoch": 6.233656882235996,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 26649
    },
    {
      "epoch": 6.233890773009005,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 26650
    },
    {
      "epoch": 6.234124663782014,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 26651
    },
    {
      "epoch": 6.234358554555023,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6775,
      "step": 26652
    },
    {
      "epoch": 6.2345924453280315,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 26653
    },
    {
      "epoch": 6.234826336101041,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 26654
    },
    {
      "epoch": 6.23506022687405,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6519,
      "step": 26655
    },
    {
      "epoch": 6.235294117647059,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 26656
    },
    {
      "epoch": 6.235528008420068,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 26657
    },
    {
      "epoch": 6.2357618991930766,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7407,
      "step": 26658
    },
    {
      "epoch": 6.235995789966086,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 26659
    },
    {
      "epoch": 6.236229680739095,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6726,
      "step": 26660
    },
    {
      "epoch": 6.236463571512104,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 26661
    },
    {
      "epoch": 6.236697462285113,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5542,
      "step": 26662
    },
    {
      "epoch": 6.236931353058122,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 26663
    },
    {
      "epoch": 6.2371652438311305,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6463,
      "step": 26664
    },
    {
      "epoch": 6.23739913460414,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 26665
    },
    {
      "epoch": 6.237633025377149,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 26666
    },
    {
      "epoch": 6.237866916150158,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6663,
      "step": 26667
    },
    {
      "epoch": 6.238100806923167,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 26668
    },
    {
      "epoch": 6.2383346976961755,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 26669
    },
    {
      "epoch": 6.238568588469185,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6956,
      "step": 26670
    },
    {
      "epoch": 6.238802479242194,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 26671
    },
    {
      "epoch": 6.239036370015203,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 26672
    },
    {
      "epoch": 6.239270260788212,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 26673
    },
    {
      "epoch": 6.239504151561221,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 26674
    },
    {
      "epoch": 6.2397380423342295,
      "grad_norm": 7.40625,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 26675
    },
    {
      "epoch": 6.239971933107239,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6423,
      "step": 26676
    },
    {
      "epoch": 6.240205823880248,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 26677
    },
    {
      "epoch": 6.240439714653257,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 26678
    },
    {
      "epoch": 6.240673605426266,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 26679
    },
    {
      "epoch": 6.2409074961992745,
      "grad_norm": 7.40625,
      "learning_rate": 3e-05,
      "loss": 1.7054,
      "step": 26680
    },
    {
      "epoch": 6.241141386972284,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6705,
      "step": 26681
    },
    {
      "epoch": 6.241375277745293,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 26682
    },
    {
      "epoch": 6.241609168518302,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5555,
      "step": 26683
    },
    {
      "epoch": 6.241843059291311,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5259,
      "step": 26684
    },
    {
      "epoch": 6.24207695006432,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 26685
    },
    {
      "epoch": 6.242310840837329,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 26686
    },
    {
      "epoch": 6.242544731610338,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 26687
    },
    {
      "epoch": 6.242778622383347,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4365,
      "step": 26688
    },
    {
      "epoch": 6.243012513156356,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 26689
    },
    {
      "epoch": 6.243246403929365,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 26690
    },
    {
      "epoch": 6.243480294702374,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7723,
      "step": 26691
    },
    {
      "epoch": 6.243714185475383,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 26692
    },
    {
      "epoch": 6.243948076248392,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9786,
      "step": 26693
    },
    {
      "epoch": 6.244181967021401,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 26694
    },
    {
      "epoch": 6.24441585779441,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.631,
      "step": 26695
    },
    {
      "epoch": 6.244649748567419,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 26696
    },
    {
      "epoch": 6.244883639340428,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 26697
    },
    {
      "epoch": 6.245117530113437,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 26698
    },
    {
      "epoch": 6.245351420886446,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7503,
      "step": 26699
    },
    {
      "epoch": 6.245585311659455,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 26700
    },
    {
      "epoch": 6.245585311659455,
      "eval_runtime": 4.5968,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 26700
    },
    {
      "epoch": 6.245819202432464,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 26701
    },
    {
      "epoch": 6.246053093205473,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7443,
      "step": 26702
    },
    {
      "epoch": 6.246286983978482,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9155,
      "step": 26703
    },
    {
      "epoch": 6.246520874751491,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7045,
      "step": 26704
    },
    {
      "epoch": 6.2467547655245,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 26705
    },
    {
      "epoch": 6.246988656297509,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 26706
    },
    {
      "epoch": 6.2472225470705185,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 26707
    },
    {
      "epoch": 6.247456437843527,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7876,
      "step": 26708
    },
    {
      "epoch": 6.247690328616536,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 26709
    },
    {
      "epoch": 6.247924219389545,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7245,
      "step": 26710
    },
    {
      "epoch": 6.248158110162554,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 26711
    },
    {
      "epoch": 6.248392000935563,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 26712
    },
    {
      "epoch": 6.248625891708572,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 26713
    },
    {
      "epoch": 6.248859782481581,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 26714
    },
    {
      "epoch": 6.24909367325459,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 26715
    },
    {
      "epoch": 6.249327564027599,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.717,
      "step": 26716
    },
    {
      "epoch": 6.249561454800608,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8075,
      "step": 26717
    },
    {
      "epoch": 6.2497953455736175,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 26718
    },
    {
      "epoch": 6.250029236346626,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 26719
    },
    {
      "epoch": 6.250263127119635,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 26720
    },
    {
      "epoch": 6.250497017892644,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.3566,
      "step": 26721
    },
    {
      "epoch": 6.250730908665653,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 26722
    },
    {
      "epoch": 6.250964799438663,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 26723
    },
    {
      "epoch": 6.251198690211671,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6527,
      "step": 26724
    },
    {
      "epoch": 6.25143258098468,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 26725
    },
    {
      "epoch": 6.251666471757689,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5279,
      "step": 26726
    },
    {
      "epoch": 6.251900362530698,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 26727
    },
    {
      "epoch": 6.252134253303707,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8411,
      "step": 26728
    },
    {
      "epoch": 6.2523681440767165,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6188,
      "step": 26729
    },
    {
      "epoch": 6.252602034849725,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 26730
    },
    {
      "epoch": 6.252835925622734,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5879,
      "step": 26731
    },
    {
      "epoch": 6.253069816395743,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 26732
    },
    {
      "epoch": 6.253303707168752,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5334,
      "step": 26733
    },
    {
      "epoch": 6.253537597941762,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 26734
    },
    {
      "epoch": 6.25377148871477,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 26735
    },
    {
      "epoch": 6.254005379487779,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.597,
      "step": 26736
    },
    {
      "epoch": 6.254239270260788,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6817,
      "step": 26737
    },
    {
      "epoch": 6.254473161033797,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 26738
    },
    {
      "epoch": 6.254707051806806,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 26739
    },
    {
      "epoch": 6.2549409425798155,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 26740
    },
    {
      "epoch": 6.255174833352824,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 26741
    },
    {
      "epoch": 6.255408724125833,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.591,
      "step": 26742
    },
    {
      "epoch": 6.255642614898842,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6547,
      "step": 26743
    },
    {
      "epoch": 6.255876505671852,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 26744
    },
    {
      "epoch": 6.2561103964448606,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5563,
      "step": 26745
    },
    {
      "epoch": 6.256344287217869,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 26746
    },
    {
      "epoch": 6.256578177990878,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7011,
      "step": 26747
    },
    {
      "epoch": 6.256812068763887,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4202,
      "step": 26748
    },
    {
      "epoch": 6.257045959536896,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5116,
      "step": 26749
    },
    {
      "epoch": 6.257279850309906,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2794,
      "step": 26750
    },
    {
      "epoch": 6.2575137410829145,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 26751
    },
    {
      "epoch": 6.257747631855923,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 26752
    },
    {
      "epoch": 6.257981522628932,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 26753
    },
    {
      "epoch": 6.258215413401941,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5986,
      "step": 26754
    },
    {
      "epoch": 6.258449304174951,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7368,
      "step": 26755
    },
    {
      "epoch": 6.2586831949479595,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 26756
    },
    {
      "epoch": 6.258917085720968,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7469,
      "step": 26757
    },
    {
      "epoch": 6.259150976493977,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5909,
      "step": 26758
    },
    {
      "epoch": 6.259384867266986,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5972,
      "step": 26759
    },
    {
      "epoch": 6.259618758039995,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8416,
      "step": 26760
    },
    {
      "epoch": 6.259852648813005,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6823,
      "step": 26761
    },
    {
      "epoch": 6.2600865395860135,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.5673,
      "step": 26762
    },
    {
      "epoch": 6.260320430359022,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5005,
      "step": 26763
    },
    {
      "epoch": 6.260554321132031,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 26764
    },
    {
      "epoch": 6.26078821190504,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 26765
    },
    {
      "epoch": 6.26102210267805,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 26766
    },
    {
      "epoch": 6.2612559934510585,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 26767
    },
    {
      "epoch": 6.261489884224067,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 26768
    },
    {
      "epoch": 6.261723774997076,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 26769
    },
    {
      "epoch": 6.261957665770085,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 26770
    },
    {
      "epoch": 6.262191556543095,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 26771
    },
    {
      "epoch": 6.262425447316104,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5391,
      "step": 26772
    },
    {
      "epoch": 6.2626593380891125,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 26773
    },
    {
      "epoch": 6.262893228862121,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.543,
      "step": 26774
    },
    {
      "epoch": 6.26312711963513,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7054,
      "step": 26775
    },
    {
      "epoch": 6.26336101040814,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 26776
    },
    {
      "epoch": 6.263594901181149,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6607,
      "step": 26777
    },
    {
      "epoch": 6.2638287919541575,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 26778
    },
    {
      "epoch": 6.264062682727166,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6823,
      "step": 26779
    },
    {
      "epoch": 6.264296573500175,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6465,
      "step": 26780
    },
    {
      "epoch": 6.264530464273184,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 26781
    },
    {
      "epoch": 6.264764355046194,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.6207,
      "step": 26782
    },
    {
      "epoch": 6.264998245819203,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 26783
    },
    {
      "epoch": 6.2652321365922115,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 26784
    },
    {
      "epoch": 6.26546602736522,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5951,
      "step": 26785
    },
    {
      "epoch": 6.265699918138229,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6011,
      "step": 26786
    },
    {
      "epoch": 6.265933808911239,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 26787
    },
    {
      "epoch": 6.266167699684248,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 26788
    },
    {
      "epoch": 6.2664015904572565,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 26789
    },
    {
      "epoch": 6.266635481230265,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6395,
      "step": 26790
    },
    {
      "epoch": 6.266869372003274,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 26791
    },
    {
      "epoch": 6.267103262776283,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 26792
    },
    {
      "epoch": 6.267337153549293,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4518,
      "step": 26793
    },
    {
      "epoch": 6.267571044322302,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7477,
      "step": 26794
    },
    {
      "epoch": 6.2678049350953104,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 26795
    },
    {
      "epoch": 6.268038825868319,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 26796
    },
    {
      "epoch": 6.268272716641328,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 26797
    },
    {
      "epoch": 6.268506607414338,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 26798
    },
    {
      "epoch": 6.268740498187347,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 26799
    },
    {
      "epoch": 6.2689743889603555,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 26800
    },
    {
      "epoch": 6.2689743889603555,
      "eval_runtime": 4.629,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 26800
    },
    {
      "epoch": 6.269208279733364,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5938,
      "step": 26801
    },
    {
      "epoch": 6.269442170506373,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 26802
    },
    {
      "epoch": 6.269676061279383,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 26803
    },
    {
      "epoch": 6.269909952052392,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6952,
      "step": 26804
    },
    {
      "epoch": 6.270143842825401,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7295,
      "step": 26805
    },
    {
      "epoch": 6.270377733598409,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7984,
      "step": 26806
    },
    {
      "epoch": 6.270611624371418,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6403,
      "step": 26807
    },
    {
      "epoch": 6.270845515144428,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 26808
    },
    {
      "epoch": 6.271079405917437,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 26809
    },
    {
      "epoch": 6.271313296690446,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 26810
    },
    {
      "epoch": 6.2715471874634545,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5609,
      "step": 26811
    },
    {
      "epoch": 6.271781078236463,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6542,
      "step": 26812
    },
    {
      "epoch": 6.272014969009472,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 26813
    },
    {
      "epoch": 6.272248859782482,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 26814
    },
    {
      "epoch": 6.272482750555491,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 26815
    },
    {
      "epoch": 6.2727166413285,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 26816
    },
    {
      "epoch": 6.272950532101508,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.5179,
      "step": 26817
    },
    {
      "epoch": 6.273184422874517,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7086,
      "step": 26818
    },
    {
      "epoch": 6.273418313647527,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4291,
      "step": 26819
    },
    {
      "epoch": 6.273652204420536,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 26820
    },
    {
      "epoch": 6.273886095193545,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 26821
    },
    {
      "epoch": 6.2741199859665535,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 26822
    },
    {
      "epoch": 6.274353876739562,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 26823
    },
    {
      "epoch": 6.274587767512571,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 26824
    },
    {
      "epoch": 6.274821658285581,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7019,
      "step": 26825
    },
    {
      "epoch": 6.27505554905859,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 26826
    },
    {
      "epoch": 6.275289439831599,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 26827
    },
    {
      "epoch": 6.275523330604607,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 26828
    },
    {
      "epoch": 6.275757221377616,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.531,
      "step": 26829
    },
    {
      "epoch": 6.275991112150626,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 26830
    },
    {
      "epoch": 6.276225002923635,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 26831
    },
    {
      "epoch": 6.276458893696644,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8011,
      "step": 26832
    },
    {
      "epoch": 6.2766927844696525,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 26833
    },
    {
      "epoch": 6.276926675242661,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7818,
      "step": 26834
    },
    {
      "epoch": 6.277160566015671,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 26835
    },
    {
      "epoch": 6.27739445678868,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 26836
    },
    {
      "epoch": 6.277628347561689,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 26837
    },
    {
      "epoch": 6.277862238334698,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9506,
      "step": 26838
    },
    {
      "epoch": 6.278096129107706,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 26839
    },
    {
      "epoch": 6.278330019880716,
      "grad_norm": 13.0,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 26840
    },
    {
      "epoch": 6.278563910653725,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 26841
    },
    {
      "epoch": 6.278797801426734,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 26842
    },
    {
      "epoch": 6.279031692199743,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6587,
      "step": 26843
    },
    {
      "epoch": 6.2792655829727515,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 26844
    },
    {
      "epoch": 6.27949947374576,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 26845
    },
    {
      "epoch": 6.27973336451877,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 26846
    },
    {
      "epoch": 6.279967255291779,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 26847
    },
    {
      "epoch": 6.280201146064788,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 26848
    },
    {
      "epoch": 6.280435036837797,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5395,
      "step": 26849
    },
    {
      "epoch": 6.280668927610805,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7826,
      "step": 26850
    },
    {
      "epoch": 6.280902818383815,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6423,
      "step": 26851
    },
    {
      "epoch": 6.281136709156824,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 26852
    },
    {
      "epoch": 6.281370599929833,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 26853
    },
    {
      "epoch": 6.281604490702842,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 26854
    },
    {
      "epoch": 6.2818383814758505,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4734,
      "step": 26855
    },
    {
      "epoch": 6.282072272248859,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 26856
    },
    {
      "epoch": 6.282306163021869,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.571,
      "step": 26857
    },
    {
      "epoch": 6.282540053794878,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 26858
    },
    {
      "epoch": 6.282773944567887,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 26859
    },
    {
      "epoch": 6.283007835340896,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7667,
      "step": 26860
    },
    {
      "epoch": 6.283241726113904,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 26861
    },
    {
      "epoch": 6.283475616886914,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 26862
    },
    {
      "epoch": 6.283709507659923,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 26863
    },
    {
      "epoch": 6.283943398432932,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 26864
    },
    {
      "epoch": 6.284177289205941,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 26865
    },
    {
      "epoch": 6.2844111799789495,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 26866
    },
    {
      "epoch": 6.284645070751959,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 26867
    },
    {
      "epoch": 6.284878961524968,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.7534,
      "step": 26868
    },
    {
      "epoch": 6.285112852297977,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 26869
    },
    {
      "epoch": 6.285346743070986,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 26870
    },
    {
      "epoch": 6.285580633843995,
      "grad_norm": 7.34375,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 26871
    },
    {
      "epoch": 6.285814524617004,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5693,
      "step": 26872
    },
    {
      "epoch": 6.286048415390013,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6667,
      "step": 26873
    },
    {
      "epoch": 6.286282306163022,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7494,
      "step": 26874
    },
    {
      "epoch": 6.286516196936031,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 26875
    },
    {
      "epoch": 6.28675008770904,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 26876
    },
    {
      "epoch": 6.2869839784820485,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6245,
      "step": 26877
    },
    {
      "epoch": 6.287217869255058,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 26878
    },
    {
      "epoch": 6.287451760028067,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7495,
      "step": 26879
    },
    {
      "epoch": 6.287685650801076,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7781,
      "step": 26880
    },
    {
      "epoch": 6.287919541574085,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 26881
    },
    {
      "epoch": 6.288153432347094,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.58,
      "step": 26882
    },
    {
      "epoch": 6.288387323120103,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5535,
      "step": 26883
    },
    {
      "epoch": 6.288621213893112,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6391,
      "step": 26884
    },
    {
      "epoch": 6.288855104666121,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6418,
      "step": 26885
    },
    {
      "epoch": 6.28908899543913,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6631,
      "step": 26886
    },
    {
      "epoch": 6.289322886212139,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5169,
      "step": 26887
    },
    {
      "epoch": 6.2895567769851475,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 26888
    },
    {
      "epoch": 6.289790667758157,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7671,
      "step": 26889
    },
    {
      "epoch": 6.290024558531166,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 26890
    },
    {
      "epoch": 6.290258449304175,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 26891
    },
    {
      "epoch": 6.290492340077184,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 26892
    },
    {
      "epoch": 6.290726230850193,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 26893
    },
    {
      "epoch": 6.290960121623202,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 26894
    },
    {
      "epoch": 6.291194012396211,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 26895
    },
    {
      "epoch": 6.29142790316922,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6246,
      "step": 26896
    },
    {
      "epoch": 6.291661793942229,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 26897
    },
    {
      "epoch": 6.291895684715238,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 26898
    },
    {
      "epoch": 6.292129575488247,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 26899
    },
    {
      "epoch": 6.292363466261256,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.4763,
      "step": 26900
    },
    {
      "epoch": 6.292363466261256,
      "eval_runtime": 4.6238,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 26900
    },
    {
      "epoch": 6.292597357034265,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 26901
    },
    {
      "epoch": 6.292831247807274,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.59,
      "step": 26902
    },
    {
      "epoch": 6.293065138580283,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4742,
      "step": 26903
    },
    {
      "epoch": 6.293299029353292,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 26904
    },
    {
      "epoch": 6.293532920126301,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4558,
      "step": 26905
    },
    {
      "epoch": 6.29376681089931,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6913,
      "step": 26906
    },
    {
      "epoch": 6.294000701672319,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 26907
    },
    {
      "epoch": 6.294234592445328,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 26908
    },
    {
      "epoch": 6.294468483218337,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 26909
    },
    {
      "epoch": 6.294702373991346,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 26910
    },
    {
      "epoch": 6.294936264764355,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 26911
    },
    {
      "epoch": 6.295170155537364,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.493,
      "step": 26912
    },
    {
      "epoch": 6.295404046310373,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 26913
    },
    {
      "epoch": 6.295637937083382,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 26914
    },
    {
      "epoch": 6.295871827856391,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 26915
    },
    {
      "epoch": 6.2961057186294,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6186,
      "step": 26916
    },
    {
      "epoch": 6.296339609402409,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 26917
    },
    {
      "epoch": 6.296573500175418,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8135,
      "step": 26918
    },
    {
      "epoch": 6.296807390948427,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 26919
    },
    {
      "epoch": 6.297041281721436,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 26920
    },
    {
      "epoch": 6.297275172494445,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 26921
    },
    {
      "epoch": 6.297509063267454,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 26922
    },
    {
      "epoch": 6.297742954040463,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7258,
      "step": 26923
    },
    {
      "epoch": 6.297976844813472,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.488,
      "step": 26924
    },
    {
      "epoch": 6.298210735586481,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 26925
    },
    {
      "epoch": 6.29844462635949,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.5798,
      "step": 26926
    },
    {
      "epoch": 6.298678517132499,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6548,
      "step": 26927
    },
    {
      "epoch": 6.298912407905508,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6124,
      "step": 26928
    },
    {
      "epoch": 6.299146298678517,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 26929
    },
    {
      "epoch": 6.299380189451526,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 26930
    },
    {
      "epoch": 6.2996140802245355,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6067,
      "step": 26931
    },
    {
      "epoch": 6.299847970997544,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7728,
      "step": 26932
    },
    {
      "epoch": 6.300081861770553,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7332,
      "step": 26933
    },
    {
      "epoch": 6.300315752543562,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 26934
    },
    {
      "epoch": 6.300549643316571,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 26935
    },
    {
      "epoch": 6.300783534089581,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 26936
    },
    {
      "epoch": 6.301017424862589,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.8156,
      "step": 26937
    },
    {
      "epoch": 6.301251315635598,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 26938
    },
    {
      "epoch": 6.301485206408607,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7906,
      "step": 26939
    },
    {
      "epoch": 6.301719097181616,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 26940
    },
    {
      "epoch": 6.301952987954625,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 26941
    },
    {
      "epoch": 6.3021868787276345,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 26942
    },
    {
      "epoch": 6.302420769500643,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 26943
    },
    {
      "epoch": 6.302654660273652,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5644,
      "step": 26944
    },
    {
      "epoch": 6.302888551046661,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 26945
    },
    {
      "epoch": 6.30312244181967,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 26946
    },
    {
      "epoch": 6.30335633259268,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.6941,
      "step": 26947
    },
    {
      "epoch": 6.303590223365688,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 26948
    },
    {
      "epoch": 6.303824114138697,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5841,
      "step": 26949
    },
    {
      "epoch": 6.304058004911706,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 26950
    },
    {
      "epoch": 6.304291895684715,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 26951
    },
    {
      "epoch": 6.304525786457725,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 26952
    },
    {
      "epoch": 6.3047596772307335,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.585,
      "step": 26953
    },
    {
      "epoch": 6.304993568003742,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5245,
      "step": 26954
    },
    {
      "epoch": 6.305227458776751,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 26955
    },
    {
      "epoch": 6.30546134954976,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7432,
      "step": 26956
    },
    {
      "epoch": 6.30569524032277,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 26957
    },
    {
      "epoch": 6.305929131095779,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5935,
      "step": 26958
    },
    {
      "epoch": 6.306163021868787,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7802,
      "step": 26959
    },
    {
      "epoch": 6.306396912641796,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 26960
    },
    {
      "epoch": 6.306630803414805,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.6015,
      "step": 26961
    },
    {
      "epoch": 6.306864694187814,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 26962
    },
    {
      "epoch": 6.307098584960824,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 26963
    },
    {
      "epoch": 6.3073324757338325,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 26964
    },
    {
      "epoch": 6.307566366506841,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 26965
    },
    {
      "epoch": 6.30780025727985,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 26966
    },
    {
      "epoch": 6.308034148052859,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7284,
      "step": 26967
    },
    {
      "epoch": 6.308268038825869,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7155,
      "step": 26968
    },
    {
      "epoch": 6.308501929598878,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 26969
    },
    {
      "epoch": 6.308735820371886,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 26970
    },
    {
      "epoch": 6.308969711144895,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 26971
    },
    {
      "epoch": 6.309203601917904,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 26972
    },
    {
      "epoch": 6.309437492690913,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 26973
    },
    {
      "epoch": 6.309671383463923,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 26974
    },
    {
      "epoch": 6.3099052742369315,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 26975
    },
    {
      "epoch": 6.31013916500994,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9998,
      "step": 26976
    },
    {
      "epoch": 6.310373055782949,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5626,
      "step": 26977
    },
    {
      "epoch": 6.310606946555958,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 26978
    },
    {
      "epoch": 6.310840837328968,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6062,
      "step": 26979
    },
    {
      "epoch": 6.311074728101977,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 26980
    },
    {
      "epoch": 6.311308618874985,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.616,
      "step": 26981
    },
    {
      "epoch": 6.311542509647994,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 26982
    },
    {
      "epoch": 6.311776400421003,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 26983
    },
    {
      "epoch": 6.312010291194013,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 26984
    },
    {
      "epoch": 6.312244181967022,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7987,
      "step": 26985
    },
    {
      "epoch": 6.3124780727400305,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 26986
    },
    {
      "epoch": 6.312711963513039,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8532,
      "step": 26987
    },
    {
      "epoch": 6.312945854286048,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 26988
    },
    {
      "epoch": 6.313179745059058,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 26989
    },
    {
      "epoch": 6.313413635832067,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 26990
    },
    {
      "epoch": 6.3136475266050756,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7056,
      "step": 26991
    },
    {
      "epoch": 6.313881417378084,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7194,
      "step": 26992
    },
    {
      "epoch": 6.314115308151093,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 26993
    },
    {
      "epoch": 6.314349198924102,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 26994
    },
    {
      "epoch": 6.314583089697112,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 26995
    },
    {
      "epoch": 6.314816980470121,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 26996
    },
    {
      "epoch": 6.3150508712431295,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 26997
    },
    {
      "epoch": 6.315284762016138,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 26998
    },
    {
      "epoch": 6.315518652789147,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 26999
    },
    {
      "epoch": 6.315752543562157,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.775,
      "step": 27000
    },
    {
      "epoch": 6.315752543562157,
      "eval_runtime": 4.6156,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 27000
    },
    {
      "epoch": 6.315986434335166,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 27001
    },
    {
      "epoch": 6.3162203251081745,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 27002
    },
    {
      "epoch": 6.316454215881183,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 27003
    },
    {
      "epoch": 6.316688106654192,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 27004
    },
    {
      "epoch": 6.316921997427201,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 27005
    },
    {
      "epoch": 6.317155888200211,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6779,
      "step": 27006
    },
    {
      "epoch": 6.31738977897322,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7981,
      "step": 27007
    },
    {
      "epoch": 6.3176236697462285,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 27008
    },
    {
      "epoch": 6.317857560519237,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 27009
    },
    {
      "epoch": 6.318091451292246,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 27010
    },
    {
      "epoch": 6.318325342065256,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.586,
      "step": 27011
    },
    {
      "epoch": 6.318559232838265,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 27012
    },
    {
      "epoch": 6.3187931236112735,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 27013
    },
    {
      "epoch": 6.319027014384282,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7173,
      "step": 27014
    },
    {
      "epoch": 6.319260905157291,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 27015
    },
    {
      "epoch": 6.319494795930301,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5203,
      "step": 27016
    },
    {
      "epoch": 6.31972868670331,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6525,
      "step": 27017
    },
    {
      "epoch": 6.319962577476319,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6978,
      "step": 27018
    },
    {
      "epoch": 6.3201964682493275,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 27019
    },
    {
      "epoch": 6.320430359022336,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 27020
    },
    {
      "epoch": 6.320664249795346,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8711,
      "step": 27021
    },
    {
      "epoch": 6.320898140568355,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 27022
    },
    {
      "epoch": 6.321132031341364,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 27023
    },
    {
      "epoch": 6.3213659221143725,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6137,
      "step": 27024
    },
    {
      "epoch": 6.321599812887381,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 27025
    },
    {
      "epoch": 6.32183370366039,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 27026
    },
    {
      "epoch": 6.3220675944334,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 27027
    },
    {
      "epoch": 6.322301485206409,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 27028
    },
    {
      "epoch": 6.322535375979418,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8728,
      "step": 27029
    },
    {
      "epoch": 6.3227692667524265,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 27030
    },
    {
      "epoch": 6.323003157525435,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.4612,
      "step": 27031
    },
    {
      "epoch": 6.323237048298445,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 27032
    },
    {
      "epoch": 6.323470939071454,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 27033
    },
    {
      "epoch": 6.323704829844463,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 27034
    },
    {
      "epoch": 6.3239387206174715,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 27035
    },
    {
      "epoch": 6.32417261139048,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 27036
    },
    {
      "epoch": 6.324406502163489,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 27037
    },
    {
      "epoch": 6.324640392936499,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 27038
    },
    {
      "epoch": 6.324874283709508,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6746,
      "step": 27039
    },
    {
      "epoch": 6.325108174482517,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 27040
    },
    {
      "epoch": 6.3253420652555254,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 27041
    },
    {
      "epoch": 6.325575956028534,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 27042
    },
    {
      "epoch": 6.325809846801544,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6672,
      "step": 27043
    },
    {
      "epoch": 6.326043737574553,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 27044
    },
    {
      "epoch": 6.326277628347562,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 27045
    },
    {
      "epoch": 6.3265115191205705,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 27046
    },
    {
      "epoch": 6.326745409893579,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 27047
    },
    {
      "epoch": 6.326979300666589,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 27048
    },
    {
      "epoch": 6.327213191439598,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.678,
      "step": 27049
    },
    {
      "epoch": 6.327447082212607,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.4868,
      "step": 27050
    },
    {
      "epoch": 6.327680972985616,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7317,
      "step": 27051
    },
    {
      "epoch": 6.327914863758624,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 27052
    },
    {
      "epoch": 6.328148754531634,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 27053
    },
    {
      "epoch": 6.328382645304643,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 27054
    },
    {
      "epoch": 6.328616536077652,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 27055
    },
    {
      "epoch": 6.328850426850661,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 27056
    },
    {
      "epoch": 6.3290843176236695,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.579,
      "step": 27057
    },
    {
      "epoch": 6.329318208396678,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 27058
    },
    {
      "epoch": 6.329552099169688,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 27059
    },
    {
      "epoch": 6.329785989942697,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 27060
    },
    {
      "epoch": 6.330019880715706,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6783,
      "step": 27061
    },
    {
      "epoch": 6.330253771488715,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 27062
    },
    {
      "epoch": 6.330487662261723,
      "grad_norm": 10.0625,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 27063
    },
    {
      "epoch": 6.330721553034733,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.4742,
      "step": 27064
    },
    {
      "epoch": 6.330955443807742,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 27065
    },
    {
      "epoch": 6.331189334580751,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 27066
    },
    {
      "epoch": 6.33142322535376,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.4709,
      "step": 27067
    },
    {
      "epoch": 6.3316571161267685,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.4881,
      "step": 27068
    },
    {
      "epoch": 6.331891006899777,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.2926,
      "step": 27069
    },
    {
      "epoch": 6.332124897672787,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 27070
    },
    {
      "epoch": 6.332358788445796,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5209,
      "step": 27071
    },
    {
      "epoch": 6.332592679218805,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6631,
      "step": 27072
    },
    {
      "epoch": 6.332826569991814,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 27073
    },
    {
      "epoch": 6.333060460764822,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4531,
      "step": 27074
    },
    {
      "epoch": 6.333294351537832,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 27075
    },
    {
      "epoch": 6.333528242310841,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 27076
    },
    {
      "epoch": 6.33376213308385,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 27077
    },
    {
      "epoch": 6.333996023856859,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 27078
    },
    {
      "epoch": 6.3342299146298675,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 27079
    },
    {
      "epoch": 6.334463805402877,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6157,
      "step": 27080
    },
    {
      "epoch": 6.334697696175886,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6987,
      "step": 27081
    },
    {
      "epoch": 6.334931586948895,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5059,
      "step": 27082
    },
    {
      "epoch": 6.335165477721904,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 27083
    },
    {
      "epoch": 6.335399368494913,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 27084
    },
    {
      "epoch": 6.335633259267922,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6713,
      "step": 27085
    },
    {
      "epoch": 6.335867150040931,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 27086
    },
    {
      "epoch": 6.33610104081394,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 27087
    },
    {
      "epoch": 6.336334931586949,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 27088
    },
    {
      "epoch": 6.336568822359958,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.3746,
      "step": 27089
    },
    {
      "epoch": 6.3368027131329665,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6545,
      "step": 27090
    },
    {
      "epoch": 6.337036603905976,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 27091
    },
    {
      "epoch": 6.337270494678985,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7865,
      "step": 27092
    },
    {
      "epoch": 6.337504385451994,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7767,
      "step": 27093
    },
    {
      "epoch": 6.337738276225003,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1715,
      "step": 27094
    },
    {
      "epoch": 6.337972166998012,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6279,
      "step": 27095
    },
    {
      "epoch": 6.338206057771021,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 27096
    },
    {
      "epoch": 6.33843994854403,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 27097
    },
    {
      "epoch": 6.338673839317039,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 27098
    },
    {
      "epoch": 6.338907730090048,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 27099
    },
    {
      "epoch": 6.339141620863057,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 27100
    },
    {
      "epoch": 6.339141620863057,
      "eval_runtime": 4.5964,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 27100
    },
    {
      "epoch": 6.3393755116360655,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6243,
      "step": 27101
    },
    {
      "epoch": 6.339609402409075,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 27102
    },
    {
      "epoch": 6.339843293182084,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6326,
      "step": 27103
    },
    {
      "epoch": 6.340077183955093,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 27104
    },
    {
      "epoch": 6.340311074728102,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6558,
      "step": 27105
    },
    {
      "epoch": 6.340544965501111,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 27106
    },
    {
      "epoch": 6.34077885627412,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.4983,
      "step": 27107
    },
    {
      "epoch": 6.341012747047129,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 27108
    },
    {
      "epoch": 6.341246637820138,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 27109
    },
    {
      "epoch": 6.341480528593147,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 27110
    },
    {
      "epoch": 6.341714419366156,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 27111
    },
    {
      "epoch": 6.341948310139165,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 27112
    },
    {
      "epoch": 6.342182200912174,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6691,
      "step": 27113
    },
    {
      "epoch": 6.342416091685183,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 27114
    },
    {
      "epoch": 6.342649982458192,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 27115
    },
    {
      "epoch": 6.342883873231201,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 27116
    },
    {
      "epoch": 6.3431177640042105,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 27117
    },
    {
      "epoch": 6.343351654777219,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 27118
    },
    {
      "epoch": 6.343585545550228,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8252,
      "step": 27119
    },
    {
      "epoch": 6.343819436323237,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 27120
    },
    {
      "epoch": 6.344053327096246,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4727,
      "step": 27121
    },
    {
      "epoch": 6.344287217869255,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 27122
    },
    {
      "epoch": 6.344521108642264,
      "grad_norm": 7.40625,
      "learning_rate": 3e-05,
      "loss": 1.6973,
      "step": 27123
    },
    {
      "epoch": 6.344754999415273,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 27124
    },
    {
      "epoch": 6.344988890188282,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7292,
      "step": 27125
    },
    {
      "epoch": 6.345222780961291,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 27126
    },
    {
      "epoch": 6.3454566717343,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 27127
    },
    {
      "epoch": 6.3456905625073095,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 27128
    },
    {
      "epoch": 6.345924453280318,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4835,
      "step": 27129
    },
    {
      "epoch": 6.346158344053327,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.715,
      "step": 27130
    },
    {
      "epoch": 6.346392234826336,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6004,
      "step": 27131
    },
    {
      "epoch": 6.346626125599345,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 27132
    },
    {
      "epoch": 6.346860016372354,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.71,
      "step": 27133
    },
    {
      "epoch": 6.347093907145363,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 27134
    },
    {
      "epoch": 6.347327797918372,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 27135
    },
    {
      "epoch": 6.347561688691381,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 27136
    },
    {
      "epoch": 6.34779557946439,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5948,
      "step": 27137
    },
    {
      "epoch": 6.348029470237399,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6847,
      "step": 27138
    },
    {
      "epoch": 6.3482633610104084,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6467,
      "step": 27139
    },
    {
      "epoch": 6.348497251783417,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 27140
    },
    {
      "epoch": 6.348731142556426,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.4937,
      "step": 27141
    },
    {
      "epoch": 6.348965033329435,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5881,
      "step": 27142
    },
    {
      "epoch": 6.349198924102444,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 27143
    },
    {
      "epoch": 6.3494328148754535,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 27144
    },
    {
      "epoch": 6.349666705648462,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 27145
    },
    {
      "epoch": 6.349900596421471,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5346,
      "step": 27146
    },
    {
      "epoch": 6.35013448719448,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 27147
    },
    {
      "epoch": 6.350368377967489,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6648,
      "step": 27148
    },
    {
      "epoch": 6.350602268740499,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5921,
      "step": 27149
    },
    {
      "epoch": 6.350836159513507,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 27150
    },
    {
      "epoch": 6.351070050286516,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 27151
    },
    {
      "epoch": 6.351303941059525,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5806,
      "step": 27152
    },
    {
      "epoch": 6.351537831832534,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 27153
    },
    {
      "epoch": 6.351771722605543,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5225,
      "step": 27154
    },
    {
      "epoch": 6.3520056133785525,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 27155
    },
    {
      "epoch": 6.352239504151561,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 27156
    },
    {
      "epoch": 6.35247339492457,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6181,
      "step": 27157
    },
    {
      "epoch": 6.352707285697579,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5346,
      "step": 27158
    },
    {
      "epoch": 6.352941176470588,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 27159
    },
    {
      "epoch": 6.353175067243598,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5756,
      "step": 27160
    },
    {
      "epoch": 6.353408958016606,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5931,
      "step": 27161
    },
    {
      "epoch": 6.353642848789615,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 27162
    },
    {
      "epoch": 6.353876739562624,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 27163
    },
    {
      "epoch": 6.354110630335633,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 27164
    },
    {
      "epoch": 6.354344521108643,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5686,
      "step": 27165
    },
    {
      "epoch": 6.3545784118816515,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 27166
    },
    {
      "epoch": 6.35481230265466,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6118,
      "step": 27167
    },
    {
      "epoch": 6.355046193427669,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 27168
    },
    {
      "epoch": 6.355280084200678,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 27169
    },
    {
      "epoch": 6.355513974973688,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.708,
      "step": 27170
    },
    {
      "epoch": 6.355747865746697,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6286,
      "step": 27171
    },
    {
      "epoch": 6.355981756519705,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 27172
    },
    {
      "epoch": 6.356215647292714,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 27173
    },
    {
      "epoch": 6.356449538065723,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6261,
      "step": 27174
    },
    {
      "epoch": 6.356683428838732,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 27175
    },
    {
      "epoch": 6.356917319611742,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.5907,
      "step": 27176
    },
    {
      "epoch": 6.3571512103847505,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 27177
    },
    {
      "epoch": 6.357385101157759,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.6993,
      "step": 27178
    },
    {
      "epoch": 6.357618991930768,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 27179
    },
    {
      "epoch": 6.357852882703777,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 27180
    },
    {
      "epoch": 6.358086773476787,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 27181
    },
    {
      "epoch": 6.358320664249796,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 27182
    },
    {
      "epoch": 6.358554555022804,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 27183
    },
    {
      "epoch": 6.358788445795813,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2558,
      "step": 27184
    },
    {
      "epoch": 6.359022336568822,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 27185
    },
    {
      "epoch": 6.359256227341831,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7003,
      "step": 27186
    },
    {
      "epoch": 6.359490118114841,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 27187
    },
    {
      "epoch": 6.3597240088878495,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 27188
    },
    {
      "epoch": 6.359957899660858,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 27189
    },
    {
      "epoch": 6.360191790433867,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 27190
    },
    {
      "epoch": 6.360425681206876,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.564,
      "step": 27191
    },
    {
      "epoch": 6.360659571979886,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6647,
      "step": 27192
    },
    {
      "epoch": 6.360893462752895,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 27193
    },
    {
      "epoch": 6.361127353525903,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 27194
    },
    {
      "epoch": 6.361361244298912,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 27195
    },
    {
      "epoch": 6.361595135071921,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7662,
      "step": 27196
    },
    {
      "epoch": 6.361829025844931,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 27197
    },
    {
      "epoch": 6.36206291661794,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 27198
    },
    {
      "epoch": 6.3622968073909485,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 27199
    },
    {
      "epoch": 6.362530698163957,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 27200
    },
    {
      "epoch": 6.362530698163957,
      "eval_runtime": 4.6365,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 27200
    },
    {
      "epoch": 6.362764588936966,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.64,
      "step": 27201
    },
    {
      "epoch": 6.362998479709976,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 27202
    },
    {
      "epoch": 6.363232370482985,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 27203
    },
    {
      "epoch": 6.363466261255994,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 27204
    },
    {
      "epoch": 6.363700152029002,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 27205
    },
    {
      "epoch": 6.363934042802011,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 27206
    },
    {
      "epoch": 6.36416793357502,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6148,
      "step": 27207
    },
    {
      "epoch": 6.36440182434803,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 27208
    },
    {
      "epoch": 6.364635715121039,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 27209
    },
    {
      "epoch": 6.3648696058940475,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.612,
      "step": 27210
    },
    {
      "epoch": 6.365103496667056,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 27211
    },
    {
      "epoch": 6.365337387440065,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6458,
      "step": 27212
    },
    {
      "epoch": 6.365571278213075,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5358,
      "step": 27213
    },
    {
      "epoch": 6.365805168986084,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6632,
      "step": 27214
    },
    {
      "epoch": 6.366039059759093,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 27215
    },
    {
      "epoch": 6.366272950532101,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 27216
    },
    {
      "epoch": 6.36650684130511,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6722,
      "step": 27217
    },
    {
      "epoch": 6.366740732078119,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.628,
      "step": 27218
    },
    {
      "epoch": 6.366974622851129,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 27219
    },
    {
      "epoch": 6.367208513624138,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 27220
    },
    {
      "epoch": 6.3674424043971465,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.4926,
      "step": 27221
    },
    {
      "epoch": 6.367676295170155,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6236,
      "step": 27222
    },
    {
      "epoch": 6.367910185943164,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5846,
      "step": 27223
    },
    {
      "epoch": 6.368144076716174,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.696,
      "step": 27224
    },
    {
      "epoch": 6.368377967489183,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8416,
      "step": 27225
    },
    {
      "epoch": 6.368611858262192,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6515,
      "step": 27226
    },
    {
      "epoch": 6.3688457490352,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 27227
    },
    {
      "epoch": 6.369079639808209,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 27228
    },
    {
      "epoch": 6.369313530581219,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 27229
    },
    {
      "epoch": 6.369547421354228,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5357,
      "step": 27230
    },
    {
      "epoch": 6.369781312127237,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6381,
      "step": 27231
    },
    {
      "epoch": 6.3700152029002455,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 27232
    },
    {
      "epoch": 6.370249093673254,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 27233
    },
    {
      "epoch": 6.370482984446264,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6523,
      "step": 27234
    },
    {
      "epoch": 6.370716875219273,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 27235
    },
    {
      "epoch": 6.370950765992282,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 27236
    },
    {
      "epoch": 6.3711846567652906,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6444,
      "step": 27237
    },
    {
      "epoch": 6.371418547538299,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 27238
    },
    {
      "epoch": 6.371652438311308,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.719,
      "step": 27239
    },
    {
      "epoch": 6.371886329084318,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5147,
      "step": 27240
    },
    {
      "epoch": 6.372120219857327,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 27241
    },
    {
      "epoch": 6.372354110630336,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 27242
    },
    {
      "epoch": 6.3725880014033445,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.77,
      "step": 27243
    },
    {
      "epoch": 6.372821892176353,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 27244
    },
    {
      "epoch": 6.373055782949363,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 27245
    },
    {
      "epoch": 6.373289673722372,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 27246
    },
    {
      "epoch": 6.373523564495381,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 27247
    },
    {
      "epoch": 6.3737574552683895,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7526,
      "step": 27248
    },
    {
      "epoch": 6.373991346041398,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 27249
    },
    {
      "epoch": 6.374225236814407,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 27250
    },
    {
      "epoch": 6.374459127587417,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 27251
    },
    {
      "epoch": 6.374693018360426,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 27252
    },
    {
      "epoch": 6.374926909133435,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7758,
      "step": 27253
    },
    {
      "epoch": 6.3751607999064435,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 27254
    },
    {
      "epoch": 6.375394690679452,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 27255
    },
    {
      "epoch": 6.375628581452462,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 27256
    },
    {
      "epoch": 6.375862472225471,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 27257
    },
    {
      "epoch": 6.37609636299848,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 27258
    },
    {
      "epoch": 6.3763302537714885,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 27259
    },
    {
      "epoch": 6.376564144544497,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6026,
      "step": 27260
    },
    {
      "epoch": 6.376798035317507,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 27261
    },
    {
      "epoch": 6.377031926090516,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 27262
    },
    {
      "epoch": 6.377265816863525,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 27263
    },
    {
      "epoch": 6.377499707636534,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 27264
    },
    {
      "epoch": 6.3777335984095425,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6643,
      "step": 27265
    },
    {
      "epoch": 6.377967489182552,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 27266
    },
    {
      "epoch": 6.378201379955561,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 27267
    },
    {
      "epoch": 6.37843527072857,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5373,
      "step": 27268
    },
    {
      "epoch": 6.378669161501579,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 27269
    },
    {
      "epoch": 6.3789030522745875,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 27270
    },
    {
      "epoch": 6.379136943047596,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 27271
    },
    {
      "epoch": 6.379370833820606,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 27272
    },
    {
      "epoch": 6.379604724593615,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 27273
    },
    {
      "epoch": 6.379838615366624,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6566,
      "step": 27274
    },
    {
      "epoch": 6.380072506139633,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6504,
      "step": 27275
    },
    {
      "epoch": 6.3803063969126415,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 27276
    },
    {
      "epoch": 6.380540287685651,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7741,
      "step": 27277
    },
    {
      "epoch": 6.38077417845866,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 27278
    },
    {
      "epoch": 6.381008069231669,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6332,
      "step": 27279
    },
    {
      "epoch": 6.381241960004678,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 27280
    },
    {
      "epoch": 6.3814758507776865,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 27281
    },
    {
      "epoch": 6.381709741550695,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5971,
      "step": 27282
    },
    {
      "epoch": 6.381943632323705,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.5334,
      "step": 27283
    },
    {
      "epoch": 6.382177523096714,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 27284
    },
    {
      "epoch": 6.382411413869723,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 27285
    },
    {
      "epoch": 6.382645304642732,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 27286
    },
    {
      "epoch": 6.3828791954157404,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7356,
      "step": 27287
    },
    {
      "epoch": 6.38311308618875,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 27288
    },
    {
      "epoch": 6.383346976961759,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 27289
    },
    {
      "epoch": 6.383580867734768,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 27290
    },
    {
      "epoch": 6.383814758507777,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 27291
    },
    {
      "epoch": 6.3840486492807855,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7007,
      "step": 27292
    },
    {
      "epoch": 6.384282540053795,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 27293
    },
    {
      "epoch": 6.384516430826804,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 27294
    },
    {
      "epoch": 6.384750321599813,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 27295
    },
    {
      "epoch": 6.384984212372822,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 27296
    },
    {
      "epoch": 6.385218103145831,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 27297
    },
    {
      "epoch": 6.38545199391884,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 27298
    },
    {
      "epoch": 6.385685884691849,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 27299
    },
    {
      "epoch": 6.385919775464858,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6818,
      "step": 27300
    },
    {
      "epoch": 6.385919775464858,
      "eval_runtime": 4.5872,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 27300
    },
    {
      "epoch": 6.386153666237867,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6691,
      "step": 27301
    },
    {
      "epoch": 6.386387557010876,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 27302
    },
    {
      "epoch": 6.3866214477838845,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 27303
    },
    {
      "epoch": 6.386855338556894,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7074,
      "step": 27304
    },
    {
      "epoch": 6.387089229329903,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.609,
      "step": 27305
    },
    {
      "epoch": 6.387323120102912,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6624,
      "step": 27306
    },
    {
      "epoch": 6.387557010875921,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 27307
    },
    {
      "epoch": 6.38779090164893,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 27308
    },
    {
      "epoch": 6.388024792421939,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 27309
    },
    {
      "epoch": 6.388258683194948,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7023,
      "step": 27310
    },
    {
      "epoch": 6.388492573967957,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 27311
    },
    {
      "epoch": 6.388726464740966,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 27312
    },
    {
      "epoch": 6.388960355513975,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6159,
      "step": 27313
    },
    {
      "epoch": 6.3891942462869835,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 27314
    },
    {
      "epoch": 6.389428137059993,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 27315
    },
    {
      "epoch": 6.389662027833002,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 27316
    },
    {
      "epoch": 6.389895918606011,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 27317
    },
    {
      "epoch": 6.39012980937902,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 27318
    },
    {
      "epoch": 6.390363700152029,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6788,
      "step": 27319
    },
    {
      "epoch": 6.390597590925038,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 27320
    },
    {
      "epoch": 6.390831481698047,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 27321
    },
    {
      "epoch": 6.391065372471056,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 27322
    },
    {
      "epoch": 6.391299263244065,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 27323
    },
    {
      "epoch": 6.391533154017074,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 27324
    },
    {
      "epoch": 6.391767044790083,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 27325
    },
    {
      "epoch": 6.392000935563092,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6483,
      "step": 27326
    },
    {
      "epoch": 6.392234826336101,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 27327
    },
    {
      "epoch": 6.39246871710911,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 27328
    },
    {
      "epoch": 6.392702607882119,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 27329
    },
    {
      "epoch": 6.3929364986551285,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 27330
    },
    {
      "epoch": 6.393170389428137,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6108,
      "step": 27331
    },
    {
      "epoch": 6.393404280201146,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 27332
    },
    {
      "epoch": 6.393638170974155,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 27333
    },
    {
      "epoch": 6.393872061747164,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 27334
    },
    {
      "epoch": 6.394105952520173,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 27335
    },
    {
      "epoch": 6.394339843293182,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6696,
      "step": 27336
    },
    {
      "epoch": 6.394573734066191,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6727,
      "step": 27337
    },
    {
      "epoch": 6.3948076248392,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 27338
    },
    {
      "epoch": 6.395041515612209,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 27339
    },
    {
      "epoch": 6.395275406385218,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 27340
    },
    {
      "epoch": 6.3955092971582275,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7297,
      "step": 27341
    },
    {
      "epoch": 6.395743187931236,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 27342
    },
    {
      "epoch": 6.395977078704245,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 27343
    },
    {
      "epoch": 6.396210969477254,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7018,
      "step": 27344
    },
    {
      "epoch": 6.396444860250263,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 27345
    },
    {
      "epoch": 6.3966787510232725,
      "grad_norm": 17.0,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 27346
    },
    {
      "epoch": 6.396912641796281,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 27347
    },
    {
      "epoch": 6.39714653256929,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 27348
    },
    {
      "epoch": 6.397380423342299,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 27349
    },
    {
      "epoch": 6.397614314115308,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 27350
    },
    {
      "epoch": 6.397848204888318,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.581,
      "step": 27351
    },
    {
      "epoch": 6.3980820956613265,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 27352
    },
    {
      "epoch": 6.398315986434335,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 27353
    },
    {
      "epoch": 6.398549877207344,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7224,
      "step": 27354
    },
    {
      "epoch": 6.398783767980353,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 27355
    },
    {
      "epoch": 6.399017658753362,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7933,
      "step": 27356
    },
    {
      "epoch": 6.3992515495263715,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 27357
    },
    {
      "epoch": 6.39948544029938,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 27358
    },
    {
      "epoch": 6.399719331072389,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 27359
    },
    {
      "epoch": 6.399953221845398,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 27360
    },
    {
      "epoch": 6.400187112618407,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 27361
    },
    {
      "epoch": 6.400421003391417,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7589,
      "step": 27362
    },
    {
      "epoch": 6.4006548941644255,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 27363
    },
    {
      "epoch": 6.400888784937434,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 27364
    },
    {
      "epoch": 6.401122675710443,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7723,
      "step": 27365
    },
    {
      "epoch": 6.401356566483452,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 27366
    },
    {
      "epoch": 6.401590457256461,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 27367
    },
    {
      "epoch": 6.4018243480294705,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6642,
      "step": 27368
    },
    {
      "epoch": 6.402058238802479,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.742,
      "step": 27369
    },
    {
      "epoch": 6.402292129575488,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8577,
      "step": 27370
    },
    {
      "epoch": 6.402526020348497,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 27371
    },
    {
      "epoch": 6.402759911121506,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 27372
    },
    {
      "epoch": 6.402993801894516,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 27373
    },
    {
      "epoch": 6.4032276926675245,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 27374
    },
    {
      "epoch": 6.403461583440533,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 27375
    },
    {
      "epoch": 6.403695474213542,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 27376
    },
    {
      "epoch": 6.403929364986551,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5112,
      "step": 27377
    },
    {
      "epoch": 6.404163255759561,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5832,
      "step": 27378
    },
    {
      "epoch": 6.4043971465325695,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 27379
    },
    {
      "epoch": 6.404631037305578,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 27380
    },
    {
      "epoch": 6.404864928078587,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 27381
    },
    {
      "epoch": 6.405098818851596,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 27382
    },
    {
      "epoch": 6.405332709624606,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 27383
    },
    {
      "epoch": 6.405566600397615,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 27384
    },
    {
      "epoch": 6.4058004911706234,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 27385
    },
    {
      "epoch": 6.406034381943632,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 27386
    },
    {
      "epoch": 6.406268272716641,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 27387
    },
    {
      "epoch": 6.40650216348965,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.776,
      "step": 27388
    },
    {
      "epoch": 6.40673605426266,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 27389
    },
    {
      "epoch": 6.4069699450356685,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 27390
    },
    {
      "epoch": 6.407203835808677,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 27391
    },
    {
      "epoch": 6.407437726581686,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 27392
    },
    {
      "epoch": 6.407671617354695,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 27393
    },
    {
      "epoch": 6.407905508127705,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7502,
      "step": 27394
    },
    {
      "epoch": 6.408139398900714,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 27395
    },
    {
      "epoch": 6.408373289673722,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 27396
    },
    {
      "epoch": 6.408607180446731,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5352,
      "step": 27397
    },
    {
      "epoch": 6.40884107121974,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 27398
    },
    {
      "epoch": 6.409074961992749,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 27399
    },
    {
      "epoch": 6.409308852765759,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6454,
      "step": 27400
    },
    {
      "epoch": 6.409308852765759,
      "eval_runtime": 4.5944,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 27400
    },
    {
      "epoch": 6.4095427435387675,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9442,
      "step": 27401
    },
    {
      "epoch": 6.409776634311776,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.6264,
      "step": 27402
    },
    {
      "epoch": 6.410010525084785,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 27403
    },
    {
      "epoch": 6.410244415857794,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 27404
    },
    {
      "epoch": 6.410478306630804,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4927,
      "step": 27405
    },
    {
      "epoch": 6.410712197403813,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 27406
    },
    {
      "epoch": 6.410946088176821,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 27407
    },
    {
      "epoch": 6.41117997894983,
      "grad_norm": 8.4375,
      "learning_rate": 3e-05,
      "loss": 1.3964,
      "step": 27408
    },
    {
      "epoch": 6.411413869722839,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 27409
    },
    {
      "epoch": 6.411647760495849,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 27410
    },
    {
      "epoch": 6.411881651268858,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 27411
    },
    {
      "epoch": 6.4121155420418665,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6803,
      "step": 27412
    },
    {
      "epoch": 6.412349432814875,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 27413
    },
    {
      "epoch": 6.412583323587884,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 27414
    },
    {
      "epoch": 6.412817214360894,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.534,
      "step": 27415
    },
    {
      "epoch": 6.413051105133903,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 27416
    },
    {
      "epoch": 6.413284995906912,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5834,
      "step": 27417
    },
    {
      "epoch": 6.41351888667992,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 27418
    },
    {
      "epoch": 6.413752777452929,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.576,
      "step": 27419
    },
    {
      "epoch": 6.413986668225938,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 27420
    },
    {
      "epoch": 6.414220558998948,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 27421
    },
    {
      "epoch": 6.414454449771957,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6794,
      "step": 27422
    },
    {
      "epoch": 6.4146883405449655,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.727,
      "step": 27423
    },
    {
      "epoch": 6.414922231317974,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8132,
      "step": 27424
    },
    {
      "epoch": 6.415156122090983,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5799,
      "step": 27425
    },
    {
      "epoch": 6.415390012863993,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7587,
      "step": 27426
    },
    {
      "epoch": 6.415623903637002,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 27427
    },
    {
      "epoch": 6.415857794410011,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6201,
      "step": 27428
    },
    {
      "epoch": 6.416091685183019,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 27429
    },
    {
      "epoch": 6.416325575956028,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7184,
      "step": 27430
    },
    {
      "epoch": 6.416559466729037,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 27431
    },
    {
      "epoch": 6.416793357502047,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 27432
    },
    {
      "epoch": 6.417027248275056,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7074,
      "step": 27433
    },
    {
      "epoch": 6.4172611390480645,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 27434
    },
    {
      "epoch": 6.417495029821073,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 27435
    },
    {
      "epoch": 6.417728920594082,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2601,
      "step": 27436
    },
    {
      "epoch": 6.417962811367092,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5555,
      "step": 27437
    },
    {
      "epoch": 6.418196702140101,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 27438
    },
    {
      "epoch": 6.41843059291311,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 27439
    },
    {
      "epoch": 6.418664483686118,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 27440
    },
    {
      "epoch": 6.418898374459127,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6427,
      "step": 27441
    },
    {
      "epoch": 6.419132265232137,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.866,
      "step": 27442
    },
    {
      "epoch": 6.419366156005146,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 27443
    },
    {
      "epoch": 6.419600046778155,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 27444
    },
    {
      "epoch": 6.4198339375511635,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.667,
      "step": 27445
    },
    {
      "epoch": 6.420067828324172,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 27446
    },
    {
      "epoch": 6.420301719097182,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 27447
    },
    {
      "epoch": 6.420535609870191,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 27448
    },
    {
      "epoch": 6.4207695006432,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3645,
      "step": 27449
    },
    {
      "epoch": 6.421003391416209,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 27450
    },
    {
      "epoch": 6.421237282189217,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.575,
      "step": 27451
    },
    {
      "epoch": 6.421471172962226,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 27452
    },
    {
      "epoch": 6.421705063735236,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4595,
      "step": 27453
    },
    {
      "epoch": 6.421938954508245,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 27454
    },
    {
      "epoch": 6.422172845281254,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 27455
    },
    {
      "epoch": 6.4224067360542625,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 27456
    },
    {
      "epoch": 6.422640626827271,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 27457
    },
    {
      "epoch": 6.422874517600281,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 27458
    },
    {
      "epoch": 6.42310840837329,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 27459
    },
    {
      "epoch": 6.423342299146299,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 27460
    },
    {
      "epoch": 6.423576189919308,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 27461
    },
    {
      "epoch": 6.423810080692316,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 27462
    },
    {
      "epoch": 6.424043971465325,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 27463
    },
    {
      "epoch": 6.424277862238335,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 27464
    },
    {
      "epoch": 6.424511753011344,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.538,
      "step": 27465
    },
    {
      "epoch": 6.424745643784353,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 27466
    },
    {
      "epoch": 6.4249795345573615,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 27467
    },
    {
      "epoch": 6.42521342533037,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8151,
      "step": 27468
    },
    {
      "epoch": 6.42544731610338,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6269,
      "step": 27469
    },
    {
      "epoch": 6.425681206876389,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 27470
    },
    {
      "epoch": 6.425915097649398,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7552,
      "step": 27471
    },
    {
      "epoch": 6.426148988422407,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 27472
    },
    {
      "epoch": 6.426382879195415,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 27473
    },
    {
      "epoch": 6.426616769968425,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 27474
    },
    {
      "epoch": 6.426850660741434,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 27475
    },
    {
      "epoch": 6.427084551514443,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.2154,
      "step": 27476
    },
    {
      "epoch": 6.427318442287452,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.769,
      "step": 27477
    },
    {
      "epoch": 6.4275523330604605,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 27478
    },
    {
      "epoch": 6.42778622383347,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7745,
      "step": 27479
    },
    {
      "epoch": 6.428020114606479,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 27480
    },
    {
      "epoch": 6.428254005379488,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5555,
      "step": 27481
    },
    {
      "epoch": 6.428487896152497,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 27482
    },
    {
      "epoch": 6.4287217869255056,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 27483
    },
    {
      "epoch": 6.428955677698514,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7114,
      "step": 27484
    },
    {
      "epoch": 6.429189568471524,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 27485
    },
    {
      "epoch": 6.429423459244533,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 27486
    },
    {
      "epoch": 6.429657350017542,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.545,
      "step": 27487
    },
    {
      "epoch": 6.429891240790551,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 27488
    },
    {
      "epoch": 6.4301251315635595,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 27489
    },
    {
      "epoch": 6.430359022336569,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 27490
    },
    {
      "epoch": 6.430592913109578,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 27491
    },
    {
      "epoch": 6.430826803882587,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 27492
    },
    {
      "epoch": 6.431060694655596,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 27493
    },
    {
      "epoch": 6.4312945854286045,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 27494
    },
    {
      "epoch": 6.431528476201613,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 27495
    },
    {
      "epoch": 6.431762366974623,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 27496
    },
    {
      "epoch": 6.431996257747632,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5746,
      "step": 27497
    },
    {
      "epoch": 6.432230148520641,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 27498
    },
    {
      "epoch": 6.43246403929365,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6398,
      "step": 27499
    },
    {
      "epoch": 6.4326979300666585,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6768,
      "step": 27500
    },
    {
      "epoch": 6.4326979300666585,
      "eval_runtime": 4.5964,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 27500
    },
    {
      "epoch": 6.432931820839668,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 27501
    },
    {
      "epoch": 6.433165711612677,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 27502
    },
    {
      "epoch": 6.433399602385686,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 27503
    },
    {
      "epoch": 6.433633493158695,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6597,
      "step": 27504
    },
    {
      "epoch": 6.4338673839317035,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8303,
      "step": 27505
    },
    {
      "epoch": 6.434101274704713,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7156,
      "step": 27506
    },
    {
      "epoch": 6.434335165477722,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 27507
    },
    {
      "epoch": 6.434569056250731,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6829,
      "step": 27508
    },
    {
      "epoch": 6.43480294702374,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 27509
    },
    {
      "epoch": 6.435036837796749,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6417,
      "step": 27510
    },
    {
      "epoch": 6.435270728569758,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 27511
    },
    {
      "epoch": 6.435504619342767,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.6728,
      "step": 27512
    },
    {
      "epoch": 6.435738510115776,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.6158,
      "step": 27513
    },
    {
      "epoch": 6.435972400888785,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 27514
    },
    {
      "epoch": 6.436206291661794,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 27515
    },
    {
      "epoch": 6.4364401824348025,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 27516
    },
    {
      "epoch": 6.436674073207812,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6745,
      "step": 27517
    },
    {
      "epoch": 6.436907963980821,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7449,
      "step": 27518
    },
    {
      "epoch": 6.43714185475383,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.59,
      "step": 27519
    },
    {
      "epoch": 6.437375745526839,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 27520
    },
    {
      "epoch": 6.437609636299848,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 27521
    },
    {
      "epoch": 6.437843527072857,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7411,
      "step": 27522
    },
    {
      "epoch": 6.438077417845866,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 27523
    },
    {
      "epoch": 6.438311308618875,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 27524
    },
    {
      "epoch": 6.438545199391884,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 27525
    },
    {
      "epoch": 6.438779090164893,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 27526
    },
    {
      "epoch": 6.4390129809379015,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 27527
    },
    {
      "epoch": 6.439246871710911,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 27528
    },
    {
      "epoch": 6.43948076248392,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 27529
    },
    {
      "epoch": 6.439714653256929,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 27530
    },
    {
      "epoch": 6.439948544029938,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 27531
    },
    {
      "epoch": 6.440182434802947,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 27532
    },
    {
      "epoch": 6.440416325575956,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 27533
    },
    {
      "epoch": 6.440650216348965,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7153,
      "step": 27534
    },
    {
      "epoch": 6.440884107121974,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 27535
    },
    {
      "epoch": 6.441117997894983,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 27536
    },
    {
      "epoch": 6.441351888667992,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 27537
    },
    {
      "epoch": 6.441585779441001,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 27538
    },
    {
      "epoch": 6.44181967021401,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5129,
      "step": 27539
    },
    {
      "epoch": 6.442053560987019,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 27540
    },
    {
      "epoch": 6.442287451760028,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 27541
    },
    {
      "epoch": 6.442521342533037,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 27542
    },
    {
      "epoch": 6.4427552333060465,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 27543
    },
    {
      "epoch": 6.442989124079055,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 27544
    },
    {
      "epoch": 6.443223014852064,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6384,
      "step": 27545
    },
    {
      "epoch": 6.443456905625073,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 27546
    },
    {
      "epoch": 6.443690796398082,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7191,
      "step": 27547
    },
    {
      "epoch": 6.443924687171091,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.5172,
      "step": 27548
    },
    {
      "epoch": 6.4441585779441,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 27549
    },
    {
      "epoch": 6.444392468717109,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6775,
      "step": 27550
    },
    {
      "epoch": 6.444626359490118,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 27551
    },
    {
      "epoch": 6.444860250263127,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 27552
    },
    {
      "epoch": 6.445094141036136,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.4915,
      "step": 27553
    },
    {
      "epoch": 6.4453280318091455,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6681,
      "step": 27554
    },
    {
      "epoch": 6.445561922582154,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 27555
    },
    {
      "epoch": 6.445795813355163,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 27556
    },
    {
      "epoch": 6.446029704128172,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 27557
    },
    {
      "epoch": 6.446263594901181,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 27558
    },
    {
      "epoch": 6.446497485674191,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6693,
      "step": 27559
    },
    {
      "epoch": 6.446731376447199,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 27560
    },
    {
      "epoch": 6.446965267220208,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4683,
      "step": 27561
    },
    {
      "epoch": 6.447199157993217,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5759,
      "step": 27562
    },
    {
      "epoch": 6.447433048766226,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6609,
      "step": 27563
    },
    {
      "epoch": 6.447666939539236,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 27564
    },
    {
      "epoch": 6.4479008303122445,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 27565
    },
    {
      "epoch": 6.448134721085253,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7154,
      "step": 27566
    },
    {
      "epoch": 6.448368611858262,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6273,
      "step": 27567
    },
    {
      "epoch": 6.448602502631271,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5208,
      "step": 27568
    },
    {
      "epoch": 6.44883639340428,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 27569
    },
    {
      "epoch": 6.44907028417729,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 27570
    },
    {
      "epoch": 6.449304174950298,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6854,
      "step": 27571
    },
    {
      "epoch": 6.449538065723307,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 27572
    },
    {
      "epoch": 6.449771956496316,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 27573
    },
    {
      "epoch": 6.450005847269325,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 27574
    },
    {
      "epoch": 6.450239738042335,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 27575
    },
    {
      "epoch": 6.4504736288153435,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7242,
      "step": 27576
    },
    {
      "epoch": 6.450707519588352,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 27577
    },
    {
      "epoch": 6.450941410361361,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 27578
    },
    {
      "epoch": 6.45117530113437,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 27579
    },
    {
      "epoch": 6.451409191907379,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 27580
    },
    {
      "epoch": 6.4516430826803886,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 27581
    },
    {
      "epoch": 6.451876973453397,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 27582
    },
    {
      "epoch": 6.452110864226406,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6342,
      "step": 27583
    },
    {
      "epoch": 6.452344754999415,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6437,
      "step": 27584
    },
    {
      "epoch": 6.452578645772424,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2618,
      "step": 27585
    },
    {
      "epoch": 6.452812536545434,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 27586
    },
    {
      "epoch": 6.4530464273184425,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 27587
    },
    {
      "epoch": 6.453280318091451,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6125,
      "step": 27588
    },
    {
      "epoch": 6.45351420886446,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 27589
    },
    {
      "epoch": 6.453748099637469,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5396,
      "step": 27590
    },
    {
      "epoch": 6.453981990410479,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6269,
      "step": 27591
    },
    {
      "epoch": 6.4542158811834875,
      "grad_norm": 8.4375,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 27592
    },
    {
      "epoch": 6.454449771956496,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 27593
    },
    {
      "epoch": 6.454683662729505,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 27594
    },
    {
      "epoch": 6.454917553502514,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 27595
    },
    {
      "epoch": 6.455151444275524,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5204,
      "step": 27596
    },
    {
      "epoch": 6.455385335048533,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 27597
    },
    {
      "epoch": 6.4556192258215415,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7509,
      "step": 27598
    },
    {
      "epoch": 6.45585311659455,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7502,
      "step": 27599
    },
    {
      "epoch": 6.456087007367559,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3167,
      "step": 27600
    },
    {
      "epoch": 6.456087007367559,
      "eval_runtime": 4.6436,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 27600
    },
    {
      "epoch": 6.456320898140568,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7123,
      "step": 27601
    },
    {
      "epoch": 6.456554788913578,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 27602
    },
    {
      "epoch": 6.4567886796865865,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 27603
    },
    {
      "epoch": 6.457022570459595,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 27604
    },
    {
      "epoch": 6.457256461232604,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6115,
      "step": 27605
    },
    {
      "epoch": 6.457490352005613,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 27606
    },
    {
      "epoch": 6.457724242778623,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 27607
    },
    {
      "epoch": 6.457958133551632,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 27608
    },
    {
      "epoch": 6.4581920243246405,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6455,
      "step": 27609
    },
    {
      "epoch": 6.458425915097649,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 27610
    },
    {
      "epoch": 6.458659805870658,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.5304,
      "step": 27611
    },
    {
      "epoch": 6.458893696643667,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1635,
      "step": 27612
    },
    {
      "epoch": 6.459127587416677,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 27613
    },
    {
      "epoch": 6.4593614781896855,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6976,
      "step": 27614
    },
    {
      "epoch": 6.459595368962694,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.502,
      "step": 27615
    },
    {
      "epoch": 6.459829259735703,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6801,
      "step": 27616
    },
    {
      "epoch": 6.460063150508712,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7607,
      "step": 27617
    },
    {
      "epoch": 6.460297041281722,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 27618
    },
    {
      "epoch": 6.460530932054731,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6862,
      "step": 27619
    },
    {
      "epoch": 6.4607648228277395,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 27620
    },
    {
      "epoch": 6.460998713600748,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 27621
    },
    {
      "epoch": 6.461232604373757,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 27622
    },
    {
      "epoch": 6.461466495146767,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6292,
      "step": 27623
    },
    {
      "epoch": 6.461700385919776,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 27624
    },
    {
      "epoch": 6.4619342766927845,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 27625
    },
    {
      "epoch": 6.462168167465793,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 27626
    },
    {
      "epoch": 6.462402058238802,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.334,
      "step": 27627
    },
    {
      "epoch": 6.462635949011812,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 27628
    },
    {
      "epoch": 6.462869839784821,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 27629
    },
    {
      "epoch": 6.46310373055783,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0437,
      "step": 27630
    },
    {
      "epoch": 6.4633376213308384,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 27631
    },
    {
      "epoch": 6.463571512103847,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.5309,
      "step": 27632
    },
    {
      "epoch": 6.463805402876856,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3692,
      "step": 27633
    },
    {
      "epoch": 6.464039293649866,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 27634
    },
    {
      "epoch": 6.464273184422875,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6403,
      "step": 27635
    },
    {
      "epoch": 6.4645070751958835,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 27636
    },
    {
      "epoch": 6.464740965968892,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 27637
    },
    {
      "epoch": 6.464974856741901,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 27638
    },
    {
      "epoch": 6.465208747514911,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 27639
    },
    {
      "epoch": 6.46544263828792,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 27640
    },
    {
      "epoch": 6.465676529060929,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8318,
      "step": 27641
    },
    {
      "epoch": 6.465910419833937,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 27642
    },
    {
      "epoch": 6.466144310606946,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9262,
      "step": 27643
    },
    {
      "epoch": 6.466378201379955,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 27644
    },
    {
      "epoch": 6.466612092152965,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7642,
      "step": 27645
    },
    {
      "epoch": 6.466845982925974,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4601,
      "step": 27646
    },
    {
      "epoch": 6.4670798736989825,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 27647
    },
    {
      "epoch": 6.467313764471991,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 27648
    },
    {
      "epoch": 6.467547655245,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 27649
    },
    {
      "epoch": 6.46778154601801,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7485,
      "step": 27650
    },
    {
      "epoch": 6.468015436791019,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 27651
    },
    {
      "epoch": 6.468249327564028,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5599,
      "step": 27652
    },
    {
      "epoch": 6.468483218337036,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 27653
    },
    {
      "epoch": 6.468717109110045,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5378,
      "step": 27654
    },
    {
      "epoch": 6.468950999883055,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 27655
    },
    {
      "epoch": 6.469184890656064,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 27656
    },
    {
      "epoch": 6.469418781429073,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4377,
      "step": 27657
    },
    {
      "epoch": 6.4696526722020815,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 27658
    },
    {
      "epoch": 6.46988656297509,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5394,
      "step": 27659
    },
    {
      "epoch": 6.4701204537481,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7504,
      "step": 27660
    },
    {
      "epoch": 6.470354344521109,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 27661
    },
    {
      "epoch": 6.470588235294118,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6779,
      "step": 27662
    },
    {
      "epoch": 6.470822126067127,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6746,
      "step": 27663
    },
    {
      "epoch": 6.471056016840135,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 27664
    },
    {
      "epoch": 6.471289907613144,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 27665
    },
    {
      "epoch": 6.471523798386154,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5147,
      "step": 27666
    },
    {
      "epoch": 6.471757689159163,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 27667
    },
    {
      "epoch": 6.471991579932172,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 27668
    },
    {
      "epoch": 6.4722254707051805,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 27669
    },
    {
      "epoch": 6.472459361478189,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 27670
    },
    {
      "epoch": 6.472693252251199,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 27671
    },
    {
      "epoch": 6.472927143024208,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7794,
      "step": 27672
    },
    {
      "epoch": 6.473161033797217,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 27673
    },
    {
      "epoch": 6.473394924570226,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 27674
    },
    {
      "epoch": 6.473628815343234,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 27675
    },
    {
      "epoch": 6.473862706116243,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 27676
    },
    {
      "epoch": 6.474096596889253,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 27677
    },
    {
      "epoch": 6.474330487662262,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6594,
      "step": 27678
    },
    {
      "epoch": 6.474564378435271,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 27679
    },
    {
      "epoch": 6.4747982692082795,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6359,
      "step": 27680
    },
    {
      "epoch": 6.475032159981288,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 27681
    },
    {
      "epoch": 6.475266050754298,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9322,
      "step": 27682
    },
    {
      "epoch": 6.475499941527307,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8329,
      "step": 27683
    },
    {
      "epoch": 6.475733832300316,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 27684
    },
    {
      "epoch": 6.475967723073325,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1725,
      "step": 27685
    },
    {
      "epoch": 6.476201613846333,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 27686
    },
    {
      "epoch": 6.476435504619343,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 27687
    },
    {
      "epoch": 6.476669395392352,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 27688
    },
    {
      "epoch": 6.476903286165361,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 27689
    },
    {
      "epoch": 6.47713717693837,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 27690
    },
    {
      "epoch": 6.4773710677113785,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 27691
    },
    {
      "epoch": 6.477604958484388,
      "grad_norm": 9.125,
      "learning_rate": 3e-05,
      "loss": 2.1647,
      "step": 27692
    },
    {
      "epoch": 6.477838849257397,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6719,
      "step": 27693
    },
    {
      "epoch": 6.478072740030406,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6829,
      "step": 27694
    },
    {
      "epoch": 6.478306630803415,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 27695
    },
    {
      "epoch": 6.478540521576424,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 27696
    },
    {
      "epoch": 6.478774412349432,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 27697
    },
    {
      "epoch": 6.479008303122442,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 27698
    },
    {
      "epoch": 6.479242193895451,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.62,
      "step": 27699
    },
    {
      "epoch": 6.47947608466846,
      "grad_norm": 7.90625,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 27700
    },
    {
      "epoch": 6.47947608466846,
      "eval_runtime": 4.6234,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 27700
    },
    {
      "epoch": 6.479709975441469,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6648,
      "step": 27701
    },
    {
      "epoch": 6.4799438662144775,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 27702
    },
    {
      "epoch": 6.480177756987487,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6718,
      "step": 27703
    },
    {
      "epoch": 6.480411647760496,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7186,
      "step": 27704
    },
    {
      "epoch": 6.480645538533505,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 27705
    },
    {
      "epoch": 6.480879429306514,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6446,
      "step": 27706
    },
    {
      "epoch": 6.481113320079523,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6655,
      "step": 27707
    },
    {
      "epoch": 6.481347210852531,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7485,
      "step": 27708
    },
    {
      "epoch": 6.481581101625541,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7372,
      "step": 27709
    },
    {
      "epoch": 6.48181499239855,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 27710
    },
    {
      "epoch": 6.482048883171559,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5875,
      "step": 27711
    },
    {
      "epoch": 6.482282773944568,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6966,
      "step": 27712
    },
    {
      "epoch": 6.4825166647175765,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 27713
    },
    {
      "epoch": 6.482750555490586,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 27714
    },
    {
      "epoch": 6.482984446263595,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 27715
    },
    {
      "epoch": 6.483218337036604,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 27716
    },
    {
      "epoch": 6.483452227809613,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.3192,
      "step": 27717
    },
    {
      "epoch": 6.483686118582622,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 27718
    },
    {
      "epoch": 6.483920009355631,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 27719
    },
    {
      "epoch": 6.48415390012864,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 27720
    },
    {
      "epoch": 6.484387790901649,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.587,
      "step": 27721
    },
    {
      "epoch": 6.484621681674658,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6782,
      "step": 27722
    },
    {
      "epoch": 6.484855572447667,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 27723
    },
    {
      "epoch": 6.485089463220676,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 27724
    },
    {
      "epoch": 6.485323353993685,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5684,
      "step": 27725
    },
    {
      "epoch": 6.485557244766694,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3189,
      "step": 27726
    },
    {
      "epoch": 6.485791135539703,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7094,
      "step": 27727
    },
    {
      "epoch": 6.486025026312712,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7092,
      "step": 27728
    },
    {
      "epoch": 6.4862589170857206,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 27729
    },
    {
      "epoch": 6.48649280785873,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 27730
    },
    {
      "epoch": 6.486726698631739,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4763,
      "step": 27731
    },
    {
      "epoch": 6.486960589404748,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 27732
    },
    {
      "epoch": 6.487194480177757,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 27733
    },
    {
      "epoch": 6.487428370950766,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 27734
    },
    {
      "epoch": 6.487662261723775,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7994,
      "step": 27735
    },
    {
      "epoch": 6.487896152496784,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 27736
    },
    {
      "epoch": 6.488130043269793,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5005,
      "step": 27737
    },
    {
      "epoch": 6.488363934042802,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 27738
    },
    {
      "epoch": 6.488597824815811,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 27739
    },
    {
      "epoch": 6.48883171558882,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 27740
    },
    {
      "epoch": 6.489065606361829,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 27741
    },
    {
      "epoch": 6.489299497134838,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 27742
    },
    {
      "epoch": 6.489533387907847,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 27743
    },
    {
      "epoch": 6.489767278680856,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 27744
    },
    {
      "epoch": 6.4900011694538655,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8073,
      "step": 27745
    },
    {
      "epoch": 6.490235060226874,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.691,
      "step": 27746
    },
    {
      "epoch": 6.490468950999883,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 27747
    },
    {
      "epoch": 6.490702841772892,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 27748
    },
    {
      "epoch": 6.490936732545901,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 27749
    },
    {
      "epoch": 6.49117062331891,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 27750
    },
    {
      "epoch": 6.491404514091919,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 27751
    },
    {
      "epoch": 6.491638404864928,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 27752
    },
    {
      "epoch": 6.491872295637937,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 27753
    },
    {
      "epoch": 6.492106186410946,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 27754
    },
    {
      "epoch": 6.492340077183955,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4334,
      "step": 27755
    },
    {
      "epoch": 6.4925739679569645,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 27756
    },
    {
      "epoch": 6.492807858729973,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 27757
    },
    {
      "epoch": 6.493041749502982,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.701,
      "step": 27758
    },
    {
      "epoch": 6.493275640275991,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 27759
    },
    {
      "epoch": 6.493509531049,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 27760
    },
    {
      "epoch": 6.493743421822009,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4994,
      "step": 27761
    },
    {
      "epoch": 6.493977312595018,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7994,
      "step": 27762
    },
    {
      "epoch": 6.494211203368027,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5852,
      "step": 27763
    },
    {
      "epoch": 6.494445094141036,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 27764
    },
    {
      "epoch": 6.494678984914045,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 27765
    },
    {
      "epoch": 6.494912875687054,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6282,
      "step": 27766
    },
    {
      "epoch": 6.4951467664600635,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 27767
    },
    {
      "epoch": 6.495380657233072,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 27768
    },
    {
      "epoch": 6.495614548006081,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 27769
    },
    {
      "epoch": 6.49584843877909,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 27770
    },
    {
      "epoch": 6.496082329552099,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 27771
    },
    {
      "epoch": 6.496316220325109,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5418,
      "step": 27772
    },
    {
      "epoch": 6.496550111098117,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 27773
    },
    {
      "epoch": 6.496784001871126,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5044,
      "step": 27774
    },
    {
      "epoch": 6.497017892644135,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5306,
      "step": 27775
    },
    {
      "epoch": 6.497251783417144,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.5721,
      "step": 27776
    },
    {
      "epoch": 6.497485674190154,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 27777
    },
    {
      "epoch": 6.4977195649631625,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6233,
      "step": 27778
    },
    {
      "epoch": 6.497953455736171,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 27779
    },
    {
      "epoch": 6.49818734650918,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7269,
      "step": 27780
    },
    {
      "epoch": 6.498421237282189,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 27781
    },
    {
      "epoch": 6.498655128055198,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 27782
    },
    {
      "epoch": 6.498889018828208,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 27783
    },
    {
      "epoch": 6.499122909601216,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 27784
    },
    {
      "epoch": 6.499356800374225,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 27785
    },
    {
      "epoch": 6.499590691147234,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 27786
    },
    {
      "epoch": 6.499824581920243,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 27787
    },
    {
      "epoch": 6.500058472693253,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5036,
      "step": 27788
    },
    {
      "epoch": 6.5002923634662615,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 27789
    },
    {
      "epoch": 6.50052625423927,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 27790
    },
    {
      "epoch": 6.500760145012279,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 27791
    },
    {
      "epoch": 6.500994035785288,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 27792
    },
    {
      "epoch": 6.501227926558297,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 27793
    },
    {
      "epoch": 6.501461817331307,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.397,
      "step": 27794
    },
    {
      "epoch": 6.501695708104315,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6471,
      "step": 27795
    },
    {
      "epoch": 6.501929598877324,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 27796
    },
    {
      "epoch": 6.502163489650333,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7558,
      "step": 27797
    },
    {
      "epoch": 6.502397380423342,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 27798
    },
    {
      "epoch": 6.502631271196352,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6796,
      "step": 27799
    },
    {
      "epoch": 6.5028651619693605,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7509,
      "step": 27800
    },
    {
      "epoch": 6.5028651619693605,
      "eval_runtime": 4.6291,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 27800
    },
    {
      "epoch": 6.503099052742369,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 27801
    },
    {
      "epoch": 6.503332943515378,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5613,
      "step": 27802
    },
    {
      "epoch": 6.503566834288387,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.6351,
      "step": 27803
    },
    {
      "epoch": 6.503800725061396,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 27804
    },
    {
      "epoch": 6.504034615834406,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 27805
    },
    {
      "epoch": 6.504268506607414,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 27806
    },
    {
      "epoch": 6.504502397380423,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 27807
    },
    {
      "epoch": 6.504736288153432,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 27808
    },
    {
      "epoch": 6.504970178926442,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5466,
      "step": 27809
    },
    {
      "epoch": 6.505204069699451,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 27810
    },
    {
      "epoch": 6.5054379604724595,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7737,
      "step": 27811
    },
    {
      "epoch": 6.505671851245468,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 27812
    },
    {
      "epoch": 6.505905742018477,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6322,
      "step": 27813
    },
    {
      "epoch": 6.506139632791486,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 27814
    },
    {
      "epoch": 6.506373523564496,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6191,
      "step": 27815
    },
    {
      "epoch": 6.506607414337505,
      "grad_norm": 10.75,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 27816
    },
    {
      "epoch": 6.506841305110513,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 27817
    },
    {
      "epoch": 6.507075195883522,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 27818
    },
    {
      "epoch": 6.507309086656531,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6461,
      "step": 27819
    },
    {
      "epoch": 6.507542977429541,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 27820
    },
    {
      "epoch": 6.50777686820255,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 27821
    },
    {
      "epoch": 6.5080107589755585,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7327,
      "step": 27822
    },
    {
      "epoch": 6.508244649748567,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 27823
    },
    {
      "epoch": 6.508478540521576,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6563,
      "step": 27824
    },
    {
      "epoch": 6.508712431294585,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 27825
    },
    {
      "epoch": 6.508946322067595,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6202,
      "step": 27826
    },
    {
      "epoch": 6.5091802128406036,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 27827
    },
    {
      "epoch": 6.509414103613612,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6202,
      "step": 27828
    },
    {
      "epoch": 6.509647994386621,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 27829
    },
    {
      "epoch": 6.50988188515963,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 27830
    },
    {
      "epoch": 6.51011577593264,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 27831
    },
    {
      "epoch": 6.510349666705649,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 27832
    },
    {
      "epoch": 6.5105835574786575,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6602,
      "step": 27833
    },
    {
      "epoch": 6.510817448251666,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.464,
      "step": 27834
    },
    {
      "epoch": 6.511051339024675,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5267,
      "step": 27835
    },
    {
      "epoch": 6.511285229797684,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 1.4166,
      "step": 27836
    },
    {
      "epoch": 6.511519120570694,
      "grad_norm": 7.03125,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 27837
    },
    {
      "epoch": 6.5117530113437025,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 27838
    },
    {
      "epoch": 6.511986902116711,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 27839
    },
    {
      "epoch": 6.51222079288972,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.69,
      "step": 27840
    },
    {
      "epoch": 6.51245468366273,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 27841
    },
    {
      "epoch": 6.512688574435739,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5679,
      "step": 27842
    },
    {
      "epoch": 6.512922465208748,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7069,
      "step": 27843
    },
    {
      "epoch": 6.5131563559817565,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 27844
    },
    {
      "epoch": 6.513390246754765,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 27845
    },
    {
      "epoch": 6.513624137527774,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 27846
    },
    {
      "epoch": 6.513858028300784,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 27847
    },
    {
      "epoch": 6.514091919073793,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 27848
    },
    {
      "epoch": 6.5143258098468015,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6068,
      "step": 27849
    },
    {
      "epoch": 6.51455970061981,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 27850
    },
    {
      "epoch": 6.514793591392819,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7916,
      "step": 27851
    },
    {
      "epoch": 6.515027482165829,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6933,
      "step": 27852
    },
    {
      "epoch": 6.515261372938838,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7054,
      "step": 27853
    },
    {
      "epoch": 6.515495263711847,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 27854
    },
    {
      "epoch": 6.5157291544848555,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 27855
    },
    {
      "epoch": 6.515963045257864,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 27856
    },
    {
      "epoch": 6.516196936030873,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 27857
    },
    {
      "epoch": 6.516430826803883,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6512,
      "step": 27858
    },
    {
      "epoch": 6.516664717576892,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 27859
    },
    {
      "epoch": 6.5168986083499005,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 27860
    },
    {
      "epoch": 6.517132499122909,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 27861
    },
    {
      "epoch": 6.517366389895919,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5444,
      "step": 27862
    },
    {
      "epoch": 6.517600280668928,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 27863
    },
    {
      "epoch": 6.517834171441937,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 27864
    },
    {
      "epoch": 6.518068062214946,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 27865
    },
    {
      "epoch": 6.5183019529879545,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 27866
    },
    {
      "epoch": 6.518535843760963,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 27867
    },
    {
      "epoch": 6.518769734533973,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6932,
      "step": 27868
    },
    {
      "epoch": 6.519003625306982,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 27869
    },
    {
      "epoch": 6.519237516079991,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4594,
      "step": 27870
    },
    {
      "epoch": 6.5194714068529995,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.603,
      "step": 27871
    },
    {
      "epoch": 6.519705297626008,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 27872
    },
    {
      "epoch": 6.519939188399018,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 27873
    },
    {
      "epoch": 6.520173079172027,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 27874
    },
    {
      "epoch": 6.520406969945036,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 27875
    },
    {
      "epoch": 6.520640860718045,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 27876
    },
    {
      "epoch": 6.5208747514910534,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 27877
    },
    {
      "epoch": 6.521108642264062,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 27878
    },
    {
      "epoch": 6.521342533037072,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 27879
    },
    {
      "epoch": 6.521576423810081,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 27880
    },
    {
      "epoch": 6.52181031458309,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 27881
    },
    {
      "epoch": 6.5220442053560985,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 27882
    },
    {
      "epoch": 6.522278096129107,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 27883
    },
    {
      "epoch": 6.522511986902117,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 27884
    },
    {
      "epoch": 6.522745877675126,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 27885
    },
    {
      "epoch": 6.522979768448135,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 27886
    },
    {
      "epoch": 6.523213659221144,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 27887
    },
    {
      "epoch": 6.523447549994152,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.726,
      "step": 27888
    },
    {
      "epoch": 6.523681440767161,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 27889
    },
    {
      "epoch": 6.523915331540171,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 27890
    },
    {
      "epoch": 6.52414922231318,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.541,
      "step": 27891
    },
    {
      "epoch": 6.524383113086189,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 27892
    },
    {
      "epoch": 6.5246170038591975,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 27893
    },
    {
      "epoch": 6.524850894632207,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 27894
    },
    {
      "epoch": 6.525084785405216,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 27895
    },
    {
      "epoch": 6.525318676178225,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 27896
    },
    {
      "epoch": 6.525552566951234,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 27897
    },
    {
      "epoch": 6.525786457724243,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 27898
    },
    {
      "epoch": 6.526020348497251,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 27899
    },
    {
      "epoch": 6.526254239270261,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6318,
      "step": 27900
    },
    {
      "epoch": 6.526254239270261,
      "eval_runtime": 4.6574,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 27900
    },
    {
      "epoch": 6.52648813004327,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 27901
    },
    {
      "epoch": 6.526722020816279,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 27902
    },
    {
      "epoch": 6.526955911589288,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6952,
      "step": 27903
    },
    {
      "epoch": 6.5271898023622965,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 27904
    },
    {
      "epoch": 6.527423693135306,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 27905
    },
    {
      "epoch": 6.527657583908315,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 27906
    },
    {
      "epoch": 6.527891474681324,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 27907
    },
    {
      "epoch": 6.528125365454333,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.547,
      "step": 27908
    },
    {
      "epoch": 6.528359256227342,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 27909
    },
    {
      "epoch": 6.52859314700035,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 27910
    },
    {
      "epoch": 6.52882703777336,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 27911
    },
    {
      "epoch": 6.529060928546369,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 27912
    },
    {
      "epoch": 6.529294819319378,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 27913
    },
    {
      "epoch": 6.529528710092387,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 27914
    },
    {
      "epoch": 6.5297626008653955,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 27915
    },
    {
      "epoch": 6.529996491638405,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8645,
      "step": 27916
    },
    {
      "epoch": 6.530230382411414,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 27917
    },
    {
      "epoch": 6.530464273184423,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.542,
      "step": 27918
    },
    {
      "epoch": 6.530698163957432,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 27919
    },
    {
      "epoch": 6.530932054730441,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 27920
    },
    {
      "epoch": 6.531165945503449,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 27921
    },
    {
      "epoch": 6.531399836276459,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8415,
      "step": 27922
    },
    {
      "epoch": 6.531633727049468,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 27923
    },
    {
      "epoch": 6.531867617822477,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6584,
      "step": 27924
    },
    {
      "epoch": 6.532101508595486,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 27925
    },
    {
      "epoch": 6.532335399368495,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7687,
      "step": 27926
    },
    {
      "epoch": 6.532569290141504,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 27927
    },
    {
      "epoch": 6.532803180914513,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 27928
    },
    {
      "epoch": 6.533037071687522,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5992,
      "step": 27929
    },
    {
      "epoch": 6.533270962460531,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 27930
    },
    {
      "epoch": 6.53350485323354,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.672,
      "step": 27931
    },
    {
      "epoch": 6.533738744006549,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 27932
    },
    {
      "epoch": 6.533972634779558,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.649,
      "step": 27933
    },
    {
      "epoch": 6.534206525552567,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7042,
      "step": 27934
    },
    {
      "epoch": 6.534440416325576,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6898,
      "step": 27935
    },
    {
      "epoch": 6.534674307098585,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.701,
      "step": 27936
    },
    {
      "epoch": 6.534908197871594,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 27937
    },
    {
      "epoch": 6.535142088644603,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7339,
      "step": 27938
    },
    {
      "epoch": 6.535375979417612,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 27939
    },
    {
      "epoch": 6.535609870190621,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 27940
    },
    {
      "epoch": 6.53584376096363,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 27941
    },
    {
      "epoch": 6.536077651736639,
      "grad_norm": 2.28125,
      "learning_rate": 3e-05,
      "loss": 1.3454,
      "step": 27942
    },
    {
      "epoch": 6.536311542509648,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6614,
      "step": 27943
    },
    {
      "epoch": 6.536545433282657,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 27944
    },
    {
      "epoch": 6.536779324055666,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 27945
    },
    {
      "epoch": 6.537013214828675,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6419,
      "step": 27946
    },
    {
      "epoch": 6.537247105601684,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 27947
    },
    {
      "epoch": 6.537480996374693,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7644,
      "step": 27948
    },
    {
      "epoch": 6.537714887147702,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 27949
    },
    {
      "epoch": 6.537948777920711,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7827,
      "step": 27950
    },
    {
      "epoch": 6.53818266869372,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 27951
    },
    {
      "epoch": 6.538416559466729,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5351,
      "step": 27952
    },
    {
      "epoch": 6.538650450239738,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 27953
    },
    {
      "epoch": 6.538884341012747,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 27954
    },
    {
      "epoch": 6.539118231785756,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 27955
    },
    {
      "epoch": 6.539352122558765,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 27956
    },
    {
      "epoch": 6.539586013331774,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 27957
    },
    {
      "epoch": 6.5398199041047835,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 27958
    },
    {
      "epoch": 6.540053794877792,
      "grad_norm": 9.9375,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 27959
    },
    {
      "epoch": 6.540287685650801,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 27960
    },
    {
      "epoch": 6.54052157642381,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.482,
      "step": 27961
    },
    {
      "epoch": 6.540755467196819,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 27962
    },
    {
      "epoch": 6.540989357969828,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6919,
      "step": 27963
    },
    {
      "epoch": 6.5412232487428374,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 27964
    },
    {
      "epoch": 6.541457139515846,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2044,
      "step": 27965
    },
    {
      "epoch": 6.541691030288855,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 27966
    },
    {
      "epoch": 6.541924921061864,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7134,
      "step": 27967
    },
    {
      "epoch": 6.542158811834873,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 27968
    },
    {
      "epoch": 6.5423927026078825,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 27969
    },
    {
      "epoch": 6.542626593380891,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9504,
      "step": 27970
    },
    {
      "epoch": 6.5428604841539,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.5384,
      "step": 27971
    },
    {
      "epoch": 6.543094374926909,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6371,
      "step": 27972
    },
    {
      "epoch": 6.543328265699918,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 27973
    },
    {
      "epoch": 6.543562156472927,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6128,
      "step": 27974
    },
    {
      "epoch": 6.543796047245936,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5615,
      "step": 27975
    },
    {
      "epoch": 6.544029938018945,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6927,
      "step": 27976
    },
    {
      "epoch": 6.544263828791954,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 27977
    },
    {
      "epoch": 6.544497719564963,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 27978
    },
    {
      "epoch": 6.544731610337972,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7818,
      "step": 27979
    },
    {
      "epoch": 6.5449655011109815,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 27980
    },
    {
      "epoch": 6.54519939188399,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 27981
    },
    {
      "epoch": 6.545433282656999,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 27982
    },
    {
      "epoch": 6.545667173430008,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 27983
    },
    {
      "epoch": 6.545901064203017,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 27984
    },
    {
      "epoch": 6.546134954976026,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7384,
      "step": 27985
    },
    {
      "epoch": 6.546368845749035,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7337,
      "step": 27986
    },
    {
      "epoch": 6.546602736522044,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 27987
    },
    {
      "epoch": 6.546836627295053,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 27988
    },
    {
      "epoch": 6.547070518068062,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 27989
    },
    {
      "epoch": 6.547304408841072,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0032,
      "step": 27990
    },
    {
      "epoch": 6.5475382996140805,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 27991
    },
    {
      "epoch": 6.547772190387089,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 27992
    },
    {
      "epoch": 6.548006081160098,
      "grad_norm": 13.625,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 27993
    },
    {
      "epoch": 6.548239971933107,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.518,
      "step": 27994
    },
    {
      "epoch": 6.548473862706116,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 27995
    },
    {
      "epoch": 6.548707753479126,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6221,
      "step": 27996
    },
    {
      "epoch": 6.548941644252134,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 27997
    },
    {
      "epoch": 6.549175535025143,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 27998
    },
    {
      "epoch": 6.549409425798152,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6683,
      "step": 27999
    },
    {
      "epoch": 6.549643316571161,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 28000
    },
    {
      "epoch": 6.549643316571161,
      "eval_runtime": 4.6019,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 28000
    },
    {
      "epoch": 6.549877207344171,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 28001
    },
    {
      "epoch": 6.5501110981171795,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 28002
    },
    {
      "epoch": 6.550344988890188,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 28003
    },
    {
      "epoch": 6.550578879663197,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5126,
      "step": 28004
    },
    {
      "epoch": 6.550812770436206,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6054,
      "step": 28005
    },
    {
      "epoch": 6.551046661209215,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 28006
    },
    {
      "epoch": 6.551280551982225,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 28007
    },
    {
      "epoch": 6.551514442755233,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 28008
    },
    {
      "epoch": 6.551748333528242,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 28009
    },
    {
      "epoch": 6.551982224301251,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 28010
    },
    {
      "epoch": 6.55221611507426,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 28011
    },
    {
      "epoch": 6.55245000584727,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 28012
    },
    {
      "epoch": 6.5526838966202785,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 28013
    },
    {
      "epoch": 6.552917787393287,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 28014
    },
    {
      "epoch": 6.553151678166296,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 28015
    },
    {
      "epoch": 6.553385568939305,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7426,
      "step": 28016
    },
    {
      "epoch": 6.553619459712314,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 28017
    },
    {
      "epoch": 6.553853350485324,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7268,
      "step": 28018
    },
    {
      "epoch": 6.554087241258332,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7083,
      "step": 28019
    },
    {
      "epoch": 6.554321132031341,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 28020
    },
    {
      "epoch": 6.55455502280435,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8748,
      "step": 28021
    },
    {
      "epoch": 6.55478891357736,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 28022
    },
    {
      "epoch": 6.555022804350369,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7391,
      "step": 28023
    },
    {
      "epoch": 6.5552566951233775,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 28024
    },
    {
      "epoch": 6.555490585896386,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5503,
      "step": 28025
    },
    {
      "epoch": 6.555724476669395,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 28026
    },
    {
      "epoch": 6.555958367442404,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6054,
      "step": 28027
    },
    {
      "epoch": 6.556192258215414,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8234,
      "step": 28028
    },
    {
      "epoch": 6.556426148988423,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 28029
    },
    {
      "epoch": 6.556660039761431,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 28030
    },
    {
      "epoch": 6.55689393053444,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 28031
    },
    {
      "epoch": 6.557127821307449,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7224,
      "step": 28032
    },
    {
      "epoch": 6.557361712080459,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 28033
    },
    {
      "epoch": 6.557595602853468,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.851,
      "step": 28034
    },
    {
      "epoch": 6.5578294936264765,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 28035
    },
    {
      "epoch": 6.558063384399485,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 28036
    },
    {
      "epoch": 6.558297275172494,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8503,
      "step": 28037
    },
    {
      "epoch": 6.558531165945503,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5979,
      "step": 28038
    },
    {
      "epoch": 6.558765056718513,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6966,
      "step": 28039
    },
    {
      "epoch": 6.558998947491522,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 28040
    },
    {
      "epoch": 6.55923283826453,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 28041
    },
    {
      "epoch": 6.559466729037539,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 28042
    },
    {
      "epoch": 6.559700619810549,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 28043
    },
    {
      "epoch": 6.559934510583558,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6997,
      "step": 28044
    },
    {
      "epoch": 6.560168401356567,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 28045
    },
    {
      "epoch": 6.5604022921295755,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 28046
    },
    {
      "epoch": 6.560636182902584,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7597,
      "step": 28047
    },
    {
      "epoch": 6.560870073675593,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 28048
    },
    {
      "epoch": 6.561103964448603,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 28049
    },
    {
      "epoch": 6.561337855221612,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 28050
    },
    {
      "epoch": 6.561571745994621,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 28051
    },
    {
      "epoch": 6.561805636767629,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 28052
    },
    {
      "epoch": 6.562039527540638,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4754,
      "step": 28053
    },
    {
      "epoch": 6.562273418313648,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 28054
    },
    {
      "epoch": 6.562507309086657,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 28055
    },
    {
      "epoch": 6.562741199859666,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 28056
    },
    {
      "epoch": 6.5629750906326745,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.6469,
      "step": 28057
    },
    {
      "epoch": 6.563208981405683,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6153,
      "step": 28058
    },
    {
      "epoch": 6.563442872178692,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7281,
      "step": 28059
    },
    {
      "epoch": 6.563676762951702,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 28060
    },
    {
      "epoch": 6.563910653724711,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6889,
      "step": 28061
    },
    {
      "epoch": 6.56414454449772,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5497,
      "step": 28062
    },
    {
      "epoch": 6.564378435270728,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 28063
    },
    {
      "epoch": 6.564612326043737,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 28064
    },
    {
      "epoch": 6.564846216816747,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6792,
      "step": 28065
    },
    {
      "epoch": 6.565080107589756,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 28066
    },
    {
      "epoch": 6.565313998362765,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 28067
    },
    {
      "epoch": 6.5655478891357735,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.634,
      "step": 28068
    },
    {
      "epoch": 6.565781779908782,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 28069
    },
    {
      "epoch": 6.566015670681791,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 28070
    },
    {
      "epoch": 6.566249561454801,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 28071
    },
    {
      "epoch": 6.56648345222781,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 28072
    },
    {
      "epoch": 6.5667173430008186,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 28073
    },
    {
      "epoch": 6.566951233773827,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6918,
      "step": 28074
    },
    {
      "epoch": 6.567185124546837,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 28075
    },
    {
      "epoch": 6.567419015319846,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6747,
      "step": 28076
    },
    {
      "epoch": 6.567652906092855,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 28077
    },
    {
      "epoch": 6.567886796865864,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 28078
    },
    {
      "epoch": 6.5681206876388725,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6589,
      "step": 28079
    },
    {
      "epoch": 6.568354578411881,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 28080
    },
    {
      "epoch": 6.568588469184891,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 28081
    },
    {
      "epoch": 6.5688223599579,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 28082
    },
    {
      "epoch": 6.569056250730909,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 28083
    },
    {
      "epoch": 6.5692901415039175,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 28084
    },
    {
      "epoch": 6.569524032276926,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.661,
      "step": 28085
    },
    {
      "epoch": 6.569757923049936,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 28086
    },
    {
      "epoch": 6.569991813822945,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5037,
      "step": 28087
    },
    {
      "epoch": 6.570225704595954,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5028,
      "step": 28088
    },
    {
      "epoch": 6.570459595368963,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 28089
    },
    {
      "epoch": 6.5706934861419715,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6869,
      "step": 28090
    },
    {
      "epoch": 6.57092737691498,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 28091
    },
    {
      "epoch": 6.57116126768799,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 28092
    },
    {
      "epoch": 6.571395158460999,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.5427,
      "step": 28093
    },
    {
      "epoch": 6.571629049234008,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 28094
    },
    {
      "epoch": 6.5718629400070165,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5714,
      "step": 28095
    },
    {
      "epoch": 6.572096830780025,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 28096
    },
    {
      "epoch": 6.572330721553035,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 28097
    },
    {
      "epoch": 6.572564612326044,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 28098
    },
    {
      "epoch": 6.572798503099053,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6595,
      "step": 28099
    },
    {
      "epoch": 6.573032393872062,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6221,
      "step": 28100
    },
    {
      "epoch": 6.573032393872062,
      "eval_runtime": 4.7038,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 28100
    },
    {
      "epoch": 6.5732662846450705,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7838,
      "step": 28101
    },
    {
      "epoch": 6.573500175418079,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6267,
      "step": 28102
    },
    {
      "epoch": 6.573734066191089,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 28103
    },
    {
      "epoch": 6.573967956964098,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 28104
    },
    {
      "epoch": 6.574201847737107,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 28105
    },
    {
      "epoch": 6.5744357385101155,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6729,
      "step": 28106
    },
    {
      "epoch": 6.574669629283125,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 28107
    },
    {
      "epoch": 6.574903520056134,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.5872,
      "step": 28108
    },
    {
      "epoch": 6.575137410829143,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 28109
    },
    {
      "epoch": 6.575371301602152,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.5308,
      "step": 28110
    },
    {
      "epoch": 6.575605192375161,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 28111
    },
    {
      "epoch": 6.5758390831481695,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.629,
      "step": 28112
    },
    {
      "epoch": 6.576072973921179,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.474,
      "step": 28113
    },
    {
      "epoch": 6.576306864694188,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5255,
      "step": 28114
    },
    {
      "epoch": 6.576540755467197,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 28115
    },
    {
      "epoch": 6.576774646240206,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 28116
    },
    {
      "epoch": 6.5770085370132145,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5271,
      "step": 28117
    },
    {
      "epoch": 6.577242427786224,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 28118
    },
    {
      "epoch": 6.577476318559233,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 28119
    },
    {
      "epoch": 6.577710209332242,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 28120
    },
    {
      "epoch": 6.577944100105251,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 28121
    },
    {
      "epoch": 6.57817799087826,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 28122
    },
    {
      "epoch": 6.5784118816512684,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 28123
    },
    {
      "epoch": 6.578645772424278,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 28124
    },
    {
      "epoch": 6.578879663197287,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 28125
    },
    {
      "epoch": 6.579113553970296,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 28126
    },
    {
      "epoch": 6.579347444743305,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8411,
      "step": 28127
    },
    {
      "epoch": 6.5795813355163135,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7609,
      "step": 28128
    },
    {
      "epoch": 6.579815226289323,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 28129
    },
    {
      "epoch": 6.580049117062332,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 28130
    },
    {
      "epoch": 6.580283007835341,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 28131
    },
    {
      "epoch": 6.58051689860835,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6361,
      "step": 28132
    },
    {
      "epoch": 6.580750789381359,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6246,
      "step": 28133
    },
    {
      "epoch": 6.580984680154367,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 28134
    },
    {
      "epoch": 6.581218570927377,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 28135
    },
    {
      "epoch": 6.581452461700386,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5449,
      "step": 28136
    },
    {
      "epoch": 6.581686352473395,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 28137
    },
    {
      "epoch": 6.581920243246404,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 28138
    },
    {
      "epoch": 6.582154134019413,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 28139
    },
    {
      "epoch": 6.582388024792422,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6828,
      "step": 28140
    },
    {
      "epoch": 6.582621915565431,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8177,
      "step": 28141
    },
    {
      "epoch": 6.58285580633844,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 28142
    },
    {
      "epoch": 6.583089697111449,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 28143
    },
    {
      "epoch": 6.583323587884458,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 28144
    },
    {
      "epoch": 6.583557478657467,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 28145
    },
    {
      "epoch": 6.583791369430476,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 28146
    },
    {
      "epoch": 6.584025260203485,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 28147
    },
    {
      "epoch": 6.584259150976494,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6369,
      "step": 28148
    },
    {
      "epoch": 6.584493041749503,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 28149
    },
    {
      "epoch": 6.584726932522512,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 28150
    },
    {
      "epoch": 6.584960823295521,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5448,
      "step": 28151
    },
    {
      "epoch": 6.58519471406853,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 28152
    },
    {
      "epoch": 6.585428604841539,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 28153
    },
    {
      "epoch": 6.585662495614548,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 28154
    },
    {
      "epoch": 6.585896386387557,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 28155
    },
    {
      "epoch": 6.586130277160566,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.678,
      "step": 28156
    },
    {
      "epoch": 6.586364167933575,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 28157
    },
    {
      "epoch": 6.586598058706584,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6367,
      "step": 28158
    },
    {
      "epoch": 6.586831949479593,
      "grad_norm": 10.0625,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 28159
    },
    {
      "epoch": 6.587065840252602,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 28160
    },
    {
      "epoch": 6.587299731025611,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.729,
      "step": 28161
    },
    {
      "epoch": 6.58753362179862,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6728,
      "step": 28162
    },
    {
      "epoch": 6.587767512571629,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 28163
    },
    {
      "epoch": 6.588001403344638,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.5636,
      "step": 28164
    },
    {
      "epoch": 6.588235294117647,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 28165
    },
    {
      "epoch": 6.588469184890656,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 28166
    },
    {
      "epoch": 6.588703075663665,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7187,
      "step": 28167
    },
    {
      "epoch": 6.588936966436674,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7988,
      "step": 28168
    },
    {
      "epoch": 6.589170857209683,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6288,
      "step": 28169
    },
    {
      "epoch": 6.589404747982692,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 28170
    },
    {
      "epoch": 6.5896386387557015,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 28171
    },
    {
      "epoch": 6.58987252952871,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8047,
      "step": 28172
    },
    {
      "epoch": 6.590106420301719,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 28173
    },
    {
      "epoch": 6.590340311074728,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5884,
      "step": 28174
    },
    {
      "epoch": 6.590574201847737,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 28175
    },
    {
      "epoch": 6.590808092620746,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 28176
    },
    {
      "epoch": 6.5910419833937555,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 28177
    },
    {
      "epoch": 6.591275874166764,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9133,
      "step": 28178
    },
    {
      "epoch": 6.591509764939773,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 28179
    },
    {
      "epoch": 6.591743655712782,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 28180
    },
    {
      "epoch": 6.591977546485791,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 28181
    },
    {
      "epoch": 6.5922114372588005,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 28182
    },
    {
      "epoch": 6.592445328031809,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 28183
    },
    {
      "epoch": 6.592679218804818,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 28184
    },
    {
      "epoch": 6.592913109577827,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 28185
    },
    {
      "epoch": 6.593147000350836,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6352,
      "step": 28186
    },
    {
      "epoch": 6.593380891123845,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 28187
    },
    {
      "epoch": 6.5936147818968545,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.58,
      "step": 28188
    },
    {
      "epoch": 6.593848672669863,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7617,
      "step": 28189
    },
    {
      "epoch": 6.594082563442872,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 28190
    },
    {
      "epoch": 6.594316454215881,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 28191
    },
    {
      "epoch": 6.59455034498889,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5496,
      "step": 28192
    },
    {
      "epoch": 6.5947842357618995,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 28193
    },
    {
      "epoch": 6.595018126534908,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 28194
    },
    {
      "epoch": 6.595252017307917,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.866,
      "step": 28195
    },
    {
      "epoch": 6.595485908080926,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.7289,
      "step": 28196
    },
    {
      "epoch": 6.595719798853935,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 28197
    },
    {
      "epoch": 6.595953689626944,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 28198
    },
    {
      "epoch": 6.5961875803999535,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 28199
    },
    {
      "epoch": 6.596421471172962,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 28200
    },
    {
      "epoch": 6.596421471172962,
      "eval_runtime": 4.6251,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 28200
    },
    {
      "epoch": 6.596655361945971,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 28201
    },
    {
      "epoch": 6.59688925271898,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5711,
      "step": 28202
    },
    {
      "epoch": 6.59712314349199,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4057,
      "step": 28203
    },
    {
      "epoch": 6.5973570342649985,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6346,
      "step": 28204
    },
    {
      "epoch": 6.597590925038007,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 28205
    },
    {
      "epoch": 6.597824815811016,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 28206
    },
    {
      "epoch": 6.598058706584025,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5677,
      "step": 28207
    },
    {
      "epoch": 6.598292597357034,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 28208
    },
    {
      "epoch": 6.598526488130044,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7916,
      "step": 28209
    },
    {
      "epoch": 6.5987603789030524,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 28210
    },
    {
      "epoch": 6.598994269676061,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2586,
      "step": 28211
    },
    {
      "epoch": 6.59922816044907,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 28212
    },
    {
      "epoch": 6.599462051222079,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 28213
    },
    {
      "epoch": 6.599695941995089,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5805,
      "step": 28214
    },
    {
      "epoch": 6.5999298327680975,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4719,
      "step": 28215
    },
    {
      "epoch": 6.600163723541106,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.628,
      "step": 28216
    },
    {
      "epoch": 6.600397614314115,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 28217
    },
    {
      "epoch": 6.600631505087124,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 28218
    },
    {
      "epoch": 6.600865395860133,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7479,
      "step": 28219
    },
    {
      "epoch": 6.601099286633143,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 28220
    },
    {
      "epoch": 6.601333177406151,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 28221
    },
    {
      "epoch": 6.60156706817916,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 28222
    },
    {
      "epoch": 6.601800958952169,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 28223
    },
    {
      "epoch": 6.602034849725178,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 28224
    },
    {
      "epoch": 6.602268740498188,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 28225
    },
    {
      "epoch": 6.6025026312711965,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 28226
    },
    {
      "epoch": 6.602736522044205,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 28227
    },
    {
      "epoch": 6.602970412817214,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5097,
      "step": 28228
    },
    {
      "epoch": 6.603204303590223,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 28229
    },
    {
      "epoch": 6.603438194363232,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.6272,
      "step": 28230
    },
    {
      "epoch": 6.603672085136242,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 28231
    },
    {
      "epoch": 6.60390597590925,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 28232
    },
    {
      "epoch": 6.604139866682259,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6619,
      "step": 28233
    },
    {
      "epoch": 6.604373757455268,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6926,
      "step": 28234
    },
    {
      "epoch": 6.604607648228278,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 28235
    },
    {
      "epoch": 6.604841539001287,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 28236
    },
    {
      "epoch": 6.6050754297742955,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 28237
    },
    {
      "epoch": 6.605309320547304,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 28238
    },
    {
      "epoch": 6.605543211320313,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 28239
    },
    {
      "epoch": 6.605777102093322,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 28240
    },
    {
      "epoch": 6.606010992866332,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 28241
    },
    {
      "epoch": 6.606244883639341,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 28242
    },
    {
      "epoch": 6.606478774412349,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 28243
    },
    {
      "epoch": 6.606712665185358,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 28244
    },
    {
      "epoch": 6.606946555958367,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5421,
      "step": 28245
    },
    {
      "epoch": 6.607180446731377,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 28246
    },
    {
      "epoch": 6.607414337504386,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9255,
      "step": 28247
    },
    {
      "epoch": 6.6076482282773945,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7472,
      "step": 28248
    },
    {
      "epoch": 6.607882119050403,
      "grad_norm": 10.875,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 28249
    },
    {
      "epoch": 6.608116009823412,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 28250
    },
    {
      "epoch": 6.608349900596421,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 28251
    },
    {
      "epoch": 6.608583791369431,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 28252
    },
    {
      "epoch": 6.60881768214244,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 28253
    },
    {
      "epoch": 6.609051572915448,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 28254
    },
    {
      "epoch": 6.609285463688457,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 28255
    },
    {
      "epoch": 6.609519354461467,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 28256
    },
    {
      "epoch": 6.609753245234476,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7251,
      "step": 28257
    },
    {
      "epoch": 6.609987136007485,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 28258
    },
    {
      "epoch": 6.6102210267804935,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 28259
    },
    {
      "epoch": 6.610454917553502,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 28260
    },
    {
      "epoch": 6.610688808326511,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 28261
    },
    {
      "epoch": 6.610922699099521,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7345,
      "step": 28262
    },
    {
      "epoch": 6.61115658987253,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 28263
    },
    {
      "epoch": 6.611390480645539,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6222,
      "step": 28264
    },
    {
      "epoch": 6.611624371418547,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 28265
    },
    {
      "epoch": 6.611858262191556,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 28266
    },
    {
      "epoch": 6.612092152964566,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 28267
    },
    {
      "epoch": 6.612326043737575,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6755,
      "step": 28268
    },
    {
      "epoch": 6.612559934510584,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7104,
      "step": 28269
    },
    {
      "epoch": 6.6127938252835925,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 28270
    },
    {
      "epoch": 6.613027716056601,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 28271
    },
    {
      "epoch": 6.61326160682961,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 28272
    },
    {
      "epoch": 6.61349549760262,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 28273
    },
    {
      "epoch": 6.613729388375629,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9563,
      "step": 28274
    },
    {
      "epoch": 6.613963279148638,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 28275
    },
    {
      "epoch": 6.614197169921646,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 28276
    },
    {
      "epoch": 6.614431060694655,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 28277
    },
    {
      "epoch": 6.614664951467665,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 28278
    },
    {
      "epoch": 6.614898842240674,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5967,
      "step": 28279
    },
    {
      "epoch": 6.615132733013683,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 28280
    },
    {
      "epoch": 6.6153666237866915,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8175,
      "step": 28281
    },
    {
      "epoch": 6.6156005145597,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 28282
    },
    {
      "epoch": 6.615834405332709,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6239,
      "step": 28283
    },
    {
      "epoch": 6.616068296105719,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7682,
      "step": 28284
    },
    {
      "epoch": 6.616302186878728,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 28285
    },
    {
      "epoch": 6.616536077651737,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 28286
    },
    {
      "epoch": 6.616769968424745,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 28287
    },
    {
      "epoch": 6.617003859197755,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7165,
      "step": 28288
    },
    {
      "epoch": 6.617237749970764,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 28289
    },
    {
      "epoch": 6.617471640743773,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7115,
      "step": 28290
    },
    {
      "epoch": 6.617705531516782,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 28291
    },
    {
      "epoch": 6.6179394222897905,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 28292
    },
    {
      "epoch": 6.618173313062799,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 28293
    },
    {
      "epoch": 6.618407203835809,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6893,
      "step": 28294
    },
    {
      "epoch": 6.618641094608818,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 28295
    },
    {
      "epoch": 6.618874985381827,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 28296
    },
    {
      "epoch": 6.619108876154836,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7244,
      "step": 28297
    },
    {
      "epoch": 6.619342766927844,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 28298
    },
    {
      "epoch": 6.619576657700854,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 28299
    },
    {
      "epoch": 6.619810548473863,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 28300
    },
    {
      "epoch": 6.619810548473863,
      "eval_runtime": 4.6436,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 28300
    },
    {
      "epoch": 6.620044439246872,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 28301
    },
    {
      "epoch": 6.620278330019881,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8203,
      "step": 28302
    },
    {
      "epoch": 6.6205122207928895,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7423,
      "step": 28303
    },
    {
      "epoch": 6.620746111565898,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 28304
    },
    {
      "epoch": 6.620980002338908,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 28305
    },
    {
      "epoch": 6.621213893111917,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 28306
    },
    {
      "epoch": 6.621447783884926,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 28307
    },
    {
      "epoch": 6.621681674657935,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6938,
      "step": 28308
    },
    {
      "epoch": 6.621915565430943,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 28309
    },
    {
      "epoch": 6.622149456203953,
      "grad_norm": 10.6875,
      "learning_rate": 3e-05,
      "loss": 1.9786,
      "step": 28310
    },
    {
      "epoch": 6.622383346976962,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6199,
      "step": 28311
    },
    {
      "epoch": 6.622617237749971,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 28312
    },
    {
      "epoch": 6.62285112852298,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 28313
    },
    {
      "epoch": 6.6230850192959885,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5889,
      "step": 28314
    },
    {
      "epoch": 6.623318910068997,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 28315
    },
    {
      "epoch": 6.623552800842007,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 28316
    },
    {
      "epoch": 6.623786691615016,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 28317
    },
    {
      "epoch": 6.624020582388025,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 28318
    },
    {
      "epoch": 6.6242544731610336,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 28319
    },
    {
      "epoch": 6.624488363934043,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2266,
      "step": 28320
    },
    {
      "epoch": 6.624722254707052,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 28321
    },
    {
      "epoch": 6.624956145480061,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 28322
    },
    {
      "epoch": 6.62519003625307,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 28323
    },
    {
      "epoch": 6.625423927026079,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 28324
    },
    {
      "epoch": 6.6256578177990875,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 28325
    },
    {
      "epoch": 6.625891708572097,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4156,
      "step": 28326
    },
    {
      "epoch": 6.626125599345106,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.5529,
      "step": 28327
    },
    {
      "epoch": 6.626359490118115,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 28328
    },
    {
      "epoch": 6.626593380891124,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6143,
      "step": 28329
    },
    {
      "epoch": 6.6268272716641325,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 28330
    },
    {
      "epoch": 6.627061162437142,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6971,
      "step": 28331
    },
    {
      "epoch": 6.627295053210151,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7215,
      "step": 28332
    },
    {
      "epoch": 6.62752894398316,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 28333
    },
    {
      "epoch": 6.627762834756169,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6832,
      "step": 28334
    },
    {
      "epoch": 6.627996725529178,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 28335
    },
    {
      "epoch": 6.6282306163021865,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 28336
    },
    {
      "epoch": 6.628464507075196,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.583,
      "step": 28337
    },
    {
      "epoch": 6.628698397848205,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 28338
    },
    {
      "epoch": 6.628932288621214,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 28339
    },
    {
      "epoch": 6.629166179394223,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 28340
    },
    {
      "epoch": 6.6294000701672315,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 28341
    },
    {
      "epoch": 6.629633960940241,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 28342
    },
    {
      "epoch": 6.62986785171325,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 28343
    },
    {
      "epoch": 6.630101742486259,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5218,
      "step": 28344
    },
    {
      "epoch": 6.630335633259268,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 28345
    },
    {
      "epoch": 6.630569524032277,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7572,
      "step": 28346
    },
    {
      "epoch": 6.6308034148052855,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 28347
    },
    {
      "epoch": 6.631037305578295,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 28348
    },
    {
      "epoch": 6.631271196351304,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6028,
      "step": 28349
    },
    {
      "epoch": 6.631505087124313,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7424,
      "step": 28350
    },
    {
      "epoch": 6.631738977897322,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 28351
    },
    {
      "epoch": 6.631972868670331,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7025,
      "step": 28352
    },
    {
      "epoch": 6.63220675944334,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 28353
    },
    {
      "epoch": 6.632440650216349,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 28354
    },
    {
      "epoch": 6.632674540989358,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 28355
    },
    {
      "epoch": 6.632908431762367,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 28356
    },
    {
      "epoch": 6.633142322535376,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 28357
    },
    {
      "epoch": 6.633376213308385,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 28358
    },
    {
      "epoch": 6.633610104081394,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 28359
    },
    {
      "epoch": 6.633843994854403,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 28360
    },
    {
      "epoch": 6.634077885627412,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5587,
      "step": 28361
    },
    {
      "epoch": 6.634311776400421,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 28362
    },
    {
      "epoch": 6.63454566717343,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 28363
    },
    {
      "epoch": 6.634779557946439,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7769,
      "step": 28364
    },
    {
      "epoch": 6.635013448719448,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 28365
    },
    {
      "epoch": 6.635247339492457,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6278,
      "step": 28366
    },
    {
      "epoch": 6.635481230265466,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 28367
    },
    {
      "epoch": 6.635715121038475,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8434,
      "step": 28368
    },
    {
      "epoch": 6.635949011811484,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 28369
    },
    {
      "epoch": 6.636182902584493,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 28370
    },
    {
      "epoch": 6.636416793357502,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6972,
      "step": 28371
    },
    {
      "epoch": 6.636650684130511,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 28372
    },
    {
      "epoch": 6.63688457490352,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6229,
      "step": 28373
    },
    {
      "epoch": 6.637118465676529,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 28374
    },
    {
      "epoch": 6.637352356449538,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 28375
    },
    {
      "epoch": 6.637586247222547,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.593,
      "step": 28376
    },
    {
      "epoch": 6.637820137995556,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6343,
      "step": 28377
    },
    {
      "epoch": 6.638054028768565,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7872,
      "step": 28378
    },
    {
      "epoch": 6.638287919541574,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 28379
    },
    {
      "epoch": 6.638521810314583,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5071,
      "step": 28380
    },
    {
      "epoch": 6.638755701087592,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 28381
    },
    {
      "epoch": 6.638989591860601,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4803,
      "step": 28382
    },
    {
      "epoch": 6.63922348263361,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5316,
      "step": 28383
    },
    {
      "epoch": 6.63945737340662,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 28384
    },
    {
      "epoch": 6.639691264179628,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5966,
      "step": 28385
    },
    {
      "epoch": 6.639925154952637,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 28386
    },
    {
      "epoch": 6.640159045725646,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 28387
    },
    {
      "epoch": 6.640392936498655,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6903,
      "step": 28388
    },
    {
      "epoch": 6.640626827271664,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 28389
    },
    {
      "epoch": 6.6408607180446735,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 28390
    },
    {
      "epoch": 6.641094608817682,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8196,
      "step": 28391
    },
    {
      "epoch": 6.641328499590691,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 28392
    },
    {
      "epoch": 6.6415623903637,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 28393
    },
    {
      "epoch": 6.641796281136709,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8447,
      "step": 28394
    },
    {
      "epoch": 6.642030171909719,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 28395
    },
    {
      "epoch": 6.642264062682727,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5627,
      "step": 28396
    },
    {
      "epoch": 6.642497953455736,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 28397
    },
    {
      "epoch": 6.642731844228745,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 28398
    },
    {
      "epoch": 6.642965735001754,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 28399
    },
    {
      "epoch": 6.643199625774763,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 28400
    },
    {
      "epoch": 6.643199625774763,
      "eval_runtime": 4.6447,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 28400
    },
    {
      "epoch": 6.6434335165477725,
      "grad_norm": 8.6875,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 28401
    },
    {
      "epoch": 6.643667407320781,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 28402
    },
    {
      "epoch": 6.64390129809379,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 28403
    },
    {
      "epoch": 6.644135188866799,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 28404
    },
    {
      "epoch": 6.644369079639808,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 28405
    },
    {
      "epoch": 6.6446029704128176,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 28406
    },
    {
      "epoch": 6.644836861185826,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 28407
    },
    {
      "epoch": 6.645070751958835,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.3636,
      "step": 28408
    },
    {
      "epoch": 6.645304642731844,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 28409
    },
    {
      "epoch": 6.645538533504853,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 28410
    },
    {
      "epoch": 6.645772424277862,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 28411
    },
    {
      "epoch": 6.6460063150508715,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 28412
    },
    {
      "epoch": 6.64624020582388,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6532,
      "step": 28413
    },
    {
      "epoch": 6.646474096596889,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 28414
    },
    {
      "epoch": 6.646707987369898,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 28415
    },
    {
      "epoch": 6.646941878142908,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5499,
      "step": 28416
    },
    {
      "epoch": 6.6471757689159165,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5665,
      "step": 28417
    },
    {
      "epoch": 6.647409659688925,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 28418
    },
    {
      "epoch": 6.647643550461934,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 28419
    },
    {
      "epoch": 6.647877441234943,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6166,
      "step": 28420
    },
    {
      "epoch": 6.648111332007952,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 28421
    },
    {
      "epoch": 6.648345222780962,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 28422
    },
    {
      "epoch": 6.6485791135539705,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 28423
    },
    {
      "epoch": 6.648813004326979,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 28424
    },
    {
      "epoch": 6.649046895099988,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.612,
      "step": 28425
    },
    {
      "epoch": 6.649280785872997,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 28426
    },
    {
      "epoch": 6.649514676646007,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6183,
      "step": 28427
    },
    {
      "epoch": 6.6497485674190155,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 28428
    },
    {
      "epoch": 6.649982458192024,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 28429
    },
    {
      "epoch": 6.650216348965033,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.3755,
      "step": 28430
    },
    {
      "epoch": 6.650450239738042,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 28431
    },
    {
      "epoch": 6.650684130511051,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 28432
    },
    {
      "epoch": 6.650918021284061,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 28433
    },
    {
      "epoch": 6.6511519120570695,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1941,
      "step": 28434
    },
    {
      "epoch": 6.651385802830078,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7377,
      "step": 28435
    },
    {
      "epoch": 6.651619693603087,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 28436
    },
    {
      "epoch": 6.651853584376097,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 28437
    },
    {
      "epoch": 6.652087475149106,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 28438
    },
    {
      "epoch": 6.6523213659221145,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 28439
    },
    {
      "epoch": 6.652555256695123,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 28440
    },
    {
      "epoch": 6.652789147468132,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 28441
    },
    {
      "epoch": 6.653023038241141,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 28442
    },
    {
      "epoch": 6.653256929014151,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.637,
      "step": 28443
    },
    {
      "epoch": 6.65349081978716,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 28444
    },
    {
      "epoch": 6.6537247105601685,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 28445
    },
    {
      "epoch": 6.653958601333177,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 28446
    },
    {
      "epoch": 6.654192492106186,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8354,
      "step": 28447
    },
    {
      "epoch": 6.654426382879196,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 28448
    },
    {
      "epoch": 6.654660273652205,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6291,
      "step": 28449
    },
    {
      "epoch": 6.6548941644252135,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 28450
    },
    {
      "epoch": 6.655128055198222,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 28451
    },
    {
      "epoch": 6.655361945971231,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 28452
    },
    {
      "epoch": 6.65559583674424,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 28453
    },
    {
      "epoch": 6.65582972751725,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 28454
    },
    {
      "epoch": 6.656063618290259,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 28455
    },
    {
      "epoch": 6.6562975090632674,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6741,
      "step": 28456
    },
    {
      "epoch": 6.656531399836276,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6399,
      "step": 28457
    },
    {
      "epoch": 6.656765290609285,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 28458
    },
    {
      "epoch": 6.656999181382295,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6914,
      "step": 28459
    },
    {
      "epoch": 6.657233072155304,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6984,
      "step": 28460
    },
    {
      "epoch": 6.6574669629283125,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 28461
    },
    {
      "epoch": 6.657700853701321,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 28462
    },
    {
      "epoch": 6.65793474447433,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 28463
    },
    {
      "epoch": 6.658168635247339,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 28464
    },
    {
      "epoch": 6.658402526020349,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7656,
      "step": 28465
    },
    {
      "epoch": 6.658636416793358,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 28466
    },
    {
      "epoch": 6.658870307566366,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 28467
    },
    {
      "epoch": 6.659104198339375,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 28468
    },
    {
      "epoch": 6.659338089112385,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5897,
      "step": 28469
    },
    {
      "epoch": 6.659571979885394,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 28470
    },
    {
      "epoch": 6.659805870658403,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 28471
    },
    {
      "epoch": 6.6600397614314115,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 28472
    },
    {
      "epoch": 6.66027365220442,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.72,
      "step": 28473
    },
    {
      "epoch": 6.660507542977429,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 28474
    },
    {
      "epoch": 6.660741433750439,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 28475
    },
    {
      "epoch": 6.660975324523448,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 28476
    },
    {
      "epoch": 6.661209215296457,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4511,
      "step": 28477
    },
    {
      "epoch": 6.661443106069465,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 28478
    },
    {
      "epoch": 6.661676996842474,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 28479
    },
    {
      "epoch": 6.661910887615484,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.4489,
      "step": 28480
    },
    {
      "epoch": 6.662144778388493,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 28481
    },
    {
      "epoch": 6.662378669161502,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 28482
    },
    {
      "epoch": 6.6626125599345105,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 28483
    },
    {
      "epoch": 6.662846450707519,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 28484
    },
    {
      "epoch": 6.663080341480528,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6301,
      "step": 28485
    },
    {
      "epoch": 6.663314232253538,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 28486
    },
    {
      "epoch": 6.663548123026547,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 28487
    },
    {
      "epoch": 6.663782013799556,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 28488
    },
    {
      "epoch": 6.664015904572564,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 28489
    },
    {
      "epoch": 6.664249795345573,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 28490
    },
    {
      "epoch": 6.664483686118583,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 28491
    },
    {
      "epoch": 6.664717576891592,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6612,
      "step": 28492
    },
    {
      "epoch": 6.664951467664601,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6182,
      "step": 28493
    },
    {
      "epoch": 6.6651853584376095,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 28494
    },
    {
      "epoch": 6.665419249210618,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.6006,
      "step": 28495
    },
    {
      "epoch": 6.665653139983627,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 28496
    },
    {
      "epoch": 6.665887030756637,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5974,
      "step": 28497
    },
    {
      "epoch": 6.666120921529646,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.5006,
      "step": 28498
    },
    {
      "epoch": 6.666354812302655,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 28499
    },
    {
      "epoch": 6.666588703075663,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 28500
    },
    {
      "epoch": 6.666588703075663,
      "eval_runtime": 4.5994,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 28500
    },
    {
      "epoch": 6.666822593848673,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5983,
      "step": 28501
    },
    {
      "epoch": 6.667056484621682,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 28502
    },
    {
      "epoch": 6.667290375394691,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 28503
    },
    {
      "epoch": 6.6675242661677,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5635,
      "step": 28504
    },
    {
      "epoch": 6.6677581569407085,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 28505
    },
    {
      "epoch": 6.667992047713717,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 28506
    },
    {
      "epoch": 6.668225938486727,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6993,
      "step": 28507
    },
    {
      "epoch": 6.668459829259736,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 28508
    },
    {
      "epoch": 6.668693720032745,
      "grad_norm": 7.78125,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 28509
    },
    {
      "epoch": 6.668927610805754,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 28510
    },
    {
      "epoch": 6.669161501578762,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 28511
    },
    {
      "epoch": 6.669395392351772,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 28512
    },
    {
      "epoch": 6.669629283124781,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 28513
    },
    {
      "epoch": 6.66986317389779,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6761,
      "step": 28514
    },
    {
      "epoch": 6.670097064670799,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 28515
    },
    {
      "epoch": 6.6703309554438075,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 28516
    },
    {
      "epoch": 6.670564846216816,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4356,
      "step": 28517
    },
    {
      "epoch": 6.670798736989826,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 28518
    },
    {
      "epoch": 6.671032627762835,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 28519
    },
    {
      "epoch": 6.671266518535844,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 28520
    },
    {
      "epoch": 6.671500409308853,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.652,
      "step": 28521
    },
    {
      "epoch": 6.671734300081861,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 28522
    },
    {
      "epoch": 6.671968190854871,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 28523
    },
    {
      "epoch": 6.67220208162788,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 28524
    },
    {
      "epoch": 6.672435972400889,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.6607,
      "step": 28525
    },
    {
      "epoch": 6.672669863173898,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8629,
      "step": 28526
    },
    {
      "epoch": 6.6729037539469065,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 28527
    },
    {
      "epoch": 6.673137644719915,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 28528
    },
    {
      "epoch": 6.673371535492925,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.5151,
      "step": 28529
    },
    {
      "epoch": 6.673605426265934,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6773,
      "step": 28530
    },
    {
      "epoch": 6.673839317038943,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 28531
    },
    {
      "epoch": 6.674073207811952,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 28532
    },
    {
      "epoch": 6.674307098584961,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 28533
    },
    {
      "epoch": 6.67454098935797,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6966,
      "step": 28534
    },
    {
      "epoch": 6.674774880130979,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 28535
    },
    {
      "epoch": 6.675008770903988,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 28536
    },
    {
      "epoch": 6.675242661676997,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 28537
    },
    {
      "epoch": 6.6754765524500055,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 28538
    },
    {
      "epoch": 6.675710443223015,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 28539
    },
    {
      "epoch": 6.675944333996024,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5002,
      "step": 28540
    },
    {
      "epoch": 6.676178224769033,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 28541
    },
    {
      "epoch": 6.676412115542042,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 28542
    },
    {
      "epoch": 6.676646006315051,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 28543
    },
    {
      "epoch": 6.67687989708806,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 28544
    },
    {
      "epoch": 6.677113787861069,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5924,
      "step": 28545
    },
    {
      "epoch": 6.677347678634078,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7718,
      "step": 28546
    },
    {
      "epoch": 6.677581569407087,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 28547
    },
    {
      "epoch": 6.677815460180096,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 28548
    },
    {
      "epoch": 6.6780493509531045,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5375,
      "step": 28549
    },
    {
      "epoch": 6.678283241726114,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6478,
      "step": 28550
    },
    {
      "epoch": 6.678517132499123,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 28551
    },
    {
      "epoch": 6.678751023272132,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 28552
    },
    {
      "epoch": 6.678984914045141,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 28553
    },
    {
      "epoch": 6.67921880481815,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 28554
    },
    {
      "epoch": 6.679452695591159,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 28555
    },
    {
      "epoch": 6.679686586364168,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 28556
    },
    {
      "epoch": 6.679920477137177,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9313,
      "step": 28557
    },
    {
      "epoch": 6.680154367910186,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 28558
    },
    {
      "epoch": 6.680388258683195,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7534,
      "step": 28559
    },
    {
      "epoch": 6.6806221494562035,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5421,
      "step": 28560
    },
    {
      "epoch": 6.680856040229213,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 28561
    },
    {
      "epoch": 6.681089931002222,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 28562
    },
    {
      "epoch": 6.681323821775231,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 28563
    },
    {
      "epoch": 6.68155771254824,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 28564
    },
    {
      "epoch": 6.681791603321249,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 28565
    },
    {
      "epoch": 6.682025494094258,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7392,
      "step": 28566
    },
    {
      "epoch": 6.682259384867267,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4209,
      "step": 28567
    },
    {
      "epoch": 6.682493275640276,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 28568
    },
    {
      "epoch": 6.682727166413285,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 28569
    },
    {
      "epoch": 6.682961057186294,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 28570
    },
    {
      "epoch": 6.683194947959303,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 28571
    },
    {
      "epoch": 6.683428838732312,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 28572
    },
    {
      "epoch": 6.683662729505321,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8929,
      "step": 28573
    },
    {
      "epoch": 6.68389662027833,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6716,
      "step": 28574
    },
    {
      "epoch": 6.684130511051339,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 28575
    },
    {
      "epoch": 6.684364401824348,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6724,
      "step": 28576
    },
    {
      "epoch": 6.684598292597357,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 28577
    },
    {
      "epoch": 6.684832183370366,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 28578
    },
    {
      "epoch": 6.685066074143375,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5938,
      "step": 28579
    },
    {
      "epoch": 6.685299964916384,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 28580
    },
    {
      "epoch": 6.685533855689393,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 28581
    },
    {
      "epoch": 6.685767746462402,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 28582
    },
    {
      "epoch": 6.686001637235411,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.608,
      "step": 28583
    },
    {
      "epoch": 6.68623552800842,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 28584
    },
    {
      "epoch": 6.686469418781429,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 28585
    },
    {
      "epoch": 6.686703309554438,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 28586
    },
    {
      "epoch": 6.686937200327447,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 28587
    },
    {
      "epoch": 6.687171091100456,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 28588
    },
    {
      "epoch": 6.687404981873465,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 28589
    },
    {
      "epoch": 6.687638872646474,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 28590
    },
    {
      "epoch": 6.687872763419483,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7245,
      "step": 28591
    },
    {
      "epoch": 6.688106654192492,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 28592
    },
    {
      "epoch": 6.688340544965501,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 28593
    },
    {
      "epoch": 6.68857443573851,
      "grad_norm": 7.34375,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 28594
    },
    {
      "epoch": 6.688808326511519,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6927,
      "step": 28595
    },
    {
      "epoch": 6.689042217284528,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 28596
    },
    {
      "epoch": 6.689276108057538,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6589,
      "step": 28597
    },
    {
      "epoch": 6.689509998830546,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 28598
    },
    {
      "epoch": 6.689743889603555,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7483,
      "step": 28599
    },
    {
      "epoch": 6.689977780376564,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5981,
      "step": 28600
    },
    {
      "epoch": 6.689977780376564,
      "eval_runtime": 4.6238,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 28600
    },
    {
      "epoch": 6.690211671149573,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9811,
      "step": 28601
    },
    {
      "epoch": 6.690445561922582,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 28602
    },
    {
      "epoch": 6.6906794526955915,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 28603
    },
    {
      "epoch": 6.6909133434686,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4864,
      "step": 28604
    },
    {
      "epoch": 6.691147234241609,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 28605
    },
    {
      "epoch": 6.691381125014618,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 28606
    },
    {
      "epoch": 6.691615015787627,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 28607
    },
    {
      "epoch": 6.691848906560637,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 28608
    },
    {
      "epoch": 6.692082797333645,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.514,
      "step": 28609
    },
    {
      "epoch": 6.692316688106654,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 28610
    },
    {
      "epoch": 6.692550578879663,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 28611
    },
    {
      "epoch": 6.692784469652672,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 28612
    },
    {
      "epoch": 6.693018360425681,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6483,
      "step": 28613
    },
    {
      "epoch": 6.6932522511986905,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 28614
    },
    {
      "epoch": 6.693486141971699,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6335,
      "step": 28615
    },
    {
      "epoch": 6.693720032744708,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.3309,
      "step": 28616
    },
    {
      "epoch": 6.693953923517717,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 28617
    },
    {
      "epoch": 6.694187814290726,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 28618
    },
    {
      "epoch": 6.694421705063736,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.4986,
      "step": 28619
    },
    {
      "epoch": 6.694655595836744,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 28620
    },
    {
      "epoch": 6.694889486609753,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 28621
    },
    {
      "epoch": 6.695123377382762,
      "grad_norm": 7.46875,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 28622
    },
    {
      "epoch": 6.695357268155771,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6948,
      "step": 28623
    },
    {
      "epoch": 6.69559115892878,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 28624
    },
    {
      "epoch": 6.6958250497017895,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.668,
      "step": 28625
    },
    {
      "epoch": 6.696058940474798,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 28626
    },
    {
      "epoch": 6.696292831247807,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 2.3346,
      "step": 28627
    },
    {
      "epoch": 6.696526722020816,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 28628
    },
    {
      "epoch": 6.696760612793826,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7592,
      "step": 28629
    },
    {
      "epoch": 6.696994503566835,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 28630
    },
    {
      "epoch": 6.697228394339843,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4938,
      "step": 28631
    },
    {
      "epoch": 6.697462285112852,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 28632
    },
    {
      "epoch": 6.697696175885861,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.8897,
      "step": 28633
    },
    {
      "epoch": 6.69793006665887,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 28634
    },
    {
      "epoch": 6.69816395743188,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.3838,
      "step": 28635
    },
    {
      "epoch": 6.6983978482048885,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 28636
    },
    {
      "epoch": 6.698631738977897,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 28637
    },
    {
      "epoch": 6.698865629750906,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7518,
      "step": 28638
    },
    {
      "epoch": 6.699099520523915,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7056,
      "step": 28639
    },
    {
      "epoch": 6.699333411296925,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 28640
    },
    {
      "epoch": 6.699567302069934,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 28641
    },
    {
      "epoch": 6.699801192842942,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 28642
    },
    {
      "epoch": 6.700035083615951,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 28643
    },
    {
      "epoch": 6.70026897438896,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 28644
    },
    {
      "epoch": 6.700502865161969,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 28645
    },
    {
      "epoch": 6.700736755934979,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 28646
    },
    {
      "epoch": 6.7009706467079875,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 28647
    },
    {
      "epoch": 6.701204537480996,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6252,
      "step": 28648
    },
    {
      "epoch": 6.701438428254005,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7614,
      "step": 28649
    },
    {
      "epoch": 6.701672319027015,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 28650
    },
    {
      "epoch": 6.701906209800024,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 28651
    },
    {
      "epoch": 6.7021401005730326,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5854,
      "step": 28652
    },
    {
      "epoch": 6.702373991346041,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.748,
      "step": 28653
    },
    {
      "epoch": 6.70260788211905,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 28654
    },
    {
      "epoch": 6.702841772892059,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4101,
      "step": 28655
    },
    {
      "epoch": 6.703075663665069,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 28656
    },
    {
      "epoch": 6.703309554438078,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 28657
    },
    {
      "epoch": 6.7035434452110865,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7085,
      "step": 28658
    },
    {
      "epoch": 6.703777335984095,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5861,
      "step": 28659
    },
    {
      "epoch": 6.704011226757104,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 28660
    },
    {
      "epoch": 6.704245117530114,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 28661
    },
    {
      "epoch": 6.704479008303123,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 28662
    },
    {
      "epoch": 6.7047128990761315,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 28663
    },
    {
      "epoch": 6.70494678984914,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 28664
    },
    {
      "epoch": 6.705180680622149,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6907,
      "step": 28665
    },
    {
      "epoch": 6.705414571395158,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5822,
      "step": 28666
    },
    {
      "epoch": 6.705648462168168,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 28667
    },
    {
      "epoch": 6.705882352941177,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 28668
    },
    {
      "epoch": 6.7061162437141855,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 28669
    },
    {
      "epoch": 6.706350134487194,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.597,
      "step": 28670
    },
    {
      "epoch": 6.706584025260203,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 28671
    },
    {
      "epoch": 6.706817916033213,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 28672
    },
    {
      "epoch": 6.707051806806222,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.49,
      "step": 28673
    },
    {
      "epoch": 6.7072856975792305,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 28674
    },
    {
      "epoch": 6.707519588352239,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 28675
    },
    {
      "epoch": 6.707753479125248,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 28676
    },
    {
      "epoch": 6.707987369898257,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 28677
    },
    {
      "epoch": 6.708221260671267,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5633,
      "step": 28678
    },
    {
      "epoch": 6.708455151444276,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 28679
    },
    {
      "epoch": 6.7086890422172845,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 28680
    },
    {
      "epoch": 6.708922932990293,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.762,
      "step": 28681
    },
    {
      "epoch": 6.709156823763303,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 28682
    },
    {
      "epoch": 6.709390714536312,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7042,
      "step": 28683
    },
    {
      "epoch": 6.709624605309321,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 28684
    },
    {
      "epoch": 6.7098584960823295,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 28685
    },
    {
      "epoch": 6.710092386855338,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 28686
    },
    {
      "epoch": 6.710326277628347,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 28687
    },
    {
      "epoch": 6.710560168401357,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 28688
    },
    {
      "epoch": 6.710794059174366,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 28689
    },
    {
      "epoch": 6.711027949947375,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 28690
    },
    {
      "epoch": 6.7112618407203835,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 28691
    },
    {
      "epoch": 6.711495731493392,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 28692
    },
    {
      "epoch": 6.711729622266402,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7479,
      "step": 28693
    },
    {
      "epoch": 6.711963513039411,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 28694
    },
    {
      "epoch": 6.71219740381242,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 28695
    },
    {
      "epoch": 6.7124312945854285,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 28696
    },
    {
      "epoch": 6.712665185358437,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 28697
    },
    {
      "epoch": 6.712899076131446,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.614,
      "step": 28698
    },
    {
      "epoch": 6.713132966904456,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5865,
      "step": 28699
    },
    {
      "epoch": 6.713366857677465,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6671,
      "step": 28700
    },
    {
      "epoch": 6.713366857677465,
      "eval_runtime": 4.6212,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 28700
    },
    {
      "epoch": 6.713600748450474,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 28701
    },
    {
      "epoch": 6.7138346392234824,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.517,
      "step": 28702
    },
    {
      "epoch": 6.714068529996491,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 28703
    },
    {
      "epoch": 6.714302420769501,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4905,
      "step": 28704
    },
    {
      "epoch": 6.71453631154251,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 28705
    },
    {
      "epoch": 6.714770202315519,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 28706
    },
    {
      "epoch": 6.7150040930885275,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 28707
    },
    {
      "epoch": 6.715237983861536,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 28708
    },
    {
      "epoch": 6.715471874634545,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7074,
      "step": 28709
    },
    {
      "epoch": 6.715705765407555,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 28710
    },
    {
      "epoch": 6.715939656180564,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 28711
    },
    {
      "epoch": 6.716173546953573,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5786,
      "step": 28712
    },
    {
      "epoch": 6.716407437726581,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 28713
    },
    {
      "epoch": 6.716641328499591,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6416,
      "step": 28714
    },
    {
      "epoch": 6.7168752192726,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7745,
      "step": 28715
    },
    {
      "epoch": 6.717109110045609,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 28716
    },
    {
      "epoch": 6.717343000818618,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7025,
      "step": 28717
    },
    {
      "epoch": 6.7175768915916265,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6492,
      "step": 28718
    },
    {
      "epoch": 6.717810782364635,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 28719
    },
    {
      "epoch": 6.718044673137645,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 28720
    },
    {
      "epoch": 6.718278563910654,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 28721
    },
    {
      "epoch": 6.718512454683663,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 28722
    },
    {
      "epoch": 6.718746345456672,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 28723
    },
    {
      "epoch": 6.71898023622968,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 28724
    },
    {
      "epoch": 6.71921412700269,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 28725
    },
    {
      "epoch": 6.719448017775699,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 28726
    },
    {
      "epoch": 6.719681908548708,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 28727
    },
    {
      "epoch": 6.719915799321717,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 28728
    },
    {
      "epoch": 6.7201496900947255,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 28729
    },
    {
      "epoch": 6.720383580867734,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 28730
    },
    {
      "epoch": 6.720617471640744,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 28731
    },
    {
      "epoch": 6.720851362413753,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8169,
      "step": 28732
    },
    {
      "epoch": 6.721085253186762,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 28733
    },
    {
      "epoch": 6.721319143959771,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7327,
      "step": 28734
    },
    {
      "epoch": 6.721553034732779,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 28735
    },
    {
      "epoch": 6.721786925505789,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 28736
    },
    {
      "epoch": 6.722020816278798,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 28737
    },
    {
      "epoch": 6.722254707051807,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6935,
      "step": 28738
    },
    {
      "epoch": 6.722488597824816,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6231,
      "step": 28739
    },
    {
      "epoch": 6.7227224885978245,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5008,
      "step": 28740
    },
    {
      "epoch": 6.722956379370833,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 28741
    },
    {
      "epoch": 6.723190270143843,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 28742
    },
    {
      "epoch": 6.723424160916852,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7018,
      "step": 28743
    },
    {
      "epoch": 6.723658051689861,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 28744
    },
    {
      "epoch": 6.72389194246287,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6793,
      "step": 28745
    },
    {
      "epoch": 6.724125833235879,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 28746
    },
    {
      "epoch": 6.724359724008888,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 28747
    },
    {
      "epoch": 6.724593614781897,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 28748
    },
    {
      "epoch": 6.724827505554906,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6694,
      "step": 28749
    },
    {
      "epoch": 6.725061396327915,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 28750
    },
    {
      "epoch": 6.7252952871009235,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 28751
    },
    {
      "epoch": 6.725529177873933,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 28752
    },
    {
      "epoch": 6.725763068646942,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 28753
    },
    {
      "epoch": 6.725996959419951,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7495,
      "step": 28754
    },
    {
      "epoch": 6.72623085019296,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 28755
    },
    {
      "epoch": 6.726464740965969,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 28756
    },
    {
      "epoch": 6.726698631738978,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 28757
    },
    {
      "epoch": 6.726932522511987,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 28758
    },
    {
      "epoch": 6.727166413284996,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5108,
      "step": 28759
    },
    {
      "epoch": 6.727400304058005,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 28760
    },
    {
      "epoch": 6.727634194831014,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 28761
    },
    {
      "epoch": 6.7278680856040225,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8289,
      "step": 28762
    },
    {
      "epoch": 6.728101976377032,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 28763
    },
    {
      "epoch": 6.728335867150041,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7188,
      "step": 28764
    },
    {
      "epoch": 6.72856975792305,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 28765
    },
    {
      "epoch": 6.728803648696059,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 28766
    },
    {
      "epoch": 6.729037539469068,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 28767
    },
    {
      "epoch": 6.729271430242077,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7988,
      "step": 28768
    },
    {
      "epoch": 6.729505321015086,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 28769
    },
    {
      "epoch": 6.729739211788095,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 28770
    },
    {
      "epoch": 6.729973102561104,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5056,
      "step": 28771
    },
    {
      "epoch": 6.730206993334113,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 28772
    },
    {
      "epoch": 6.7304408841071215,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 28773
    },
    {
      "epoch": 6.730674774880131,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 28774
    },
    {
      "epoch": 6.73090866565314,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 28775
    },
    {
      "epoch": 6.731142556426149,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 28776
    },
    {
      "epoch": 6.731376447199158,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 28777
    },
    {
      "epoch": 6.7316103379721675,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 28778
    },
    {
      "epoch": 6.731844228745176,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 28779
    },
    {
      "epoch": 6.732078119518185,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 28780
    },
    {
      "epoch": 6.732312010291194,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.628,
      "step": 28781
    },
    {
      "epoch": 6.732545901064203,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 28782
    },
    {
      "epoch": 6.732779791837212,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 28783
    },
    {
      "epoch": 6.733013682610221,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 28784
    },
    {
      "epoch": 6.73324757338323,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 28785
    },
    {
      "epoch": 6.733481464156239,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 28786
    },
    {
      "epoch": 6.733715354929248,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6241,
      "step": 28787
    },
    {
      "epoch": 6.733949245702257,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 28788
    },
    {
      "epoch": 6.7341831364752665,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 28789
    },
    {
      "epoch": 6.734417027248275,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 28790
    },
    {
      "epoch": 6.734650918021284,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6149,
      "step": 28791
    },
    {
      "epoch": 6.734884808794293,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.43,
      "step": 28792
    },
    {
      "epoch": 6.735118699567302,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 28793
    },
    {
      "epoch": 6.735352590340311,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 28794
    },
    {
      "epoch": 6.73558648111332,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 28795
    },
    {
      "epoch": 6.735820371886329,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5286,
      "step": 28796
    },
    {
      "epoch": 6.736054262659338,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5294,
      "step": 28797
    },
    {
      "epoch": 6.736288153432347,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 28798
    },
    {
      "epoch": 6.736522044205356,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7475,
      "step": 28799
    },
    {
      "epoch": 6.7367559349783654,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.4724,
      "step": 28800
    },
    {
      "epoch": 6.7367559349783654,
      "eval_runtime": 4.6132,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 28800
    },
    {
      "epoch": 6.736989825751374,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 28801
    },
    {
      "epoch": 6.737223716524383,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.761,
      "step": 28802
    },
    {
      "epoch": 6.737457607297392,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 28803
    },
    {
      "epoch": 6.737691498070401,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8329,
      "step": 28804
    },
    {
      "epoch": 6.73792538884341,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.652,
      "step": 28805
    },
    {
      "epoch": 6.738159279616419,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4365,
      "step": 28806
    },
    {
      "epoch": 6.738393170389428,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 28807
    },
    {
      "epoch": 6.738627061162437,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 28808
    },
    {
      "epoch": 6.738860951935446,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 28809
    },
    {
      "epoch": 6.739094842708456,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 28810
    },
    {
      "epoch": 6.739328733481464,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 28811
    },
    {
      "epoch": 6.739562624254473,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 28812
    },
    {
      "epoch": 6.739796515027482,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9912,
      "step": 28813
    },
    {
      "epoch": 6.740030405800491,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6507,
      "step": 28814
    },
    {
      "epoch": 6.7402642965735,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 28815
    },
    {
      "epoch": 6.7404981873465095,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 28816
    },
    {
      "epoch": 6.740732078119518,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 28817
    },
    {
      "epoch": 6.740965968892527,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 28818
    },
    {
      "epoch": 6.741199859665536,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 28819
    },
    {
      "epoch": 6.741433750438545,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6297,
      "step": 28820
    },
    {
      "epoch": 6.741667641211555,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 28821
    },
    {
      "epoch": 6.741901531984563,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7204,
      "step": 28822
    },
    {
      "epoch": 6.742135422757572,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 28823
    },
    {
      "epoch": 6.742369313530581,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6428,
      "step": 28824
    },
    {
      "epoch": 6.74260320430359,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 28825
    },
    {
      "epoch": 6.742837095076599,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 28826
    },
    {
      "epoch": 6.7430709858496085,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 28827
    },
    {
      "epoch": 6.743304876622617,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 28828
    },
    {
      "epoch": 6.743538767395626,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 28829
    },
    {
      "epoch": 6.743772658168635,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 28830
    },
    {
      "epoch": 6.744006548941644,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6759,
      "step": 28831
    },
    {
      "epoch": 6.744240439714654,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.5671,
      "step": 28832
    },
    {
      "epoch": 6.744474330487662,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.598,
      "step": 28833
    },
    {
      "epoch": 6.744708221260671,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6705,
      "step": 28834
    },
    {
      "epoch": 6.74494211203368,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7642,
      "step": 28835
    },
    {
      "epoch": 6.745176002806689,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6152,
      "step": 28836
    },
    {
      "epoch": 6.745409893579699,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 28837
    },
    {
      "epoch": 6.7456437843527075,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 28838
    },
    {
      "epoch": 6.745877675125716,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7147,
      "step": 28839
    },
    {
      "epoch": 6.746111565898725,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 28840
    },
    {
      "epoch": 6.746345456671734,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 28841
    },
    {
      "epoch": 6.746579347444744,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 28842
    },
    {
      "epoch": 6.746813238217753,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 28843
    },
    {
      "epoch": 6.747047128990761,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 28844
    },
    {
      "epoch": 6.74728101976377,
      "grad_norm": 9.9375,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 28845
    },
    {
      "epoch": 6.747514910536779,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 28846
    },
    {
      "epoch": 6.747748801309788,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6747,
      "step": 28847
    },
    {
      "epoch": 6.747982692082798,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 28848
    },
    {
      "epoch": 6.7482165828558065,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 28849
    },
    {
      "epoch": 6.748450473628815,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 28850
    },
    {
      "epoch": 6.748684364401824,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 28851
    },
    {
      "epoch": 6.748918255174833,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 28852
    },
    {
      "epoch": 6.749152145947843,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 28853
    },
    {
      "epoch": 6.749386036720852,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 28854
    },
    {
      "epoch": 6.74961992749386,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7311,
      "step": 28855
    },
    {
      "epoch": 6.749853818266869,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7082,
      "step": 28856
    },
    {
      "epoch": 6.750087709039878,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.628,
      "step": 28857
    },
    {
      "epoch": 6.750321599812887,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 28858
    },
    {
      "epoch": 6.750555490585897,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 28859
    },
    {
      "epoch": 6.7507893813589055,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7946,
      "step": 28860
    },
    {
      "epoch": 6.751023272131914,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 28861
    },
    {
      "epoch": 6.751257162904923,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 28862
    },
    {
      "epoch": 6.751491053677933,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 28863
    },
    {
      "epoch": 6.751724944450942,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5942,
      "step": 28864
    },
    {
      "epoch": 6.751958835223951,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7403,
      "step": 28865
    },
    {
      "epoch": 6.752192725996959,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 28866
    },
    {
      "epoch": 6.752426616769968,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 28867
    },
    {
      "epoch": 6.752660507542977,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5264,
      "step": 28868
    },
    {
      "epoch": 6.752894398315987,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 28869
    },
    {
      "epoch": 6.753128289088996,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 28870
    },
    {
      "epoch": 6.7533621798620045,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.851,
      "step": 28871
    },
    {
      "epoch": 6.753596070635013,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 28872
    },
    {
      "epoch": 6.753829961408022,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5804,
      "step": 28873
    },
    {
      "epoch": 6.754063852181032,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 28874
    },
    {
      "epoch": 6.754297742954041,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5662,
      "step": 28875
    },
    {
      "epoch": 6.75453163372705,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 28876
    },
    {
      "epoch": 6.754765524500058,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7586,
      "step": 28877
    },
    {
      "epoch": 6.754999415273067,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 28878
    },
    {
      "epoch": 6.755233306046076,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 28879
    },
    {
      "epoch": 6.755467196819086,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 28880
    },
    {
      "epoch": 6.755701087592095,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 28881
    },
    {
      "epoch": 6.7559349783651035,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 28882
    },
    {
      "epoch": 6.756168869138112,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 28883
    },
    {
      "epoch": 6.756402759911121,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7112,
      "step": 28884
    },
    {
      "epoch": 6.756636650684131,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9751,
      "step": 28885
    },
    {
      "epoch": 6.75687054145714,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7278,
      "step": 28886
    },
    {
      "epoch": 6.757104432230149,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5148,
      "step": 28887
    },
    {
      "epoch": 6.757338323003157,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4521,
      "step": 28888
    },
    {
      "epoch": 6.757572213776166,
      "grad_norm": 8.5,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 28889
    },
    {
      "epoch": 6.757806104549175,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 28890
    },
    {
      "epoch": 6.758039995322185,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 28891
    },
    {
      "epoch": 6.758273886095194,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.4091,
      "step": 28892
    },
    {
      "epoch": 6.7585077768682025,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 28893
    },
    {
      "epoch": 6.758741667641211,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 28894
    },
    {
      "epoch": 6.758975558414221,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 28895
    },
    {
      "epoch": 6.75920944918723,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 28896
    },
    {
      "epoch": 6.759443339960239,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 28897
    },
    {
      "epoch": 6.7596772307332476,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7997,
      "step": 28898
    },
    {
      "epoch": 6.759911121506256,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6404,
      "step": 28899
    },
    {
      "epoch": 6.760145012279265,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 28900
    },
    {
      "epoch": 6.760145012279265,
      "eval_runtime": 4.6065,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 28900
    },
    {
      "epoch": 6.760378903052275,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7253,
      "step": 28901
    },
    {
      "epoch": 6.760612793825284,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 28902
    },
    {
      "epoch": 6.760846684598293,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 28903
    },
    {
      "epoch": 6.7610805753713015,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 28904
    },
    {
      "epoch": 6.76131446614431,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 28905
    },
    {
      "epoch": 6.76154835691732,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 28906
    },
    {
      "epoch": 6.761782247690329,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 28907
    },
    {
      "epoch": 6.762016138463338,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 28908
    },
    {
      "epoch": 6.7622500292363465,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7714,
      "step": 28909
    },
    {
      "epoch": 6.762483920009355,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 28910
    },
    {
      "epoch": 6.762717810782364,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5946,
      "step": 28911
    },
    {
      "epoch": 6.762951701555374,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 28912
    },
    {
      "epoch": 6.763185592328383,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 28913
    },
    {
      "epoch": 6.763419483101392,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 28914
    },
    {
      "epoch": 6.7636533738744005,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 28915
    },
    {
      "epoch": 6.763887264647409,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6592,
      "step": 28916
    },
    {
      "epoch": 6.764121155420419,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 28917
    },
    {
      "epoch": 6.764355046193428,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 28918
    },
    {
      "epoch": 6.764588936966437,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 28919
    },
    {
      "epoch": 6.7648228277394455,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.712,
      "step": 28920
    },
    {
      "epoch": 6.765056718512454,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7447,
      "step": 28921
    },
    {
      "epoch": 6.765290609285463,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 28922
    },
    {
      "epoch": 6.765524500058473,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.3781,
      "step": 28923
    },
    {
      "epoch": 6.765758390831482,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8028,
      "step": 28924
    },
    {
      "epoch": 6.765992281604491,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 28925
    },
    {
      "epoch": 6.7662261723774995,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 28926
    },
    {
      "epoch": 6.766460063150509,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 28927
    },
    {
      "epoch": 6.766693953923518,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 28928
    },
    {
      "epoch": 6.766927844696527,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7701,
      "step": 28929
    },
    {
      "epoch": 6.767161735469536,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.563,
      "step": 28930
    },
    {
      "epoch": 6.7673956262425445,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 28931
    },
    {
      "epoch": 6.767629517015553,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 28932
    },
    {
      "epoch": 6.767863407788563,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 28933
    },
    {
      "epoch": 6.768097298561572,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 28934
    },
    {
      "epoch": 6.768331189334581,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6503,
      "step": 28935
    },
    {
      "epoch": 6.76856508010759,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 28936
    },
    {
      "epoch": 6.7687989708805985,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 28937
    },
    {
      "epoch": 6.769032861653608,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 28938
    },
    {
      "epoch": 6.769266752426617,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.5866,
      "step": 28939
    },
    {
      "epoch": 6.769500643199626,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 28940
    },
    {
      "epoch": 6.769734533972635,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 28941
    },
    {
      "epoch": 6.7699684247456435,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6391,
      "step": 28942
    },
    {
      "epoch": 6.770202315518652,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7092,
      "step": 28943
    },
    {
      "epoch": 6.770436206291662,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 28944
    },
    {
      "epoch": 6.770670097064671,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 28945
    },
    {
      "epoch": 6.77090398783768,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.5214,
      "step": 28946
    },
    {
      "epoch": 6.771137878610689,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 28947
    },
    {
      "epoch": 6.7713717693836974,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 28948
    },
    {
      "epoch": 6.771605660156707,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 28949
    },
    {
      "epoch": 6.771839550929716,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 28950
    },
    {
      "epoch": 6.772073441702725,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7934,
      "step": 28951
    },
    {
      "epoch": 6.772307332475734,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 28952
    },
    {
      "epoch": 6.7725412232487425,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 28953
    },
    {
      "epoch": 6.772775114021751,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 28954
    },
    {
      "epoch": 6.773009004794761,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.7876,
      "step": 28955
    },
    {
      "epoch": 6.77324289556777,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6774,
      "step": 28956
    },
    {
      "epoch": 6.773476786340779,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 28957
    },
    {
      "epoch": 6.773710677113788,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5217,
      "step": 28958
    },
    {
      "epoch": 6.773944567886797,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 28959
    },
    {
      "epoch": 6.774178458659806,
      "grad_norm": 10.25,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 28960
    },
    {
      "epoch": 6.774412349432815,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 28961
    },
    {
      "epoch": 6.774646240205824,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 28962
    },
    {
      "epoch": 6.774880130978833,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 28963
    },
    {
      "epoch": 6.7751140217518415,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 28964
    },
    {
      "epoch": 6.775347912524851,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 28965
    },
    {
      "epoch": 6.77558180329786,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 28966
    },
    {
      "epoch": 6.775815694070869,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 28967
    },
    {
      "epoch": 6.776049584843878,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 28968
    },
    {
      "epoch": 6.776283475616887,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 28969
    },
    {
      "epoch": 6.776517366389896,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 28970
    },
    {
      "epoch": 6.776751257162905,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 28971
    },
    {
      "epoch": 6.776985147935914,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5752,
      "step": 28972
    },
    {
      "epoch": 6.777219038708923,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 28973
    },
    {
      "epoch": 6.777452929481932,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 28974
    },
    {
      "epoch": 6.7776868202549405,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7482,
      "step": 28975
    },
    {
      "epoch": 6.77792071102795,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 28976
    },
    {
      "epoch": 6.778154601800959,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 28977
    },
    {
      "epoch": 6.778388492573968,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 28978
    },
    {
      "epoch": 6.778622383346977,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 28979
    },
    {
      "epoch": 6.778856274119986,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 28980
    },
    {
      "epoch": 6.779090164892995,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 28981
    },
    {
      "epoch": 6.779324055666004,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6012,
      "step": 28982
    },
    {
      "epoch": 6.779557946439013,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 28983
    },
    {
      "epoch": 6.779791837212022,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 28984
    },
    {
      "epoch": 6.780025727985031,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 28985
    },
    {
      "epoch": 6.7802596187580395,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 28986
    },
    {
      "epoch": 6.780493509531049,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 28987
    },
    {
      "epoch": 6.780727400304058,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 28988
    },
    {
      "epoch": 6.780961291077067,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 28989
    },
    {
      "epoch": 6.781195181850076,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6482,
      "step": 28990
    },
    {
      "epoch": 6.7814290726230855,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 28991
    },
    {
      "epoch": 6.781662963396094,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 28992
    },
    {
      "epoch": 6.781896854169103,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7118,
      "step": 28993
    },
    {
      "epoch": 6.782130744942112,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8377,
      "step": 28994
    },
    {
      "epoch": 6.782364635715121,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 28995
    },
    {
      "epoch": 6.78259852648813,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6649,
      "step": 28996
    },
    {
      "epoch": 6.782832417261139,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 28997
    },
    {
      "epoch": 6.783066308034148,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 28998
    },
    {
      "epoch": 6.783300198807157,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7058,
      "step": 28999
    },
    {
      "epoch": 6.783534089580166,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 29000
    },
    {
      "epoch": 6.783534089580166,
      "eval_runtime": 4.6204,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 29000
    },
    {
      "epoch": 6.783767980353175,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 29001
    },
    {
      "epoch": 6.7840018711261845,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 29002
    },
    {
      "epoch": 6.784235761899193,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 29003
    },
    {
      "epoch": 6.784469652672202,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 29004
    },
    {
      "epoch": 6.784703543445211,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 29005
    },
    {
      "epoch": 6.78493743421822,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.4316,
      "step": 29006
    },
    {
      "epoch": 6.785171324991229,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 29007
    },
    {
      "epoch": 6.785405215764238,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7865,
      "step": 29008
    },
    {
      "epoch": 6.785639106537247,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 29009
    },
    {
      "epoch": 6.785872997310256,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.5935,
      "step": 29010
    },
    {
      "epoch": 6.786106888083265,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 29011
    },
    {
      "epoch": 6.786340778856274,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 29012
    },
    {
      "epoch": 6.7865746696292835,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 29013
    },
    {
      "epoch": 6.786808560402292,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 29014
    },
    {
      "epoch": 6.787042451175301,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6172,
      "step": 29015
    },
    {
      "epoch": 6.78727634194831,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7258,
      "step": 29016
    },
    {
      "epoch": 6.787510232721319,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6476,
      "step": 29017
    },
    {
      "epoch": 6.787744123494328,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7794,
      "step": 29018
    },
    {
      "epoch": 6.787978014267337,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7172,
      "step": 29019
    },
    {
      "epoch": 6.788211905040346,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2307,
      "step": 29020
    },
    {
      "epoch": 6.788445795813355,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8824,
      "step": 29021
    },
    {
      "epoch": 6.788679686586364,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 29022
    },
    {
      "epoch": 6.788913577359374,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 29023
    },
    {
      "epoch": 6.7891474681323825,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.761,
      "step": 29024
    },
    {
      "epoch": 6.789381358905391,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 29025
    },
    {
      "epoch": 6.7896152496784,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7698,
      "step": 29026
    },
    {
      "epoch": 6.789849140451409,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 29027
    },
    {
      "epoch": 6.790083031224418,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6483,
      "step": 29028
    },
    {
      "epoch": 6.7903169219974275,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 29029
    },
    {
      "epoch": 6.790550812770436,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 29030
    },
    {
      "epoch": 6.790784703543445,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 29031
    },
    {
      "epoch": 6.791018594316454,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3811,
      "step": 29032
    },
    {
      "epoch": 6.791252485089463,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8508,
      "step": 29033
    },
    {
      "epoch": 6.791486375862473,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 29034
    },
    {
      "epoch": 6.7917202666354815,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6456,
      "step": 29035
    },
    {
      "epoch": 6.79195415740849,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 29036
    },
    {
      "epoch": 6.792188048181499,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2634,
      "step": 29037
    },
    {
      "epoch": 6.792421938954508,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6698,
      "step": 29038
    },
    {
      "epoch": 6.792655829727517,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7316,
      "step": 29039
    },
    {
      "epoch": 6.7928897205005265,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 29040
    },
    {
      "epoch": 6.793123611273535,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5898,
      "step": 29041
    },
    {
      "epoch": 6.793357502046544,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 29042
    },
    {
      "epoch": 6.793591392819553,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.497,
      "step": 29043
    },
    {
      "epoch": 6.793825283592563,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 29044
    },
    {
      "epoch": 6.794059174365572,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7215,
      "step": 29045
    },
    {
      "epoch": 6.7942930651385804,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 29046
    },
    {
      "epoch": 6.794526955911589,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 29047
    },
    {
      "epoch": 6.794760846684598,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.622,
      "step": 29048
    },
    {
      "epoch": 6.794994737457607,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.788,
      "step": 29049
    },
    {
      "epoch": 6.795228628230617,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 29050
    },
    {
      "epoch": 6.7954625190036255,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4987,
      "step": 29051
    },
    {
      "epoch": 6.795696409776634,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 29052
    },
    {
      "epoch": 6.795930300549643,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6276,
      "step": 29053
    },
    {
      "epoch": 6.796164191322652,
      "grad_norm": 8.0,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 29054
    },
    {
      "epoch": 6.796398082095662,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6773,
      "step": 29055
    },
    {
      "epoch": 6.796631972868671,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 29056
    },
    {
      "epoch": 6.796865863641679,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 29057
    },
    {
      "epoch": 6.797099754414688,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6188,
      "step": 29058
    },
    {
      "epoch": 6.797333645187697,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8451,
      "step": 29059
    },
    {
      "epoch": 6.797567535960706,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 29060
    },
    {
      "epoch": 6.797801426733716,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 29061
    },
    {
      "epoch": 6.7980353175067245,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 29062
    },
    {
      "epoch": 6.798269208279733,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 29063
    },
    {
      "epoch": 6.798503099052742,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6184,
      "step": 29064
    },
    {
      "epoch": 6.798736989825751,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 29065
    },
    {
      "epoch": 6.798970880598761,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 29066
    },
    {
      "epoch": 6.79920477137177,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 29067
    },
    {
      "epoch": 6.799438662144778,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 29068
    },
    {
      "epoch": 6.799672552917787,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 29069
    },
    {
      "epoch": 6.799906443690796,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8801,
      "step": 29070
    },
    {
      "epoch": 6.800140334463805,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.575,
      "step": 29071
    },
    {
      "epoch": 6.800374225236815,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 29072
    },
    {
      "epoch": 6.8006081160098235,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 29073
    },
    {
      "epoch": 6.800842006782832,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 29074
    },
    {
      "epoch": 6.801075897555841,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 29075
    },
    {
      "epoch": 6.801309788328851,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7478,
      "step": 29076
    },
    {
      "epoch": 6.80154367910186,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 29077
    },
    {
      "epoch": 6.801777569874869,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6579,
      "step": 29078
    },
    {
      "epoch": 6.802011460647877,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 29079
    },
    {
      "epoch": 6.802245351420886,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 29080
    },
    {
      "epoch": 6.802479242193895,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 29081
    },
    {
      "epoch": 6.802713132966905,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 29082
    },
    {
      "epoch": 6.802947023739914,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 29083
    },
    {
      "epoch": 6.8031809145129225,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 29084
    },
    {
      "epoch": 6.803414805285931,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6098,
      "step": 29085
    },
    {
      "epoch": 6.80364869605894,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 29086
    },
    {
      "epoch": 6.80388258683195,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 29087
    },
    {
      "epoch": 6.804116477604959,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 29088
    },
    {
      "epoch": 6.804350368377968,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 29089
    },
    {
      "epoch": 6.804584259150976,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 29090
    },
    {
      "epoch": 6.804818149923985,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.7699,
      "step": 29091
    },
    {
      "epoch": 6.805052040696994,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5734,
      "step": 29092
    },
    {
      "epoch": 6.805285931470004,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5286,
      "step": 29093
    },
    {
      "epoch": 6.805519822243013,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.857,
      "step": 29094
    },
    {
      "epoch": 6.8057537130160215,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 29095
    },
    {
      "epoch": 6.80598760378903,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 29096
    },
    {
      "epoch": 6.806221494562039,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8617,
      "step": 29097
    },
    {
      "epoch": 6.806455385335049,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 29098
    },
    {
      "epoch": 6.806689276108058,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 29099
    },
    {
      "epoch": 6.806923166881067,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 29100
    },
    {
      "epoch": 6.806923166881067,
      "eval_runtime": 4.5865,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 29100
    },
    {
      "epoch": 6.807157057654075,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7089,
      "step": 29101
    },
    {
      "epoch": 6.807390948427084,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 29102
    },
    {
      "epoch": 6.807624839200093,
      "grad_norm": 2.5,
      "learning_rate": 3e-05,
      "loss": 1.5425,
      "step": 29103
    },
    {
      "epoch": 6.807858729973103,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 29104
    },
    {
      "epoch": 6.808092620746112,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.4207,
      "step": 29105
    },
    {
      "epoch": 6.8083265115191205,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 29106
    },
    {
      "epoch": 6.808560402292129,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 29107
    },
    {
      "epoch": 6.808794293065139,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4783,
      "step": 29108
    },
    {
      "epoch": 6.809028183838148,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6857,
      "step": 29109
    },
    {
      "epoch": 6.809262074611157,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 29110
    },
    {
      "epoch": 6.809495965384166,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 29111
    },
    {
      "epoch": 6.809729856157174,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7893,
      "step": 29112
    },
    {
      "epoch": 6.809963746930183,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 29113
    },
    {
      "epoch": 6.810197637703193,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 29114
    },
    {
      "epoch": 6.810431528476202,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 29115
    },
    {
      "epoch": 6.810665419249211,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 29116
    },
    {
      "epoch": 6.8108993100222195,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 29117
    },
    {
      "epoch": 6.811133200795228,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 29118
    },
    {
      "epoch": 6.811367091568238,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7355,
      "step": 29119
    },
    {
      "epoch": 6.811600982341247,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 29120
    },
    {
      "epoch": 6.811834873114256,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6967,
      "step": 29121
    },
    {
      "epoch": 6.812068763887265,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7162,
      "step": 29122
    },
    {
      "epoch": 6.812302654660273,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 29123
    },
    {
      "epoch": 6.812536545433282,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 29124
    },
    {
      "epoch": 6.812770436206292,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 29125
    },
    {
      "epoch": 6.813004326979301,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7669,
      "step": 29126
    },
    {
      "epoch": 6.81323821775231,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 29127
    },
    {
      "epoch": 6.8134721085253185,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9622,
      "step": 29128
    },
    {
      "epoch": 6.813705999298327,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6914,
      "step": 29129
    },
    {
      "epoch": 6.813939890071337,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.748,
      "step": 29130
    },
    {
      "epoch": 6.814173780844346,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 29131
    },
    {
      "epoch": 6.814407671617355,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 29132
    },
    {
      "epoch": 6.814641562390364,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6251,
      "step": 29133
    },
    {
      "epoch": 6.814875453163372,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7267,
      "step": 29134
    },
    {
      "epoch": 6.815109343936381,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 29135
    },
    {
      "epoch": 6.815343234709391,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2802,
      "step": 29136
    },
    {
      "epoch": 6.8155771254824,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 29137
    },
    {
      "epoch": 6.815811016255409,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 29138
    },
    {
      "epoch": 6.8160449070284175,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 29139
    },
    {
      "epoch": 6.816278797801427,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6006,
      "step": 29140
    },
    {
      "epoch": 6.816512688574436,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 29141
    },
    {
      "epoch": 6.816746579347445,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 29142
    },
    {
      "epoch": 6.816980470120454,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 29143
    },
    {
      "epoch": 6.8172143608934626,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 29144
    },
    {
      "epoch": 6.817448251666471,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 29145
    },
    {
      "epoch": 6.817682142439481,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 29146
    },
    {
      "epoch": 6.81791603321249,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 29147
    },
    {
      "epoch": 6.818149923985499,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6727,
      "step": 29148
    },
    {
      "epoch": 6.818383814758508,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 29149
    },
    {
      "epoch": 6.8186177055315165,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6346,
      "step": 29150
    },
    {
      "epoch": 6.818851596304526,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 29151
    },
    {
      "epoch": 6.819085487077535,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 29152
    },
    {
      "epoch": 6.819319377850544,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 29153
    },
    {
      "epoch": 6.819553268623553,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 29154
    },
    {
      "epoch": 6.8197871593965615,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7969,
      "step": 29155
    },
    {
      "epoch": 6.82002105016957,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 29156
    },
    {
      "epoch": 6.82025494094258,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 29157
    },
    {
      "epoch": 6.820488831715589,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4731,
      "step": 29158
    },
    {
      "epoch": 6.820722722488598,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 29159
    },
    {
      "epoch": 6.820956613261607,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7576,
      "step": 29160
    },
    {
      "epoch": 6.8211905040346155,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 29161
    },
    {
      "epoch": 6.821424394807625,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9937,
      "step": 29162
    },
    {
      "epoch": 6.821658285580634,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 29163
    },
    {
      "epoch": 6.821892176353643,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.586,
      "step": 29164
    },
    {
      "epoch": 6.822126067126652,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 29165
    },
    {
      "epoch": 6.8223599578996605,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 29166
    },
    {
      "epoch": 6.822593848672669,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 29167
    },
    {
      "epoch": 6.822827739445679,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 29168
    },
    {
      "epoch": 6.823061630218688,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 29169
    },
    {
      "epoch": 6.823295520991697,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 29170
    },
    {
      "epoch": 6.823529411764706,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 29171
    },
    {
      "epoch": 6.823763302537715,
      "grad_norm": 9.4375,
      "learning_rate": 3e-05,
      "loss": 2.3495,
      "step": 29172
    },
    {
      "epoch": 6.823997193310724,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6571,
      "step": 29173
    },
    {
      "epoch": 6.824231084083733,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5575,
      "step": 29174
    },
    {
      "epoch": 6.824464974856742,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 29175
    },
    {
      "epoch": 6.824698865629751,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7598,
      "step": 29176
    },
    {
      "epoch": 6.8249327564027595,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 29177
    },
    {
      "epoch": 6.825166647175769,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7092,
      "step": 29178
    },
    {
      "epoch": 6.825400537948778,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 29179
    },
    {
      "epoch": 6.825634428721787,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7329,
      "step": 29180
    },
    {
      "epoch": 6.825868319494796,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 29181
    },
    {
      "epoch": 6.826102210267805,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7969,
      "step": 29182
    },
    {
      "epoch": 6.826336101040814,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6386,
      "step": 29183
    },
    {
      "epoch": 6.826569991813823,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 29184
    },
    {
      "epoch": 6.826803882586832,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 29185
    },
    {
      "epoch": 6.827037773359841,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 29186
    },
    {
      "epoch": 6.82727166413285,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 29187
    },
    {
      "epoch": 6.8275055549058585,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.7461,
      "step": 29188
    },
    {
      "epoch": 6.827739445678868,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6673,
      "step": 29189
    },
    {
      "epoch": 6.827973336451877,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 29190
    },
    {
      "epoch": 6.828207227224886,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 29191
    },
    {
      "epoch": 6.828441117997895,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7395,
      "step": 29192
    },
    {
      "epoch": 6.828675008770904,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 29193
    },
    {
      "epoch": 6.828908899543913,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 29194
    },
    {
      "epoch": 6.829142790316922,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 29195
    },
    {
      "epoch": 6.829376681089931,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 29196
    },
    {
      "epoch": 6.82961057186294,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 29197
    },
    {
      "epoch": 6.829844462635949,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 29198
    },
    {
      "epoch": 6.8300783534089575,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 29199
    },
    {
      "epoch": 6.830312244181967,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 29200
    },
    {
      "epoch": 6.830312244181967,
      "eval_runtime": 4.5823,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 29200
    },
    {
      "epoch": 6.830546134954976,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6074,
      "step": 29201
    },
    {
      "epoch": 6.830780025727985,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 29202
    },
    {
      "epoch": 6.831013916500994,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7769,
      "step": 29203
    },
    {
      "epoch": 6.8312478072740035,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 29204
    },
    {
      "epoch": 6.831481698047012,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6748,
      "step": 29205
    },
    {
      "epoch": 6.831715588820021,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.3972,
      "step": 29206
    },
    {
      "epoch": 6.83194947959303,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7033,
      "step": 29207
    },
    {
      "epoch": 6.832183370366039,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 29208
    },
    {
      "epoch": 6.832417261139048,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6962,
      "step": 29209
    },
    {
      "epoch": 6.832651151912057,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 29210
    },
    {
      "epoch": 6.832885042685066,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5194,
      "step": 29211
    },
    {
      "epoch": 6.833118933458075,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6761,
      "step": 29212
    },
    {
      "epoch": 6.833352824231084,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8902,
      "step": 29213
    },
    {
      "epoch": 6.833586715004093,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 29214
    },
    {
      "epoch": 6.8338206057771025,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 29215
    },
    {
      "epoch": 6.834054496550111,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6573,
      "step": 29216
    },
    {
      "epoch": 6.83428838732312,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9133,
      "step": 29217
    },
    {
      "epoch": 6.834522278096129,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 29218
    },
    {
      "epoch": 6.834756168869138,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 29219
    },
    {
      "epoch": 6.834990059642147,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.632,
      "step": 29220
    },
    {
      "epoch": 6.835223950415156,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 29221
    },
    {
      "epoch": 6.835457841188165,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 29222
    },
    {
      "epoch": 6.835691731961174,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 29223
    },
    {
      "epoch": 6.835925622734183,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 29224
    },
    {
      "epoch": 6.836159513507192,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 29225
    },
    {
      "epoch": 6.8363934042802015,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6771,
      "step": 29226
    },
    {
      "epoch": 6.83662729505321,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 29227
    },
    {
      "epoch": 6.836861185826219,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7289,
      "step": 29228
    },
    {
      "epoch": 6.837095076599228,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 29229
    },
    {
      "epoch": 6.837328967372237,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7154,
      "step": 29230
    },
    {
      "epoch": 6.837562858145247,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 29231
    },
    {
      "epoch": 6.837796748918255,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6909,
      "step": 29232
    },
    {
      "epoch": 6.838030639691264,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.652,
      "step": 29233
    },
    {
      "epoch": 6.838264530464273,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 29234
    },
    {
      "epoch": 6.838498421237282,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 29235
    },
    {
      "epoch": 6.838732312010292,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 29236
    },
    {
      "epoch": 6.8389662027833005,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 29237
    },
    {
      "epoch": 6.839200093556309,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6337,
      "step": 29238
    },
    {
      "epoch": 6.839433984329318,
      "grad_norm": 11.0625,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 29239
    },
    {
      "epoch": 6.839667875102327,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 29240
    },
    {
      "epoch": 6.839901765875336,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6298,
      "step": 29241
    },
    {
      "epoch": 6.8401356566483456,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6593,
      "step": 29242
    },
    {
      "epoch": 6.840369547421354,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.2247,
      "step": 29243
    },
    {
      "epoch": 6.840603438194363,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 29244
    },
    {
      "epoch": 6.840837328967372,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 29245
    },
    {
      "epoch": 6.841071219740381,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8061,
      "step": 29246
    },
    {
      "epoch": 6.841305110513391,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 29247
    },
    {
      "epoch": 6.8415390012863995,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 29248
    },
    {
      "epoch": 6.841772892059408,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 29249
    },
    {
      "epoch": 6.842006782832417,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 29250
    },
    {
      "epoch": 6.842240673605426,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 29251
    },
    {
      "epoch": 6.842474564378435,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 29252
    },
    {
      "epoch": 6.8427084551514445,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 29253
    },
    {
      "epoch": 6.842942345924453,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 29254
    },
    {
      "epoch": 6.843176236697462,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6222,
      "step": 29255
    },
    {
      "epoch": 6.843410127470471,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 29256
    },
    {
      "epoch": 6.843644018243481,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 29257
    },
    {
      "epoch": 6.84387790901649,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 29258
    },
    {
      "epoch": 6.8441117997894985,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 29259
    },
    {
      "epoch": 6.844345690562507,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2689,
      "step": 29260
    },
    {
      "epoch": 6.844579581335516,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 29261
    },
    {
      "epoch": 6.844813472108525,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 29262
    },
    {
      "epoch": 6.845047362881535,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 29263
    },
    {
      "epoch": 6.8452812536545435,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7794,
      "step": 29264
    },
    {
      "epoch": 6.845515144427552,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 29265
    },
    {
      "epoch": 6.845749035200561,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7662,
      "step": 29266
    },
    {
      "epoch": 6.84598292597357,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 29267
    },
    {
      "epoch": 6.84621681674658,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 29268
    },
    {
      "epoch": 6.846450707519589,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6913,
      "step": 29269
    },
    {
      "epoch": 6.8466845982925975,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6267,
      "step": 29270
    },
    {
      "epoch": 6.846918489065606,
      "grad_norm": 10.625,
      "learning_rate": 3e-05,
      "loss": 2.4034,
      "step": 29271
    },
    {
      "epoch": 6.847152379838615,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 29272
    },
    {
      "epoch": 6.847386270611624,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 29273
    },
    {
      "epoch": 6.847620161384634,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 29274
    },
    {
      "epoch": 6.8478540521576425,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 29275
    },
    {
      "epoch": 6.848087942930651,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 29276
    },
    {
      "epoch": 6.84832183370366,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5797,
      "step": 29277
    },
    {
      "epoch": 6.848555724476669,
      "grad_norm": 7.9375,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 29278
    },
    {
      "epoch": 6.848789615249679,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6073,
      "step": 29279
    },
    {
      "epoch": 6.849023506022688,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 29280
    },
    {
      "epoch": 6.8492573967956965,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.676,
      "step": 29281
    },
    {
      "epoch": 6.849491287568705,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7047,
      "step": 29282
    },
    {
      "epoch": 6.849725178341714,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 29283
    },
    {
      "epoch": 6.849959069114723,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.3603,
      "step": 29284
    },
    {
      "epoch": 6.850192959887733,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 29285
    },
    {
      "epoch": 6.8504268506607415,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7862,
      "step": 29286
    },
    {
      "epoch": 6.85066074143375,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6117,
      "step": 29287
    },
    {
      "epoch": 6.850894632206759,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 29288
    },
    {
      "epoch": 6.851128522979769,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 29289
    },
    {
      "epoch": 6.851362413752778,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 29290
    },
    {
      "epoch": 6.851596304525787,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 29291
    },
    {
      "epoch": 6.8518301952987954,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 29292
    },
    {
      "epoch": 6.852064086071804,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 29293
    },
    {
      "epoch": 6.852297976844813,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6819,
      "step": 29294
    },
    {
      "epoch": 6.852531867617823,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 29295
    },
    {
      "epoch": 6.852765758390832,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6477,
      "step": 29296
    },
    {
      "epoch": 6.8529996491638405,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 29297
    },
    {
      "epoch": 6.853233539936849,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 29298
    },
    {
      "epoch": 6.853467430709858,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 29299
    },
    {
      "epoch": 6.853701321482868,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7038,
      "step": 29300
    },
    {
      "epoch": 6.853701321482868,
      "eval_runtime": 4.6576,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 29300
    },
    {
      "epoch": 6.853935212255877,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 29301
    },
    {
      "epoch": 6.854169103028886,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.687,
      "step": 29302
    },
    {
      "epoch": 6.854402993801894,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8711,
      "step": 29303
    },
    {
      "epoch": 6.854636884574903,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 29304
    },
    {
      "epoch": 6.854870775347912,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6275,
      "step": 29305
    },
    {
      "epoch": 6.855104666120922,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 29306
    },
    {
      "epoch": 6.855338556893931,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.679,
      "step": 29307
    },
    {
      "epoch": 6.8555724476669395,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 29308
    },
    {
      "epoch": 6.855806338439948,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 29309
    },
    {
      "epoch": 6.856040229212957,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 29310
    },
    {
      "epoch": 6.856274119985967,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 29311
    },
    {
      "epoch": 6.856508010758976,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 29312
    },
    {
      "epoch": 6.856741901531985,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8577,
      "step": 29313
    },
    {
      "epoch": 6.856975792304993,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 29314
    },
    {
      "epoch": 6.857209683078002,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 29315
    },
    {
      "epoch": 6.857443573851011,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 29316
    },
    {
      "epoch": 6.857677464624021,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 29317
    },
    {
      "epoch": 6.85791135539703,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8532,
      "step": 29318
    },
    {
      "epoch": 6.8581452461700385,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 29319
    },
    {
      "epoch": 6.858379136943047,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 29320
    },
    {
      "epoch": 6.858613027716057,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 29321
    },
    {
      "epoch": 6.858846918489066,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 29322
    },
    {
      "epoch": 6.859080809262075,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 29323
    },
    {
      "epoch": 6.859314700035084,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7866,
      "step": 29324
    },
    {
      "epoch": 6.859548590808092,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 29325
    },
    {
      "epoch": 6.859782481581101,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0932,
      "step": 29326
    },
    {
      "epoch": 6.860016372354111,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 29327
    },
    {
      "epoch": 6.86025026312712,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 29328
    },
    {
      "epoch": 6.860484153900129,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7655,
      "step": 29329
    },
    {
      "epoch": 6.8607180446731375,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6263,
      "step": 29330
    },
    {
      "epoch": 6.860951935446146,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 29331
    },
    {
      "epoch": 6.861185826219156,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 29332
    },
    {
      "epoch": 6.861419716992165,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 29333
    },
    {
      "epoch": 6.861653607765174,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 29334
    },
    {
      "epoch": 6.861887498538183,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 29335
    },
    {
      "epoch": 6.862121389311191,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 29336
    },
    {
      "epoch": 6.8623552800842,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7254,
      "step": 29337
    },
    {
      "epoch": 6.86258917085721,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 29338
    },
    {
      "epoch": 6.862823061630219,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 29339
    },
    {
      "epoch": 6.863056952403228,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 29340
    },
    {
      "epoch": 6.8632908431762365,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 29341
    },
    {
      "epoch": 6.863524733949245,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6056,
      "step": 29342
    },
    {
      "epoch": 6.863758624722255,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8002,
      "step": 29343
    },
    {
      "epoch": 6.863992515495264,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 29344
    },
    {
      "epoch": 6.864226406268273,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 29345
    },
    {
      "epoch": 6.864460297041282,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 29346
    },
    {
      "epoch": 6.86469418781429,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9998,
      "step": 29347
    },
    {
      "epoch": 6.864928078587299,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7186,
      "step": 29348
    },
    {
      "epoch": 6.865161969360309,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 29349
    },
    {
      "epoch": 6.865395860133318,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 29350
    },
    {
      "epoch": 6.865629750906327,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 29351
    },
    {
      "epoch": 6.8658636416793355,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 29352
    },
    {
      "epoch": 6.866097532452345,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 29353
    },
    {
      "epoch": 6.866331423225354,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 29354
    },
    {
      "epoch": 6.866565313998363,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 29355
    },
    {
      "epoch": 6.866799204771372,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7777,
      "step": 29356
    },
    {
      "epoch": 6.867033095544381,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5803,
      "step": 29357
    },
    {
      "epoch": 6.867266986317389,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 29358
    },
    {
      "epoch": 6.867500877090399,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 29359
    },
    {
      "epoch": 6.867734767863408,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 29360
    },
    {
      "epoch": 6.867968658636417,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 29361
    },
    {
      "epoch": 6.868202549409426,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 29362
    },
    {
      "epoch": 6.8684364401824345,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7343,
      "step": 29363
    },
    {
      "epoch": 6.868670330955444,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6929,
      "step": 29364
    },
    {
      "epoch": 6.868904221728453,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 29365
    },
    {
      "epoch": 6.869138112501462,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4566,
      "step": 29366
    },
    {
      "epoch": 6.869372003274471,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 29367
    },
    {
      "epoch": 6.86960589404748,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 29368
    },
    {
      "epoch": 6.869839784820488,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.761,
      "step": 29369
    },
    {
      "epoch": 6.870073675593498,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 29370
    },
    {
      "epoch": 6.870307566366507,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.5692,
      "step": 29371
    },
    {
      "epoch": 6.870541457139516,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 29372
    },
    {
      "epoch": 6.870775347912525,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 29373
    },
    {
      "epoch": 6.8710092386855335,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 29374
    },
    {
      "epoch": 6.871243129458543,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 29375
    },
    {
      "epoch": 6.871477020231552,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7819,
      "step": 29376
    },
    {
      "epoch": 6.871710911004561,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 29377
    },
    {
      "epoch": 6.87194480177757,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6938,
      "step": 29378
    },
    {
      "epoch": 6.872178692550579,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 29379
    },
    {
      "epoch": 6.872412583323587,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 29380
    },
    {
      "epoch": 6.872646474096597,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 29381
    },
    {
      "epoch": 6.872880364869606,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 29382
    },
    {
      "epoch": 6.873114255642615,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7656,
      "step": 29383
    },
    {
      "epoch": 6.873348146415624,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9262,
      "step": 29384
    },
    {
      "epoch": 6.873582037188633,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 29385
    },
    {
      "epoch": 6.873815927961642,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7178,
      "step": 29386
    },
    {
      "epoch": 6.874049818734651,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7198,
      "step": 29387
    },
    {
      "epoch": 6.87428370950766,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 29388
    },
    {
      "epoch": 6.874517600280669,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4255,
      "step": 29389
    },
    {
      "epoch": 6.8747514910536776,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 29390
    },
    {
      "epoch": 6.874985381826687,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6032,
      "step": 29391
    },
    {
      "epoch": 6.875219272599696,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 29392
    },
    {
      "epoch": 6.875453163372705,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 29393
    },
    {
      "epoch": 6.875687054145714,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 29394
    },
    {
      "epoch": 6.875920944918723,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 29395
    },
    {
      "epoch": 6.876154835691732,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 29396
    },
    {
      "epoch": 6.876388726464741,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 29397
    },
    {
      "epoch": 6.87662261723775,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 29398
    },
    {
      "epoch": 6.876856508010759,
      "grad_norm": 2.453125,
      "learning_rate": 3e-05,
      "loss": 1.3703,
      "step": 29399
    },
    {
      "epoch": 6.877090398783768,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 29400
    },
    {
      "epoch": 6.877090398783768,
      "eval_runtime": 4.6004,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 29400
    },
    {
      "epoch": 6.8773242895567765,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 29401
    },
    {
      "epoch": 6.877558180329786,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5666,
      "step": 29402
    },
    {
      "epoch": 6.877792071102795,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 29403
    },
    {
      "epoch": 6.878025961875804,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.686,
      "step": 29404
    },
    {
      "epoch": 6.878259852648813,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 29405
    },
    {
      "epoch": 6.878493743421822,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7001,
      "step": 29406
    },
    {
      "epoch": 6.878727634194831,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 29407
    },
    {
      "epoch": 6.87896152496784,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7733,
      "step": 29408
    },
    {
      "epoch": 6.879195415740849,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7327,
      "step": 29409
    },
    {
      "epoch": 6.879429306513858,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8187,
      "step": 29410
    },
    {
      "epoch": 6.879663197286867,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 29411
    },
    {
      "epoch": 6.8798970880598755,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.4978,
      "step": 29412
    },
    {
      "epoch": 6.880130978832885,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7923,
      "step": 29413
    },
    {
      "epoch": 6.880364869605894,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 29414
    },
    {
      "epoch": 6.880598760378903,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 29415
    },
    {
      "epoch": 6.880832651151912,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6674,
      "step": 29416
    },
    {
      "epoch": 6.8810665419249215,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 29417
    },
    {
      "epoch": 6.88130043269793,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 29418
    },
    {
      "epoch": 6.881534323470939,
      "grad_norm": 10.75,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 29419
    },
    {
      "epoch": 6.881768214243948,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 29420
    },
    {
      "epoch": 6.882002105016957,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6146,
      "step": 29421
    },
    {
      "epoch": 6.882235995789966,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6048,
      "step": 29422
    },
    {
      "epoch": 6.882469886562975,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 29423
    },
    {
      "epoch": 6.882703777335984,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 29424
    },
    {
      "epoch": 6.882937668108993,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5971,
      "step": 29425
    },
    {
      "epoch": 6.883171558882002,
      "grad_norm": 11.5625,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 29426
    },
    {
      "epoch": 6.883405449655011,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 29427
    },
    {
      "epoch": 6.8836393404280205,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 29428
    },
    {
      "epoch": 6.883873231201029,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.682,
      "step": 29429
    },
    {
      "epoch": 6.884107121974038,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 29430
    },
    {
      "epoch": 6.884341012747047,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 29431
    },
    {
      "epoch": 6.884574903520056,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 29432
    },
    {
      "epoch": 6.884808794293065,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 29433
    },
    {
      "epoch": 6.885042685066074,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8151,
      "step": 29434
    },
    {
      "epoch": 6.885276575839083,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 29435
    },
    {
      "epoch": 6.885510466612092,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 29436
    },
    {
      "epoch": 6.885744357385101,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9824,
      "step": 29437
    },
    {
      "epoch": 6.885978248158111,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 29438
    },
    {
      "epoch": 6.8862121389311195,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 29439
    },
    {
      "epoch": 6.886446029704128,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 29440
    },
    {
      "epoch": 6.886679920477137,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6925,
      "step": 29441
    },
    {
      "epoch": 6.886913811250146,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 29442
    },
    {
      "epoch": 6.887147702023155,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 29443
    },
    {
      "epoch": 6.887381592796165,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 29444
    },
    {
      "epoch": 6.887615483569173,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 29445
    },
    {
      "epoch": 6.887849374342182,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 29446
    },
    {
      "epoch": 6.888083265115191,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 29447
    },
    {
      "epoch": 6.8883171558882,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 29448
    },
    {
      "epoch": 6.88855104666121,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 29449
    },
    {
      "epoch": 6.8887849374342185,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 29450
    },
    {
      "epoch": 6.889018828207227,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 29451
    },
    {
      "epoch": 6.889252718980236,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2892,
      "step": 29452
    },
    {
      "epoch": 6.889486609753245,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 29453
    },
    {
      "epoch": 6.889720500526254,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 29454
    },
    {
      "epoch": 6.889954391299264,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 29455
    },
    {
      "epoch": 6.890188282072272,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 29456
    },
    {
      "epoch": 6.890422172845281,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 29457
    },
    {
      "epoch": 6.89065606361829,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5854,
      "step": 29458
    },
    {
      "epoch": 6.890889954391299,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7175,
      "step": 29459
    },
    {
      "epoch": 6.891123845164309,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 29460
    },
    {
      "epoch": 6.8913577359373175,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.444,
      "step": 29461
    },
    {
      "epoch": 6.891591626710326,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 29462
    },
    {
      "epoch": 6.891825517483335,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6701,
      "step": 29463
    },
    {
      "epoch": 6.892059408256344,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6065,
      "step": 29464
    },
    {
      "epoch": 6.892293299029353,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 29465
    },
    {
      "epoch": 6.892527189802363,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5885,
      "step": 29466
    },
    {
      "epoch": 6.892761080575371,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 29467
    },
    {
      "epoch": 6.89299497134838,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 29468
    },
    {
      "epoch": 6.893228862121389,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 29469
    },
    {
      "epoch": 6.893462752894399,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 29470
    },
    {
      "epoch": 6.893696643667408,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 29471
    },
    {
      "epoch": 6.8939305344404165,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 29472
    },
    {
      "epoch": 6.894164425213425,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 29473
    },
    {
      "epoch": 6.894398315986434,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7757,
      "step": 29474
    },
    {
      "epoch": 6.894632206759443,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 29475
    },
    {
      "epoch": 6.894866097532453,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8135,
      "step": 29476
    },
    {
      "epoch": 6.895099988305462,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 29477
    },
    {
      "epoch": 6.89533387907847,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7184,
      "step": 29478
    },
    {
      "epoch": 6.895567769851479,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 29479
    },
    {
      "epoch": 6.895801660624488,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5784,
      "step": 29480
    },
    {
      "epoch": 6.896035551397498,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 29481
    },
    {
      "epoch": 6.896269442170507,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 29482
    },
    {
      "epoch": 6.8965033329435155,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 29483
    },
    {
      "epoch": 6.896737223716524,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 29484
    },
    {
      "epoch": 6.896971114489533,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.717,
      "step": 29485
    },
    {
      "epoch": 6.897205005262542,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 29486
    },
    {
      "epoch": 6.897438896035552,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5922,
      "step": 29487
    },
    {
      "epoch": 6.8976727868085606,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 29488
    },
    {
      "epoch": 6.897906677581569,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 29489
    },
    {
      "epoch": 6.898140568354578,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 29490
    },
    {
      "epoch": 6.898374459127587,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 29491
    },
    {
      "epoch": 6.898608349900597,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 29492
    },
    {
      "epoch": 6.898842240673606,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 29493
    },
    {
      "epoch": 6.8990761314466145,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 29494
    },
    {
      "epoch": 6.899310022219623,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 29495
    },
    {
      "epoch": 6.899543912992632,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3521,
      "step": 29496
    },
    {
      "epoch": 6.899777803765641,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.545,
      "step": 29497
    },
    {
      "epoch": 6.900011694538651,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 29498
    },
    {
      "epoch": 6.9002455853116595,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 29499
    },
    {
      "epoch": 6.900479476084668,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 29500
    },
    {
      "epoch": 6.900479476084668,
      "eval_runtime": 4.6359,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 29500
    },
    {
      "epoch": 6.900713366857677,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 29501
    },
    {
      "epoch": 6.900947257630687,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 29502
    },
    {
      "epoch": 6.901181148403696,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 29503
    },
    {
      "epoch": 6.901415039176705,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2963,
      "step": 29504
    },
    {
      "epoch": 6.9016489299497135,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 29505
    },
    {
      "epoch": 6.901882820722722,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 29506
    },
    {
      "epoch": 6.902116711495731,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 29507
    },
    {
      "epoch": 6.902350602268741,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6177,
      "step": 29508
    },
    {
      "epoch": 6.90258449304175,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 29509
    },
    {
      "epoch": 6.9028183838147585,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 29510
    },
    {
      "epoch": 6.903052274587767,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 29511
    },
    {
      "epoch": 6.903286165360776,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7423,
      "step": 29512
    },
    {
      "epoch": 6.903520056133786,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5987,
      "step": 29513
    },
    {
      "epoch": 6.903753946906795,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.685,
      "step": 29514
    },
    {
      "epoch": 6.903987837679804,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 29515
    },
    {
      "epoch": 6.9042217284528125,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4939,
      "step": 29516
    },
    {
      "epoch": 6.904455619225821,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6459,
      "step": 29517
    },
    {
      "epoch": 6.90468950999883,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 29518
    },
    {
      "epoch": 6.90492340077184,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 29519
    },
    {
      "epoch": 6.905157291544849,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 29520
    },
    {
      "epoch": 6.9053911823178575,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 29521
    },
    {
      "epoch": 6.905625073090866,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 29522
    },
    {
      "epoch": 6.905858963863875,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 29523
    },
    {
      "epoch": 6.906092854636885,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 29524
    },
    {
      "epoch": 6.906326745409894,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.4964,
      "step": 29525
    },
    {
      "epoch": 6.906560636182903,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 29526
    },
    {
      "epoch": 6.9067945269559115,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 29527
    },
    {
      "epoch": 6.90702841772892,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 29528
    },
    {
      "epoch": 6.907262308501929,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 29529
    },
    {
      "epoch": 6.907496199274939,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7123,
      "step": 29530
    },
    {
      "epoch": 6.907730090047948,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 29531
    },
    {
      "epoch": 6.9079639808209565,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 29532
    },
    {
      "epoch": 6.908197871593965,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6698,
      "step": 29533
    },
    {
      "epoch": 6.908431762366975,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6255,
      "step": 29534
    },
    {
      "epoch": 6.908665653139984,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 29535
    },
    {
      "epoch": 6.908899543912993,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6981,
      "step": 29536
    },
    {
      "epoch": 6.909133434686002,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 29537
    },
    {
      "epoch": 6.9093673254590104,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 29538
    },
    {
      "epoch": 6.909601216232019,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 29539
    },
    {
      "epoch": 6.909835107005029,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7298,
      "step": 29540
    },
    {
      "epoch": 6.910068997778038,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4404,
      "step": 29541
    },
    {
      "epoch": 6.910302888551047,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 29542
    },
    {
      "epoch": 6.9105367793240555,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 29543
    },
    {
      "epoch": 6.910770670097064,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 29544
    },
    {
      "epoch": 6.911004560870074,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.752,
      "step": 29545
    },
    {
      "epoch": 6.911238451643083,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8532,
      "step": 29546
    },
    {
      "epoch": 6.911472342416092,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 29547
    },
    {
      "epoch": 6.911706233189101,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 29548
    },
    {
      "epoch": 6.911940123962109,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5916,
      "step": 29549
    },
    {
      "epoch": 6.912174014735118,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5344,
      "step": 29550
    },
    {
      "epoch": 6.912407905508128,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 29551
    },
    {
      "epoch": 6.912641796281137,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 29552
    },
    {
      "epoch": 6.912875687054146,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7166,
      "step": 29553
    },
    {
      "epoch": 6.9131095778271545,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.3647,
      "step": 29554
    },
    {
      "epoch": 6.913343468600163,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 29555
    },
    {
      "epoch": 6.913577359373173,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 29556
    },
    {
      "epoch": 6.913811250146182,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7271,
      "step": 29557
    },
    {
      "epoch": 6.914045140919191,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 29558
    },
    {
      "epoch": 6.9142790316922,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8748,
      "step": 29559
    },
    {
      "epoch": 6.914512922465208,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 29560
    },
    {
      "epoch": 6.914746813238217,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.6946,
      "step": 29561
    },
    {
      "epoch": 6.914980704011227,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 29562
    },
    {
      "epoch": 6.915214594784236,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.513,
      "step": 29563
    },
    {
      "epoch": 6.915448485557245,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5702,
      "step": 29564
    },
    {
      "epoch": 6.9156823763302535,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 29565
    },
    {
      "epoch": 6.915916267103263,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 29566
    },
    {
      "epoch": 6.916150157876272,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 29567
    },
    {
      "epoch": 6.916384048649281,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 29568
    },
    {
      "epoch": 6.91661793942229,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 29569
    },
    {
      "epoch": 6.916851830195299,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7714,
      "step": 29570
    },
    {
      "epoch": 6.917085720968307,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.3681,
      "step": 29571
    },
    {
      "epoch": 6.917319611741317,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.591,
      "step": 29572
    },
    {
      "epoch": 6.917553502514326,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 29573
    },
    {
      "epoch": 6.917787393287335,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 29574
    },
    {
      "epoch": 6.918021284060344,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1948,
      "step": 29575
    },
    {
      "epoch": 6.9182551748333525,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6522,
      "step": 29576
    },
    {
      "epoch": 6.918489065606362,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6274,
      "step": 29577
    },
    {
      "epoch": 6.918722956379371,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 29578
    },
    {
      "epoch": 6.91895684715238,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7826,
      "step": 29579
    },
    {
      "epoch": 6.919190737925389,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 29580
    },
    {
      "epoch": 6.919424628698398,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2182,
      "step": 29581
    },
    {
      "epoch": 6.919658519471406,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 29582
    },
    {
      "epoch": 6.919892410244416,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 29583
    },
    {
      "epoch": 6.920126301017425,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6125,
      "step": 29584
    },
    {
      "epoch": 6.920360191790434,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 29585
    },
    {
      "epoch": 6.920594082563443,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8303,
      "step": 29586
    },
    {
      "epoch": 6.9208279733364515,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 29587
    },
    {
      "epoch": 6.921061864109461,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 29588
    },
    {
      "epoch": 6.92129575488247,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6778,
      "step": 29589
    },
    {
      "epoch": 6.921529645655479,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 29590
    },
    {
      "epoch": 6.921763536428488,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 29591
    },
    {
      "epoch": 6.921997427201497,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7095,
      "step": 29592
    },
    {
      "epoch": 6.922231317974505,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 29593
    },
    {
      "epoch": 6.922465208747515,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6937,
      "step": 29594
    },
    {
      "epoch": 6.922699099520524,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 29595
    },
    {
      "epoch": 6.922932990293533,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6735,
      "step": 29596
    },
    {
      "epoch": 6.923166881066542,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 29597
    },
    {
      "epoch": 6.923400771839551,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 29598
    },
    {
      "epoch": 6.92363466261256,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 29599
    },
    {
      "epoch": 6.923868553385569,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 29600
    },
    {
      "epoch": 6.923868553385569,
      "eval_runtime": 4.6449,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 29600
    },
    {
      "epoch": 6.924102444158578,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5977,
      "step": 29601
    },
    {
      "epoch": 6.924336334931587,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 29602
    },
    {
      "epoch": 6.924570225704596,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 29603
    },
    {
      "epoch": 6.924804116477605,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 29604
    },
    {
      "epoch": 6.925038007250614,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7729,
      "step": 29605
    },
    {
      "epoch": 6.925271898023623,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 29606
    },
    {
      "epoch": 6.925505788796632,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6005,
      "step": 29607
    },
    {
      "epoch": 6.925739679569641,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.543,
      "step": 29608
    },
    {
      "epoch": 6.92597357034265,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 29609
    },
    {
      "epoch": 6.926207461115659,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 29610
    },
    {
      "epoch": 6.926441351888668,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6631,
      "step": 29611
    },
    {
      "epoch": 6.926675242661677,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6784,
      "step": 29612
    },
    {
      "epoch": 6.926909133434686,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 29613
    },
    {
      "epoch": 6.927143024207695,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5543,
      "step": 29614
    },
    {
      "epoch": 6.927376914980704,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 29615
    },
    {
      "epoch": 6.927610805753713,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 29616
    },
    {
      "epoch": 6.927844696526722,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7448,
      "step": 29617
    },
    {
      "epoch": 6.928078587299731,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 29618
    },
    {
      "epoch": 6.92831247807274,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5987,
      "step": 29619
    },
    {
      "epoch": 6.928546368845749,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8751,
      "step": 29620
    },
    {
      "epoch": 6.928780259618758,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 29621
    },
    {
      "epoch": 6.929014150391767,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 29622
    },
    {
      "epoch": 6.929248041164776,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 29623
    },
    {
      "epoch": 6.929481931937785,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5769,
      "step": 29624
    },
    {
      "epoch": 6.929715822710794,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7453,
      "step": 29625
    },
    {
      "epoch": 6.929949713483803,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 29626
    },
    {
      "epoch": 6.930183604256812,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4667,
      "step": 29627
    },
    {
      "epoch": 6.930417495029821,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 29628
    },
    {
      "epoch": 6.93065138580283,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 29629
    },
    {
      "epoch": 6.9308852765758395,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 29630
    },
    {
      "epoch": 6.931119167348848,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 29631
    },
    {
      "epoch": 6.931353058121857,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 29632
    },
    {
      "epoch": 6.931586948894866,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 29633
    },
    {
      "epoch": 6.931820839667875,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.72,
      "step": 29634
    },
    {
      "epoch": 6.932054730440884,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 29635
    },
    {
      "epoch": 6.932288621213893,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5868,
      "step": 29636
    },
    {
      "epoch": 6.932522511986902,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5071,
      "step": 29637
    },
    {
      "epoch": 6.932756402759911,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7269,
      "step": 29638
    },
    {
      "epoch": 6.93299029353292,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 29639
    },
    {
      "epoch": 6.933224184305929,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 29640
    },
    {
      "epoch": 6.9334580750789385,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6547,
      "step": 29641
    },
    {
      "epoch": 6.933691965851947,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 29642
    },
    {
      "epoch": 6.933925856624956,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 29643
    },
    {
      "epoch": 6.934159747397965,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 29644
    },
    {
      "epoch": 6.934393638170974,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.672,
      "step": 29645
    },
    {
      "epoch": 6.934627528943983,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6466,
      "step": 29646
    },
    {
      "epoch": 6.934861419716992,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 29647
    },
    {
      "epoch": 6.935095310490001,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 29648
    },
    {
      "epoch": 6.93532920126301,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 29649
    },
    {
      "epoch": 6.935563092036019,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 29650
    },
    {
      "epoch": 6.935796982809029,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 29651
    },
    {
      "epoch": 6.9360308735820375,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7318,
      "step": 29652
    },
    {
      "epoch": 6.936264764355046,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4647,
      "step": 29653
    },
    {
      "epoch": 6.936498655128055,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5652,
      "step": 29654
    },
    {
      "epoch": 6.936732545901064,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 29655
    },
    {
      "epoch": 6.936966436674073,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 29656
    },
    {
      "epoch": 6.937200327447083,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6143,
      "step": 29657
    },
    {
      "epoch": 6.937434218220091,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4513,
      "step": 29658
    },
    {
      "epoch": 6.9376681089931,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5964,
      "step": 29659
    },
    {
      "epoch": 6.937901999766109,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8304,
      "step": 29660
    },
    {
      "epoch": 6.938135890539118,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7391,
      "step": 29661
    },
    {
      "epoch": 6.938369781312128,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 29662
    },
    {
      "epoch": 6.9386036720851365,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 29663
    },
    {
      "epoch": 6.938837562858145,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 29664
    },
    {
      "epoch": 6.939071453631154,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6114,
      "step": 29665
    },
    {
      "epoch": 6.939305344404163,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 29666
    },
    {
      "epoch": 6.939539235177172,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5532,
      "step": 29667
    },
    {
      "epoch": 6.939773125950182,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 29668
    },
    {
      "epoch": 6.94000701672319,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 29669
    },
    {
      "epoch": 6.940240907496199,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 29670
    },
    {
      "epoch": 6.940474798269208,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 29671
    },
    {
      "epoch": 6.940708689042217,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 29672
    },
    {
      "epoch": 6.940942579815227,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 29673
    },
    {
      "epoch": 6.9411764705882355,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 29674
    },
    {
      "epoch": 6.941410361361244,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 29675
    },
    {
      "epoch": 6.941644252134253,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 29676
    },
    {
      "epoch": 6.941878142907262,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.634,
      "step": 29677
    },
    {
      "epoch": 6.942112033680271,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 29678
    },
    {
      "epoch": 6.942345924453281,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 29679
    },
    {
      "epoch": 6.942579815226289,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4539,
      "step": 29680
    },
    {
      "epoch": 6.942813705999298,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5197,
      "step": 29681
    },
    {
      "epoch": 6.943047596772307,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 29682
    },
    {
      "epoch": 6.943281487545317,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 29683
    },
    {
      "epoch": 6.943515378318326,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.3603,
      "step": 29684
    },
    {
      "epoch": 6.9437492690913345,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3014,
      "step": 29685
    },
    {
      "epoch": 6.943983159864343,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 29686
    },
    {
      "epoch": 6.944217050637352,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 29687
    },
    {
      "epoch": 6.944450941410361,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7209,
      "step": 29688
    },
    {
      "epoch": 6.944684832183371,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7866,
      "step": 29689
    },
    {
      "epoch": 6.94491872295638,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5057,
      "step": 29690
    },
    {
      "epoch": 6.945152613729388,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 29691
    },
    {
      "epoch": 6.945386504502397,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 29692
    },
    {
      "epoch": 6.945620395275406,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 29693
    },
    {
      "epoch": 6.945854286048416,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 29694
    },
    {
      "epoch": 6.946088176821425,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 29695
    },
    {
      "epoch": 6.9463220675944335,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7205,
      "step": 29696
    },
    {
      "epoch": 6.946555958367442,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 29697
    },
    {
      "epoch": 6.946789849140451,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6718,
      "step": 29698
    },
    {
      "epoch": 6.94702373991346,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 29699
    },
    {
      "epoch": 6.94725763068647,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 29700
    },
    {
      "epoch": 6.94725763068647,
      "eval_runtime": 4.6651,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 29700
    },
    {
      "epoch": 6.947491521459479,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 29701
    },
    {
      "epoch": 6.947725412232487,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6657,
      "step": 29702
    },
    {
      "epoch": 6.947959303005496,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.69,
      "step": 29703
    },
    {
      "epoch": 6.948193193778505,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6487,
      "step": 29704
    },
    {
      "epoch": 6.948427084551515,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 29705
    },
    {
      "epoch": 6.948660975324524,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6318,
      "step": 29706
    },
    {
      "epoch": 6.9488948660975325,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 29707
    },
    {
      "epoch": 6.949128756870541,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 29708
    },
    {
      "epoch": 6.94936264764355,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 29709
    },
    {
      "epoch": 6.949596538416559,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6479,
      "step": 29710
    },
    {
      "epoch": 6.949830429189569,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 29711
    },
    {
      "epoch": 6.950064319962578,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 29712
    },
    {
      "epoch": 6.950298210735586,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 29713
    },
    {
      "epoch": 6.950532101508595,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 29714
    },
    {
      "epoch": 6.950765992281605,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 29715
    },
    {
      "epoch": 6.950999883054614,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 29716
    },
    {
      "epoch": 6.951233773827623,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 29717
    },
    {
      "epoch": 6.9514676646006315,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 29718
    },
    {
      "epoch": 6.95170155537364,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 29719
    },
    {
      "epoch": 6.951935446146649,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 29720
    },
    {
      "epoch": 6.952169336919659,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 29721
    },
    {
      "epoch": 6.952403227692668,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 29722
    },
    {
      "epoch": 6.952637118465677,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 29723
    },
    {
      "epoch": 6.952871009238685,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5924,
      "step": 29724
    },
    {
      "epoch": 6.953104900011694,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 29725
    },
    {
      "epoch": 6.953338790784704,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 29726
    },
    {
      "epoch": 6.953572681557713,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 29727
    },
    {
      "epoch": 6.953806572330722,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 29728
    },
    {
      "epoch": 6.9540404631037305,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 29729
    },
    {
      "epoch": 6.954274353876739,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 29730
    },
    {
      "epoch": 6.954508244649748,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 29731
    },
    {
      "epoch": 6.954742135422758,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 29732
    },
    {
      "epoch": 6.954976026195767,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 29733
    },
    {
      "epoch": 6.9552099169687756,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.5715,
      "step": 29734
    },
    {
      "epoch": 6.955443807741784,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7292,
      "step": 29735
    },
    {
      "epoch": 6.955677698514793,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 29736
    },
    {
      "epoch": 6.955911589287803,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5821,
      "step": 29737
    },
    {
      "epoch": 6.956145480060812,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 29738
    },
    {
      "epoch": 6.956379370833821,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 29739
    },
    {
      "epoch": 6.9566132616068295,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7729,
      "step": 29740
    },
    {
      "epoch": 6.956847152379838,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 29741
    },
    {
      "epoch": 6.957081043152847,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6898,
      "step": 29742
    },
    {
      "epoch": 6.957314933925857,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 29743
    },
    {
      "epoch": 6.957548824698866,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 29744
    },
    {
      "epoch": 6.9577827154718745,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 29745
    },
    {
      "epoch": 6.958016606244883,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7084,
      "step": 29746
    },
    {
      "epoch": 6.958250497017893,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 29747
    },
    {
      "epoch": 6.958484387790902,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 29748
    },
    {
      "epoch": 6.958718278563911,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 29749
    },
    {
      "epoch": 6.95895216933692,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 29750
    },
    {
      "epoch": 6.9591860601099285,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 29751
    },
    {
      "epoch": 6.959419950882937,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 29752
    },
    {
      "epoch": 6.959653841655947,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.3755,
      "step": 29753
    },
    {
      "epoch": 6.959887732428956,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 29754
    },
    {
      "epoch": 6.960121623201965,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 29755
    },
    {
      "epoch": 6.9603555139749735,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 29756
    },
    {
      "epoch": 6.960589404747982,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.479,
      "step": 29757
    },
    {
      "epoch": 6.960823295520992,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6743,
      "step": 29758
    },
    {
      "epoch": 6.961057186294001,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 29759
    },
    {
      "epoch": 6.96129107706701,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5776,
      "step": 29760
    },
    {
      "epoch": 6.961524967840019,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 29761
    },
    {
      "epoch": 6.9617588586130275,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.679,
      "step": 29762
    },
    {
      "epoch": 6.961992749386036,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 29763
    },
    {
      "epoch": 6.962226640159046,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 29764
    },
    {
      "epoch": 6.962460530932055,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 29765
    },
    {
      "epoch": 6.962694421705064,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 29766
    },
    {
      "epoch": 6.9629283124780725,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.686,
      "step": 29767
    },
    {
      "epoch": 6.963162203251081,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 29768
    },
    {
      "epoch": 6.963396094024091,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 29769
    },
    {
      "epoch": 6.9636299847971,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.3754,
      "step": 29770
    },
    {
      "epoch": 6.963863875570109,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 29771
    },
    {
      "epoch": 6.964097766343118,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 29772
    },
    {
      "epoch": 6.9643316571161265,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 29773
    },
    {
      "epoch": 6.964565547889135,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 29774
    },
    {
      "epoch": 6.964799438662145,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 29775
    },
    {
      "epoch": 6.965033329435154,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 29776
    },
    {
      "epoch": 6.965267220208163,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 29777
    },
    {
      "epoch": 6.9655011109811715,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 29778
    },
    {
      "epoch": 6.965735001754181,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7364,
      "step": 29779
    },
    {
      "epoch": 6.96596889252719,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6345,
      "step": 29780
    },
    {
      "epoch": 6.966202783300199,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 29781
    },
    {
      "epoch": 6.966436674073208,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 29782
    },
    {
      "epoch": 6.966670564846217,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 29783
    },
    {
      "epoch": 6.9669044556192254,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7933,
      "step": 29784
    },
    {
      "epoch": 6.967138346392235,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.5192,
      "step": 29785
    },
    {
      "epoch": 6.967372237165244,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 29786
    },
    {
      "epoch": 6.967606127938253,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8945,
      "step": 29787
    },
    {
      "epoch": 6.967840018711262,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 29788
    },
    {
      "epoch": 6.9680739094842705,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 29789
    },
    {
      "epoch": 6.96830780025728,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 29790
    },
    {
      "epoch": 6.968541691030289,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 29791
    },
    {
      "epoch": 6.968775581803298,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 29792
    },
    {
      "epoch": 6.969009472576307,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.557,
      "step": 29793
    },
    {
      "epoch": 6.969243363349316,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.4655,
      "step": 29794
    },
    {
      "epoch": 6.969477254122324,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 29795
    },
    {
      "epoch": 6.969711144895334,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7335,
      "step": 29796
    },
    {
      "epoch": 6.969945035668343,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 29797
    },
    {
      "epoch": 6.970178926441352,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6641,
      "step": 29798
    },
    {
      "epoch": 6.970412817214361,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7777,
      "step": 29799
    },
    {
      "epoch": 6.9706467079873695,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 29800
    },
    {
      "epoch": 6.9706467079873695,
      "eval_runtime": 4.5846,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 29800
    },
    {
      "epoch": 6.970880598760379,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 29801
    },
    {
      "epoch": 6.971114489533388,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 29802
    },
    {
      "epoch": 6.971348380306397,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.673,
      "step": 29803
    },
    {
      "epoch": 6.971582271079406,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4043,
      "step": 29804
    },
    {
      "epoch": 6.971816161852415,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 29805
    },
    {
      "epoch": 6.972050052625423,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 29806
    },
    {
      "epoch": 6.972283943398433,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 29807
    },
    {
      "epoch": 6.972517834171442,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.504,
      "step": 29808
    },
    {
      "epoch": 6.972751724944451,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 29809
    },
    {
      "epoch": 6.97298561571746,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 29810
    },
    {
      "epoch": 6.973219506490469,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 29811
    },
    {
      "epoch": 6.973453397263478,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 29812
    },
    {
      "epoch": 6.973687288036487,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 29813
    },
    {
      "epoch": 6.973921178809496,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 29814
    },
    {
      "epoch": 6.974155069582505,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9605,
      "step": 29815
    },
    {
      "epoch": 6.974388960355514,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 29816
    },
    {
      "epoch": 6.974622851128523,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6572,
      "step": 29817
    },
    {
      "epoch": 6.974856741901532,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 29818
    },
    {
      "epoch": 6.975090632674541,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 29819
    },
    {
      "epoch": 6.97532452344755,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 29820
    },
    {
      "epoch": 6.975558414220559,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.7499,
      "step": 29821
    },
    {
      "epoch": 6.975792304993568,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 29822
    },
    {
      "epoch": 6.976026195766577,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5204,
      "step": 29823
    },
    {
      "epoch": 6.976260086539586,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 29824
    },
    {
      "epoch": 6.976493977312595,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 29825
    },
    {
      "epoch": 6.976727868085604,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 29826
    },
    {
      "epoch": 6.976961758858613,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 29827
    },
    {
      "epoch": 6.977195649631622,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 29828
    },
    {
      "epoch": 6.977429540404631,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 29829
    },
    {
      "epoch": 6.97766343117764,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8107,
      "step": 29830
    },
    {
      "epoch": 6.977897321950649,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4636,
      "step": 29831
    },
    {
      "epoch": 6.9781312127236585,
      "grad_norm": 9.3125,
      "learning_rate": 3e-05,
      "loss": 1.4664,
      "step": 29832
    },
    {
      "epoch": 6.978365103496667,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5262,
      "step": 29833
    },
    {
      "epoch": 6.978598994269676,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 29834
    },
    {
      "epoch": 6.978832885042685,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 29835
    },
    {
      "epoch": 6.979066775815694,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 29836
    },
    {
      "epoch": 6.979300666588703,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5717,
      "step": 29837
    },
    {
      "epoch": 6.9795345573617125,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 1.7208,
      "step": 29838
    },
    {
      "epoch": 6.979768448134721,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 29839
    },
    {
      "epoch": 6.98000233890773,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 29840
    },
    {
      "epoch": 6.980236229680739,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 29841
    },
    {
      "epoch": 6.980470120453748,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7789,
      "step": 29842
    },
    {
      "epoch": 6.9807040112267575,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 29843
    },
    {
      "epoch": 6.980937901999766,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 29844
    },
    {
      "epoch": 6.981171792772775,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 29845
    },
    {
      "epoch": 6.981405683545784,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 29846
    },
    {
      "epoch": 6.981639574318793,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 29847
    },
    {
      "epoch": 6.981873465091802,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6159,
      "step": 29848
    },
    {
      "epoch": 6.9821073558648115,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6908,
      "step": 29849
    },
    {
      "epoch": 6.98234124663782,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 29850
    },
    {
      "epoch": 6.982575137410829,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5563,
      "step": 29851
    },
    {
      "epoch": 6.982809028183838,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6619,
      "step": 29852
    },
    {
      "epoch": 6.983042918956847,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 29853
    },
    {
      "epoch": 6.9832768097298565,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6967,
      "step": 29854
    },
    {
      "epoch": 6.983510700502865,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 29855
    },
    {
      "epoch": 6.983744591275874,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 29856
    },
    {
      "epoch": 6.983978482048883,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7794,
      "step": 29857
    },
    {
      "epoch": 6.984212372821892,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 29858
    },
    {
      "epoch": 6.984446263594901,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 29859
    },
    {
      "epoch": 6.9846801543679105,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 29860
    },
    {
      "epoch": 6.984914045140919,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 29861
    },
    {
      "epoch": 6.985147935913928,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 29862
    },
    {
      "epoch": 6.985381826686937,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 29863
    },
    {
      "epoch": 6.985615717459947,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 29864
    },
    {
      "epoch": 6.9858496082329555,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8156,
      "step": 29865
    },
    {
      "epoch": 6.986083499005964,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 29866
    },
    {
      "epoch": 6.986317389778973,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 29867
    },
    {
      "epoch": 6.986551280551982,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6018,
      "step": 29868
    },
    {
      "epoch": 6.986785171324991,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 29869
    },
    {
      "epoch": 6.987019062098001,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 29870
    },
    {
      "epoch": 6.9872529528710094,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 29871
    },
    {
      "epoch": 6.987486843644018,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 29872
    },
    {
      "epoch": 6.987720734417027,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 29873
    },
    {
      "epoch": 6.987954625190036,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 29874
    },
    {
      "epoch": 6.988188515963046,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 29875
    },
    {
      "epoch": 6.9884224067360545,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 29876
    },
    {
      "epoch": 6.988656297509063,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 29877
    },
    {
      "epoch": 6.988890188282072,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 29878
    },
    {
      "epoch": 6.989124079055081,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5533,
      "step": 29879
    },
    {
      "epoch": 6.98935796982809,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 29880
    },
    {
      "epoch": 6.9895918606011,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6689,
      "step": 29881
    },
    {
      "epoch": 6.989825751374108,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 29882
    },
    {
      "epoch": 6.990059642147117,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 29883
    },
    {
      "epoch": 6.990293532920126,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 29884
    },
    {
      "epoch": 6.990527423693135,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 29885
    },
    {
      "epoch": 6.990761314466145,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 29886
    },
    {
      "epoch": 6.9909952052391535,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 29887
    },
    {
      "epoch": 6.991229096012162,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 29888
    },
    {
      "epoch": 6.991462986785171,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 29889
    },
    {
      "epoch": 6.99169687755818,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 29890
    },
    {
      "epoch": 6.991930768331189,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5874,
      "step": 29891
    },
    {
      "epoch": 6.992164659104199,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 29892
    },
    {
      "epoch": 6.992398549877207,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6679,
      "step": 29893
    },
    {
      "epoch": 6.992632440650216,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4652,
      "step": 29894
    },
    {
      "epoch": 6.992866331423225,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 29895
    },
    {
      "epoch": 6.993100222196235,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 29896
    },
    {
      "epoch": 6.993334112969244,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7872,
      "step": 29897
    },
    {
      "epoch": 6.9935680037422525,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 29898
    },
    {
      "epoch": 6.993801894515261,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 29899
    },
    {
      "epoch": 6.99403578528827,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 29900
    },
    {
      "epoch": 6.99403578528827,
      "eval_runtime": 4.6567,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 29900
    },
    {
      "epoch": 6.994269676061279,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 29901
    },
    {
      "epoch": 6.994503566834289,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 29902
    },
    {
      "epoch": 6.994737457607298,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 29903
    },
    {
      "epoch": 6.994971348380306,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.65,
      "step": 29904
    },
    {
      "epoch": 6.995205239153315,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 29905
    },
    {
      "epoch": 6.995439129926324,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6903,
      "step": 29906
    },
    {
      "epoch": 6.995673020699334,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 29907
    },
    {
      "epoch": 6.995906911472343,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 29908
    },
    {
      "epoch": 6.9961408022453515,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 29909
    },
    {
      "epoch": 6.99637469301836,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7289,
      "step": 29910
    },
    {
      "epoch": 6.996608583791369,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 29911
    },
    {
      "epoch": 6.996842474564378,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 29912
    },
    {
      "epoch": 6.997076365337388,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 29913
    },
    {
      "epoch": 6.997310256110397,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.63,
      "step": 29914
    },
    {
      "epoch": 6.997544146883405,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 29915
    },
    {
      "epoch": 6.997778037656414,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 29916
    },
    {
      "epoch": 6.998011928429423,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7406,
      "step": 29917
    },
    {
      "epoch": 6.998245819202433,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 29918
    },
    {
      "epoch": 6.998479709975442,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 29919
    },
    {
      "epoch": 6.9987136007484505,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 29920
    },
    {
      "epoch": 6.998947491521459,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6351,
      "step": 29921
    },
    {
      "epoch": 6.999181382294468,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 29922
    },
    {
      "epoch": 6.999415273067477,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 29923
    },
    {
      "epoch": 6.999649163840487,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 29924
    },
    {
      "epoch": 6.999883054613496,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5286,
      "step": 29925
    },
    {
      "epoch": 7.000233890773009,
      "grad_norm": 11.125,
      "learning_rate": 3e-05,
      "loss": 4.2465,
      "step": 29926
    },
    {
      "epoch": 7.000467781546018,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 29927
    },
    {
      "epoch": 7.000701672319027,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 29928
    },
    {
      "epoch": 7.000935563092036,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5665,
      "step": 29929
    },
    {
      "epoch": 7.001169453865045,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5731,
      "step": 29930
    },
    {
      "epoch": 7.001403344638054,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.619,
      "step": 29931
    },
    {
      "epoch": 7.001637235411063,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6622,
      "step": 29932
    },
    {
      "epoch": 7.0018711261840725,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 29933
    },
    {
      "epoch": 7.002105016957081,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3241,
      "step": 29934
    },
    {
      "epoch": 7.00233890773009,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 29935
    },
    {
      "epoch": 7.002572798503099,
      "grad_norm": 9.75,
      "learning_rate": 3e-05,
      "loss": 1.6495,
      "step": 29936
    },
    {
      "epoch": 7.002806689276108,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 29937
    },
    {
      "epoch": 7.003040580049117,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6028,
      "step": 29938
    },
    {
      "epoch": 7.003274470822126,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8196,
      "step": 29939
    },
    {
      "epoch": 7.003508361595135,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5696,
      "step": 29940
    },
    {
      "epoch": 7.003742252368144,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5963,
      "step": 29941
    },
    {
      "epoch": 7.003976143141153,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 29942
    },
    {
      "epoch": 7.004210033914162,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5493,
      "step": 29943
    },
    {
      "epoch": 7.0044439246871715,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 29944
    },
    {
      "epoch": 7.00467781546018,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6201,
      "step": 29945
    },
    {
      "epoch": 7.004911706233189,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6788,
      "step": 29946
    },
    {
      "epoch": 7.005145597006198,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4652,
      "step": 29947
    },
    {
      "epoch": 7.005379487779207,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6759,
      "step": 29948
    },
    {
      "epoch": 7.0056133785522166,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 29949
    },
    {
      "epoch": 7.005847269325225,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 29950
    },
    {
      "epoch": 7.006081160098234,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9017,
      "step": 29951
    },
    {
      "epoch": 7.006315050871243,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.5932,
      "step": 29952
    },
    {
      "epoch": 7.006548941644252,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 29953
    },
    {
      "epoch": 7.006782832417261,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7175,
      "step": 29954
    },
    {
      "epoch": 7.0070167231902705,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 29955
    },
    {
      "epoch": 7.007250613963279,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 29956
    },
    {
      "epoch": 7.007484504736288,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 29957
    },
    {
      "epoch": 7.007718395509297,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4344,
      "step": 29958
    },
    {
      "epoch": 7.007952286282306,
      "grad_norm": 2.46875,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 29959
    },
    {
      "epoch": 7.0081861770553155,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 29960
    },
    {
      "epoch": 7.008420067828324,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 29961
    },
    {
      "epoch": 7.008653958601333,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7244,
      "step": 29962
    },
    {
      "epoch": 7.008887849374342,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6474,
      "step": 29963
    },
    {
      "epoch": 7.009121740147351,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 29964
    },
    {
      "epoch": 7.009355630920361,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 29965
    },
    {
      "epoch": 7.0095895216933695,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.3979,
      "step": 29966
    },
    {
      "epoch": 7.009823412466378,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 29967
    },
    {
      "epoch": 7.010057303239387,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 29968
    },
    {
      "epoch": 7.010291194012396,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6593,
      "step": 29969
    },
    {
      "epoch": 7.010525084785405,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.4059,
      "step": 29970
    },
    {
      "epoch": 7.0107589755584145,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 29971
    },
    {
      "epoch": 7.010992866331423,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.3672,
      "step": 29972
    },
    {
      "epoch": 7.011226757104432,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 29973
    },
    {
      "epoch": 7.011460647877441,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 29974
    },
    {
      "epoch": 7.01169453865045,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 29975
    },
    {
      "epoch": 7.01192842942346,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 29976
    },
    {
      "epoch": 7.0121623201964685,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 29977
    },
    {
      "epoch": 7.012396210969477,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6207,
      "step": 29978
    },
    {
      "epoch": 7.012630101742486,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6316,
      "step": 29979
    },
    {
      "epoch": 7.012863992515495,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 29980
    },
    {
      "epoch": 7.013097883288505,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 29981
    },
    {
      "epoch": 7.0133317740615135,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6233,
      "step": 29982
    },
    {
      "epoch": 7.013565664834522,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6124,
      "step": 29983
    },
    {
      "epoch": 7.013799555607531,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 29984
    },
    {
      "epoch": 7.01403344638054,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 29985
    },
    {
      "epoch": 7.014267337153549,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6739,
      "step": 29986
    },
    {
      "epoch": 7.014501227926559,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5588,
      "step": 29987
    },
    {
      "epoch": 7.0147351186995675,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6622,
      "step": 29988
    },
    {
      "epoch": 7.014969009472576,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 29989
    },
    {
      "epoch": 7.015202900245585,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 29990
    },
    {
      "epoch": 7.015436791018594,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5875,
      "step": 29991
    },
    {
      "epoch": 7.015670681791604,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 29992
    },
    {
      "epoch": 7.0159045725646125,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 29993
    },
    {
      "epoch": 7.016138463337621,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 29994
    },
    {
      "epoch": 7.01637235411063,
      "grad_norm": 12.0625,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 29995
    },
    {
      "epoch": 7.016606244883639,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5626,
      "step": 29996
    },
    {
      "epoch": 7.016840135656649,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 29997
    },
    {
      "epoch": 7.017074026429658,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7038,
      "step": 29998
    },
    {
      "epoch": 7.0173079172026664,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 29999
    },
    {
      "epoch": 7.017541807975675,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8684,
      "step": 30000
    },
    {
      "epoch": 7.017541807975675,
      "eval_runtime": 4.6407,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 30000
    },
    {
      "epoch": 7.017775698748684,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 30001
    },
    {
      "epoch": 7.018009589521693,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 30002
    },
    {
      "epoch": 7.018243480294703,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 30003
    },
    {
      "epoch": 7.0184773710677115,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7456,
      "step": 30004
    },
    {
      "epoch": 7.01871126184072,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 30005
    },
    {
      "epoch": 7.018945152613729,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6937,
      "step": 30006
    },
    {
      "epoch": 7.019179043386738,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 30007
    },
    {
      "epoch": 7.019412934159748,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 30008
    },
    {
      "epoch": 7.019646824932757,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 30009
    },
    {
      "epoch": 7.019880715705765,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4918,
      "step": 30010
    },
    {
      "epoch": 7.020114606478774,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6377,
      "step": 30011
    },
    {
      "epoch": 7.020348497251783,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 30012
    },
    {
      "epoch": 7.020582388024793,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6114,
      "step": 30013
    },
    {
      "epoch": 7.020816278797802,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 30014
    },
    {
      "epoch": 7.0210501695708105,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 30015
    },
    {
      "epoch": 7.021284060343819,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5722,
      "step": 30016
    },
    {
      "epoch": 7.021517951116828,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 30017
    },
    {
      "epoch": 7.021751841889837,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 30018
    },
    {
      "epoch": 7.021985732662847,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 30019
    },
    {
      "epoch": 7.022219623435856,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7729,
      "step": 30020
    },
    {
      "epoch": 7.022453514208864,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5054,
      "step": 30021
    },
    {
      "epoch": 7.022687404981873,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 30022
    },
    {
      "epoch": 7.022921295754882,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 30023
    },
    {
      "epoch": 7.023155186527892,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 30024
    },
    {
      "epoch": 7.023389077300901,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 30025
    },
    {
      "epoch": 7.0236229680739095,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 30026
    },
    {
      "epoch": 7.023856858846918,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5905,
      "step": 30027
    },
    {
      "epoch": 7.024090749619927,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 30028
    },
    {
      "epoch": 7.024324640392937,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5374,
      "step": 30029
    },
    {
      "epoch": 7.024558531165946,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6565,
      "step": 30030
    },
    {
      "epoch": 7.024792421938955,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 30031
    },
    {
      "epoch": 7.025026312711963,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 30032
    },
    {
      "epoch": 7.025260203484972,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4802,
      "step": 30033
    },
    {
      "epoch": 7.025494094257981,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 30034
    },
    {
      "epoch": 7.025727985030991,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 30035
    },
    {
      "epoch": 7.025961875804,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 30036
    },
    {
      "epoch": 7.0261957665770085,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.7708,
      "step": 30037
    },
    {
      "epoch": 7.026429657350017,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 30038
    },
    {
      "epoch": 7.026663548123026,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 30039
    },
    {
      "epoch": 7.026897438896036,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7059,
      "step": 30040
    },
    {
      "epoch": 7.027131329669045,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5073,
      "step": 30041
    },
    {
      "epoch": 7.027365220442054,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4339,
      "step": 30042
    },
    {
      "epoch": 7.027599111215062,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 30043
    },
    {
      "epoch": 7.027833001988071,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7791,
      "step": 30044
    },
    {
      "epoch": 7.028066892761081,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 30045
    },
    {
      "epoch": 7.02830078353409,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 30046
    },
    {
      "epoch": 7.028534674307099,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6421,
      "step": 30047
    },
    {
      "epoch": 7.0287685650801075,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6471,
      "step": 30048
    },
    {
      "epoch": 7.029002455853116,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6001,
      "step": 30049
    },
    {
      "epoch": 7.029236346626125,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5335,
      "step": 30050
    },
    {
      "epoch": 7.029470237399135,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 30051
    },
    {
      "epoch": 7.029704128172144,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 30052
    },
    {
      "epoch": 7.029938018945153,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 30053
    },
    {
      "epoch": 7.030171909718161,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 30054
    },
    {
      "epoch": 7.03040580049117,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 30055
    },
    {
      "epoch": 7.03063969126418,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6429,
      "step": 30056
    },
    {
      "epoch": 7.030873582037189,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7078,
      "step": 30057
    },
    {
      "epoch": 7.031107472810198,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5332,
      "step": 30058
    },
    {
      "epoch": 7.0313413635832065,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 30059
    },
    {
      "epoch": 7.031575254356215,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 30060
    },
    {
      "epoch": 7.031809145129225,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 30061
    },
    {
      "epoch": 7.032043035902234,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.532,
      "step": 30062
    },
    {
      "epoch": 7.032276926675243,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 30063
    },
    {
      "epoch": 7.032510817448252,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 30064
    },
    {
      "epoch": 7.03274470822126,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 30065
    },
    {
      "epoch": 7.032978598994269,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.4808,
      "step": 30066
    },
    {
      "epoch": 7.033212489767279,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 30067
    },
    {
      "epoch": 7.033446380540288,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 30068
    },
    {
      "epoch": 7.033680271313297,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 30069
    },
    {
      "epoch": 7.0339141620863055,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6679,
      "step": 30070
    },
    {
      "epoch": 7.034148052859314,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 30071
    },
    {
      "epoch": 7.034381943632324,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 30072
    },
    {
      "epoch": 7.034615834405333,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 30073
    },
    {
      "epoch": 7.034849725178342,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 30074
    },
    {
      "epoch": 7.035083615951351,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8466,
      "step": 30075
    },
    {
      "epoch": 7.035317506724359,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5265,
      "step": 30076
    },
    {
      "epoch": 7.035551397497369,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 30077
    },
    {
      "epoch": 7.035785288270378,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 30078
    },
    {
      "epoch": 7.036019179043387,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 30079
    },
    {
      "epoch": 7.036253069816396,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6829,
      "step": 30080
    },
    {
      "epoch": 7.0364869605894045,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 30081
    },
    {
      "epoch": 7.036720851362413,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 30082
    },
    {
      "epoch": 7.036954742135423,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 30083
    },
    {
      "epoch": 7.037188632908432,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4785,
      "step": 30084
    },
    {
      "epoch": 7.037422523681441,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 30085
    },
    {
      "epoch": 7.03765641445445,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 30086
    },
    {
      "epoch": 7.037890305227458,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7285,
      "step": 30087
    },
    {
      "epoch": 7.038124196000468,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6344,
      "step": 30088
    },
    {
      "epoch": 7.038358086773477,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6914,
      "step": 30089
    },
    {
      "epoch": 7.038591977546486,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.4089,
      "step": 30090
    },
    {
      "epoch": 7.038825868319495,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 30091
    },
    {
      "epoch": 7.0390597590925035,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6734,
      "step": 30092
    },
    {
      "epoch": 7.039293649865513,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 30093
    },
    {
      "epoch": 7.039527540638522,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.689,
      "step": 30094
    },
    {
      "epoch": 7.039761431411531,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 30095
    },
    {
      "epoch": 7.03999532218454,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.662,
      "step": 30096
    },
    {
      "epoch": 7.0402292129575486,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 30097
    },
    {
      "epoch": 7.040463103730557,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.687,
      "step": 30098
    },
    {
      "epoch": 7.040696994503567,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5356,
      "step": 30099
    },
    {
      "epoch": 7.040930885276576,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 30100
    },
    {
      "epoch": 7.040930885276576,
      "eval_runtime": 4.5683,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 30100
    },
    {
      "epoch": 7.041164776049585,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 30101
    },
    {
      "epoch": 7.041398666822594,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 30102
    },
    {
      "epoch": 7.0416325575956025,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6186,
      "step": 30103
    },
    {
      "epoch": 7.041866448368612,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.4391,
      "step": 30104
    },
    {
      "epoch": 7.042100339141621,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 30105
    },
    {
      "epoch": 7.04233422991463,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7265,
      "step": 30106
    },
    {
      "epoch": 7.042568120687639,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 30107
    },
    {
      "epoch": 7.0428020114606475,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 30108
    },
    {
      "epoch": 7.043035902233657,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7463,
      "step": 30109
    },
    {
      "epoch": 7.043269793006666,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9976,
      "step": 30110
    },
    {
      "epoch": 7.043503683779675,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5648,
      "step": 30111
    },
    {
      "epoch": 7.043737574552684,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5896,
      "step": 30112
    },
    {
      "epoch": 7.043971465325693,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 30113
    },
    {
      "epoch": 7.044205356098702,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 30114
    },
    {
      "epoch": 7.044439246871711,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5985,
      "step": 30115
    },
    {
      "epoch": 7.04467313764472,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 30116
    },
    {
      "epoch": 7.044907028417729,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 30117
    },
    {
      "epoch": 7.045140919190738,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6045,
      "step": 30118
    },
    {
      "epoch": 7.0453748099637465,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 30119
    },
    {
      "epoch": 7.045608700736756,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 30120
    },
    {
      "epoch": 7.045842591509765,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6088,
      "step": 30121
    },
    {
      "epoch": 7.046076482282774,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5253,
      "step": 30122
    },
    {
      "epoch": 7.046310373055783,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.528,
      "step": 30123
    },
    {
      "epoch": 7.046544263828792,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.2824,
      "step": 30124
    },
    {
      "epoch": 7.046778154601801,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5718,
      "step": 30125
    },
    {
      "epoch": 7.04701204537481,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 30126
    },
    {
      "epoch": 7.047245936147819,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 30127
    },
    {
      "epoch": 7.047479826920828,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4765,
      "step": 30128
    },
    {
      "epoch": 7.047713717693837,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 30129
    },
    {
      "epoch": 7.0479476084668455,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 30130
    },
    {
      "epoch": 7.048181499239855,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4812,
      "step": 30131
    },
    {
      "epoch": 7.048415390012864,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 30132
    },
    {
      "epoch": 7.048649280785873,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6483,
      "step": 30133
    },
    {
      "epoch": 7.048883171558882,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 30134
    },
    {
      "epoch": 7.049117062331891,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6343,
      "step": 30135
    },
    {
      "epoch": 7.0493509531049,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8417,
      "step": 30136
    },
    {
      "epoch": 7.049584843877909,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 30137
    },
    {
      "epoch": 7.049818734650918,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 30138
    },
    {
      "epoch": 7.050052625423927,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8456,
      "step": 30139
    },
    {
      "epoch": 7.050286516196936,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.5425,
      "step": 30140
    },
    {
      "epoch": 7.050520406969945,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 30141
    },
    {
      "epoch": 7.050754297742954,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6323,
      "step": 30142
    },
    {
      "epoch": 7.050988188515963,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 30143
    },
    {
      "epoch": 7.051222079288972,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6787,
      "step": 30144
    },
    {
      "epoch": 7.051455970061981,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 30145
    },
    {
      "epoch": 7.0516898608349905,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 30146
    },
    {
      "epoch": 7.051923751607999,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 30147
    },
    {
      "epoch": 7.052157642381008,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5794,
      "step": 30148
    },
    {
      "epoch": 7.052391533154017,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5821,
      "step": 30149
    },
    {
      "epoch": 7.052625423927026,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 30150
    },
    {
      "epoch": 7.052859314700035,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 30151
    },
    {
      "epoch": 7.053093205473044,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5696,
      "step": 30152
    },
    {
      "epoch": 7.053327096246053,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7226,
      "step": 30153
    },
    {
      "epoch": 7.053560987019062,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 30154
    },
    {
      "epoch": 7.053794877792071,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5229,
      "step": 30155
    },
    {
      "epoch": 7.05402876856508,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 30156
    },
    {
      "epoch": 7.0542626593380895,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 30157
    },
    {
      "epoch": 7.054496550111098,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.712,
      "step": 30158
    },
    {
      "epoch": 7.054730440884107,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 30159
    },
    {
      "epoch": 7.054964331657116,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7114,
      "step": 30160
    },
    {
      "epoch": 7.055198222430125,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 30161
    },
    {
      "epoch": 7.055432113203135,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 30162
    },
    {
      "epoch": 7.055666003976143,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 30163
    },
    {
      "epoch": 7.055899894749152,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 30164
    },
    {
      "epoch": 7.056133785522161,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 30165
    },
    {
      "epoch": 7.05636767629517,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 30166
    },
    {
      "epoch": 7.056601567068179,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.7231,
      "step": 30167
    },
    {
      "epoch": 7.0568354578411885,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5083,
      "step": 30168
    },
    {
      "epoch": 7.057069348614197,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6268,
      "step": 30169
    },
    {
      "epoch": 7.057303239387206,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4523,
      "step": 30170
    },
    {
      "epoch": 7.057537130160215,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 30171
    },
    {
      "epoch": 7.057771020933224,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 30172
    },
    {
      "epoch": 7.058004911706234,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 30173
    },
    {
      "epoch": 7.058238802479242,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 30174
    },
    {
      "epoch": 7.058472693252251,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6387,
      "step": 30175
    },
    {
      "epoch": 7.05870658402526,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5424,
      "step": 30176
    },
    {
      "epoch": 7.058940474798269,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7595,
      "step": 30177
    },
    {
      "epoch": 7.059174365571279,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6874,
      "step": 30178
    },
    {
      "epoch": 7.0594082563442875,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7406,
      "step": 30179
    },
    {
      "epoch": 7.059642147117296,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5006,
      "step": 30180
    },
    {
      "epoch": 7.059876037890305,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 30181
    },
    {
      "epoch": 7.060109928663314,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.8415,
      "step": 30182
    },
    {
      "epoch": 7.060343819436323,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7152,
      "step": 30183
    },
    {
      "epoch": 7.060577710209333,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 30184
    },
    {
      "epoch": 7.060811600982341,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 30185
    },
    {
      "epoch": 7.06104549175535,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 30186
    },
    {
      "epoch": 7.061279382528359,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 30187
    },
    {
      "epoch": 7.061513273301368,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 30188
    },
    {
      "epoch": 7.061747164074378,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1765,
      "step": 30189
    },
    {
      "epoch": 7.0619810548473865,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 30190
    },
    {
      "epoch": 7.062214945620395,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5588,
      "step": 30191
    },
    {
      "epoch": 7.062448836393404,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 30192
    },
    {
      "epoch": 7.062682727166413,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 30193
    },
    {
      "epoch": 7.062916617939423,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7298,
      "step": 30194
    },
    {
      "epoch": 7.0631505087124316,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6027,
      "step": 30195
    },
    {
      "epoch": 7.06338439948544,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6057,
      "step": 30196
    },
    {
      "epoch": 7.063618290258449,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5598,
      "step": 30197
    },
    {
      "epoch": 7.063852181031458,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5981,
      "step": 30198
    },
    {
      "epoch": 7.064086071804467,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 30199
    },
    {
      "epoch": 7.064319962577477,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 30200
    },
    {
      "epoch": 7.064319962577477,
      "eval_runtime": 4.6519,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 30200
    },
    {
      "epoch": 7.0645538533504855,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5979,
      "step": 30201
    },
    {
      "epoch": 7.064787744123494,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 30202
    },
    {
      "epoch": 7.065021634896503,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5207,
      "step": 30203
    },
    {
      "epoch": 7.065255525669512,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 30204
    },
    {
      "epoch": 7.065489416442522,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 30205
    },
    {
      "epoch": 7.0657233072155305,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 30206
    },
    {
      "epoch": 7.065957197988539,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 30207
    },
    {
      "epoch": 7.066191088761548,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6253,
      "step": 30208
    },
    {
      "epoch": 7.066424979534557,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6785,
      "step": 30209
    },
    {
      "epoch": 7.066658870307567,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.6921,
      "step": 30210
    },
    {
      "epoch": 7.066892761080576,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 30211
    },
    {
      "epoch": 7.0671266518535845,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 30212
    },
    {
      "epoch": 7.067360542626593,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6519,
      "step": 30213
    },
    {
      "epoch": 7.067594433399602,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 30214
    },
    {
      "epoch": 7.067828324172611,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 30215
    },
    {
      "epoch": 7.068062214945621,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6336,
      "step": 30216
    },
    {
      "epoch": 7.0682961057186295,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 30217
    },
    {
      "epoch": 7.068529996491638,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 30218
    },
    {
      "epoch": 7.068763887264647,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 30219
    },
    {
      "epoch": 7.068997778037656,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 30220
    },
    {
      "epoch": 7.069231668810666,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 30221
    },
    {
      "epoch": 7.069465559583675,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 30222
    },
    {
      "epoch": 7.0696994503566835,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6289,
      "step": 30223
    },
    {
      "epoch": 7.069933341129692,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 30224
    },
    {
      "epoch": 7.070167231902701,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6885,
      "step": 30225
    },
    {
      "epoch": 7.070401122675711,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6413,
      "step": 30226
    },
    {
      "epoch": 7.07063501344872,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 30227
    },
    {
      "epoch": 7.0708689042217285,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.643,
      "step": 30228
    },
    {
      "epoch": 7.071102794994737,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4519,
      "step": 30229
    },
    {
      "epoch": 7.071336685767746,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 30230
    },
    {
      "epoch": 7.071570576540755,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 30231
    },
    {
      "epoch": 7.071804467313765,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 30232
    },
    {
      "epoch": 7.072038358086774,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 30233
    },
    {
      "epoch": 7.0722722488597825,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7131,
      "step": 30234
    },
    {
      "epoch": 7.072506139632791,
      "grad_norm": 8.4375,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 30235
    },
    {
      "epoch": 7.0727400304058,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5897,
      "step": 30236
    },
    {
      "epoch": 7.07297392117881,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 30237
    },
    {
      "epoch": 7.073207811951819,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 30238
    },
    {
      "epoch": 7.0734417027248275,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7112,
      "step": 30239
    },
    {
      "epoch": 7.073675593497836,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 30240
    },
    {
      "epoch": 7.073909484270845,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.615,
      "step": 30241
    },
    {
      "epoch": 7.074143375043855,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 30242
    },
    {
      "epoch": 7.074377265816864,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 30243
    },
    {
      "epoch": 7.074611156589873,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7121,
      "step": 30244
    },
    {
      "epoch": 7.0748450473628814,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 30245
    },
    {
      "epoch": 7.07507893813589,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5401,
      "step": 30246
    },
    {
      "epoch": 7.075312828908899,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6,
      "step": 30247
    },
    {
      "epoch": 7.075546719681909,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 30248
    },
    {
      "epoch": 7.075780610454918,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 30249
    },
    {
      "epoch": 7.0760145012279265,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4227,
      "step": 30250
    },
    {
      "epoch": 7.076248392000935,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6764,
      "step": 30251
    },
    {
      "epoch": 7.076482282773944,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 30252
    },
    {
      "epoch": 7.076716173546954,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7333,
      "step": 30253
    },
    {
      "epoch": 7.076950064319963,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6179,
      "step": 30254
    },
    {
      "epoch": 7.077183955092972,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5271,
      "step": 30255
    },
    {
      "epoch": 7.07741784586598,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 30256
    },
    {
      "epoch": 7.077651736638989,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 30257
    },
    {
      "epoch": 7.077885627411999,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 30258
    },
    {
      "epoch": 7.078119518185008,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6601,
      "step": 30259
    },
    {
      "epoch": 7.078353408958017,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 30260
    },
    {
      "epoch": 7.0785872997310255,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 30261
    },
    {
      "epoch": 7.078821190504034,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 30262
    },
    {
      "epoch": 7.079055081277043,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7093,
      "step": 30263
    },
    {
      "epoch": 7.079288972050053,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7625,
      "step": 30264
    },
    {
      "epoch": 7.079522862823062,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 30265
    },
    {
      "epoch": 7.079756753596071,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.4994,
      "step": 30266
    },
    {
      "epoch": 7.079990644369079,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 30267
    },
    {
      "epoch": 7.080224535142088,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 30268
    },
    {
      "epoch": 7.080458425915098,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 30269
    },
    {
      "epoch": 7.080692316688107,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 30270
    },
    {
      "epoch": 7.080926207461116,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6251,
      "step": 30271
    },
    {
      "epoch": 7.0811600982341245,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4112,
      "step": 30272
    },
    {
      "epoch": 7.081393989007133,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 30273
    },
    {
      "epoch": 7.081627879780143,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.589,
      "step": 30274
    },
    {
      "epoch": 7.081861770553152,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 30275
    },
    {
      "epoch": 7.082095661326161,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 30276
    },
    {
      "epoch": 7.08232955209917,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7854,
      "step": 30277
    },
    {
      "epoch": 7.082563442872178,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 30278
    },
    {
      "epoch": 7.082797333645187,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8234,
      "step": 30279
    },
    {
      "epoch": 7.083031224418197,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 30280
    },
    {
      "epoch": 7.083265115191206,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.5341,
      "step": 30281
    },
    {
      "epoch": 7.083499005964215,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 30282
    },
    {
      "epoch": 7.0837328967372235,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 30283
    },
    {
      "epoch": 7.083966787510232,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5699,
      "step": 30284
    },
    {
      "epoch": 7.084200678283242,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5377,
      "step": 30285
    },
    {
      "epoch": 7.084434569056251,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 30286
    },
    {
      "epoch": 7.08466845982926,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 30287
    },
    {
      "epoch": 7.084902350602269,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.4864,
      "step": 30288
    },
    {
      "epoch": 7.085136241375277,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 30289
    },
    {
      "epoch": 7.085370132148287,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8061,
      "step": 30290
    },
    {
      "epoch": 7.085604022921296,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 30291
    },
    {
      "epoch": 7.085837913694305,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7442,
      "step": 30292
    },
    {
      "epoch": 7.086071804467314,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9204,
      "step": 30293
    },
    {
      "epoch": 7.0863056952403225,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 30294
    },
    {
      "epoch": 7.086539586013331,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 30295
    },
    {
      "epoch": 7.086773476786341,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 30296
    },
    {
      "epoch": 7.08700736755935,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5921,
      "step": 30297
    },
    {
      "epoch": 7.087241258332359,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 30298
    },
    {
      "epoch": 7.087475149105368,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.599,
      "step": 30299
    },
    {
      "epoch": 7.087709039878376,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6765,
      "step": 30300
    },
    {
      "epoch": 7.087709039878376,
      "eval_runtime": 4.653,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 30300
    },
    {
      "epoch": 7.087942930651386,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 30301
    },
    {
      "epoch": 7.088176821424395,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 30302
    },
    {
      "epoch": 7.088410712197404,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 30303
    },
    {
      "epoch": 7.088644602970413,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 30304
    },
    {
      "epoch": 7.0888784937434215,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 30305
    },
    {
      "epoch": 7.089112384516431,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 30306
    },
    {
      "epoch": 7.08934627528944,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 30307
    },
    {
      "epoch": 7.089580166062449,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 30308
    },
    {
      "epoch": 7.089814056835458,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 30309
    },
    {
      "epoch": 7.090047947608467,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 30310
    },
    {
      "epoch": 7.090281838381475,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6347,
      "step": 30311
    },
    {
      "epoch": 7.090515729154485,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 30312
    },
    {
      "epoch": 7.090749619927494,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.613,
      "step": 30313
    },
    {
      "epoch": 7.090983510700503,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 30314
    },
    {
      "epoch": 7.091217401473512,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7088,
      "step": 30315
    },
    {
      "epoch": 7.0914512922465205,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4357,
      "step": 30316
    },
    {
      "epoch": 7.09168518301953,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 30317
    },
    {
      "epoch": 7.091919073792539,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8349,
      "step": 30318
    },
    {
      "epoch": 7.092152964565548,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6359,
      "step": 30319
    },
    {
      "epoch": 7.092386855338557,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 30320
    },
    {
      "epoch": 7.092620746111566,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 30321
    },
    {
      "epoch": 7.092854636884575,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 30322
    },
    {
      "epoch": 7.093088527657584,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6433,
      "step": 30323
    },
    {
      "epoch": 7.093322418430593,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6284,
      "step": 30324
    },
    {
      "epoch": 7.093556309203602,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 30325
    },
    {
      "epoch": 7.093790199976611,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6459,
      "step": 30326
    },
    {
      "epoch": 7.09402409074962,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 30327
    },
    {
      "epoch": 7.094257981522629,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6573,
      "step": 30328
    },
    {
      "epoch": 7.094491872295638,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 30329
    },
    {
      "epoch": 7.094725763068647,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7061,
      "step": 30330
    },
    {
      "epoch": 7.094959653841656,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 30331
    },
    {
      "epoch": 7.095193544614665,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 30332
    },
    {
      "epoch": 7.095427435387674,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 30333
    },
    {
      "epoch": 7.095661326160683,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 30334
    },
    {
      "epoch": 7.095895216933692,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 30335
    },
    {
      "epoch": 7.096129107706701,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.4735,
      "step": 30336
    },
    {
      "epoch": 7.09636299847971,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 30337
    },
    {
      "epoch": 7.096596889252719,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 30338
    },
    {
      "epoch": 7.096830780025728,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 30339
    },
    {
      "epoch": 7.097064670798737,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 30340
    },
    {
      "epoch": 7.097298561571746,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 30341
    },
    {
      "epoch": 7.097532452344755,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6725,
      "step": 30342
    },
    {
      "epoch": 7.0977663431177636,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 30343
    },
    {
      "epoch": 7.098000233890773,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9478,
      "step": 30344
    },
    {
      "epoch": 7.098234124663782,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6627,
      "step": 30345
    },
    {
      "epoch": 7.098468015436791,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.5826,
      "step": 30346
    },
    {
      "epoch": 7.0987019062098,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 30347
    },
    {
      "epoch": 7.098935796982809,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6563,
      "step": 30348
    },
    {
      "epoch": 7.099169687755818,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 30349
    },
    {
      "epoch": 7.099403578528827,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.726,
      "step": 30350
    },
    {
      "epoch": 7.099637469301836,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5504,
      "step": 30351
    },
    {
      "epoch": 7.099871360074845,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 30352
    },
    {
      "epoch": 7.100105250847854,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 30353
    },
    {
      "epoch": 7.100339141620863,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 30354
    },
    {
      "epoch": 7.100573032393872,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 30355
    },
    {
      "epoch": 7.100806923166881,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6884,
      "step": 30356
    },
    {
      "epoch": 7.10104081393989,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 30357
    },
    {
      "epoch": 7.101274704712899,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5048,
      "step": 30358
    },
    {
      "epoch": 7.1015085954859085,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 30359
    },
    {
      "epoch": 7.101742486258917,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7173,
      "step": 30360
    },
    {
      "epoch": 7.101976377031926,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 30361
    },
    {
      "epoch": 7.102210267804935,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 30362
    },
    {
      "epoch": 7.102444158577944,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 30363
    },
    {
      "epoch": 7.102678049350953,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3155,
      "step": 30364
    },
    {
      "epoch": 7.102911940123962,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7597,
      "step": 30365
    },
    {
      "epoch": 7.103145830896971,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5757,
      "step": 30366
    },
    {
      "epoch": 7.10337972166998,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 30367
    },
    {
      "epoch": 7.103613612442989,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 30368
    },
    {
      "epoch": 7.103847503215998,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.4589,
      "step": 30369
    },
    {
      "epoch": 7.1040813939890075,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 30370
    },
    {
      "epoch": 7.104315284762016,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6045,
      "step": 30371
    },
    {
      "epoch": 7.104549175535025,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6952,
      "step": 30372
    },
    {
      "epoch": 7.104783066308034,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4973,
      "step": 30373
    },
    {
      "epoch": 7.105016957081043,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8038,
      "step": 30374
    },
    {
      "epoch": 7.105250847854053,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5721,
      "step": 30375
    },
    {
      "epoch": 7.105484738627061,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8568,
      "step": 30376
    },
    {
      "epoch": 7.10571862940007,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 30377
    },
    {
      "epoch": 7.105952520173079,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4579,
      "step": 30378
    },
    {
      "epoch": 7.106186410946088,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 30379
    },
    {
      "epoch": 7.106420301719097,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 30380
    },
    {
      "epoch": 7.1066541924921065,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 30381
    },
    {
      "epoch": 7.106888083265115,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 30382
    },
    {
      "epoch": 7.107121974038124,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5805,
      "step": 30383
    },
    {
      "epoch": 7.107355864811133,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7377,
      "step": 30384
    },
    {
      "epoch": 7.107589755584142,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6231,
      "step": 30385
    },
    {
      "epoch": 7.107823646357152,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 30386
    },
    {
      "epoch": 7.10805753713016,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 30387
    },
    {
      "epoch": 7.108291427903169,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 30388
    },
    {
      "epoch": 7.108525318676178,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 30389
    },
    {
      "epoch": 7.108759209449187,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.684,
      "step": 30390
    },
    {
      "epoch": 7.108993100222197,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5808,
      "step": 30391
    },
    {
      "epoch": 7.1092269909952055,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5602,
      "step": 30392
    },
    {
      "epoch": 7.109460881768214,
      "grad_norm": 2.453125,
      "learning_rate": 3e-05,
      "loss": 1.3375,
      "step": 30393
    },
    {
      "epoch": 7.109694772541223,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 30394
    },
    {
      "epoch": 7.109928663314232,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 30395
    },
    {
      "epoch": 7.110162554087241,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5327,
      "step": 30396
    },
    {
      "epoch": 7.110396444860251,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7163,
      "step": 30397
    },
    {
      "epoch": 7.110630335633259,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 30398
    },
    {
      "epoch": 7.110864226406268,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7415,
      "step": 30399
    },
    {
      "epoch": 7.111098117179277,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 30400
    },
    {
      "epoch": 7.111098117179277,
      "eval_runtime": 4.6459,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 30400
    },
    {
      "epoch": 7.111332007952286,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6889,
      "step": 30401
    },
    {
      "epoch": 7.111565898725296,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 30402
    },
    {
      "epoch": 7.1117997894983045,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 30403
    },
    {
      "epoch": 7.112033680271313,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5974,
      "step": 30404
    },
    {
      "epoch": 7.112267571044322,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 30405
    },
    {
      "epoch": 7.112501461817331,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6162,
      "step": 30406
    },
    {
      "epoch": 7.112735352590341,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7507,
      "step": 30407
    },
    {
      "epoch": 7.11296924336335,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6108,
      "step": 30408
    },
    {
      "epoch": 7.113203134136358,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4344,
      "step": 30409
    },
    {
      "epoch": 7.113437024909367,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7172,
      "step": 30410
    },
    {
      "epoch": 7.113670915682376,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 30411
    },
    {
      "epoch": 7.113904806455385,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6023,
      "step": 30412
    },
    {
      "epoch": 7.114138697228395,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 30413
    },
    {
      "epoch": 7.1143725880014035,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 30414
    },
    {
      "epoch": 7.114606478774412,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5962,
      "step": 30415
    },
    {
      "epoch": 7.114840369547421,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 30416
    },
    {
      "epoch": 7.11507426032043,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 30417
    },
    {
      "epoch": 7.11530815109344,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 30418
    },
    {
      "epoch": 7.115542041866449,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6565,
      "step": 30419
    },
    {
      "epoch": 7.115775932639457,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6904,
      "step": 30420
    },
    {
      "epoch": 7.116009823412466,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5597,
      "step": 30421
    },
    {
      "epoch": 7.116243714185475,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 30422
    },
    {
      "epoch": 7.116477604958485,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 30423
    },
    {
      "epoch": 7.116711495731494,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.716,
      "step": 30424
    },
    {
      "epoch": 7.1169453865045025,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 30425
    },
    {
      "epoch": 7.117179277277511,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 30426
    },
    {
      "epoch": 7.11741316805052,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 30427
    },
    {
      "epoch": 7.117647058823529,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7245,
      "step": 30428
    },
    {
      "epoch": 7.117880949596539,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 30429
    },
    {
      "epoch": 7.118114840369548,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 30430
    },
    {
      "epoch": 7.118348731142556,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6735,
      "step": 30431
    },
    {
      "epoch": 7.118582621915565,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 30432
    },
    {
      "epoch": 7.118816512688574,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 30433
    },
    {
      "epoch": 7.119050403461584,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7914,
      "step": 30434
    },
    {
      "epoch": 7.119284294234593,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 30435
    },
    {
      "epoch": 7.1195181850076015,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 30436
    },
    {
      "epoch": 7.11975207578061,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 30437
    },
    {
      "epoch": 7.119985966553619,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 30438
    },
    {
      "epoch": 7.120219857326629,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 30439
    },
    {
      "epoch": 7.120453748099638,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 30440
    },
    {
      "epoch": 7.1206876388726466,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 30441
    },
    {
      "epoch": 7.120921529645655,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7239,
      "step": 30442
    },
    {
      "epoch": 7.121155420418664,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 30443
    },
    {
      "epoch": 7.121389311191673,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7592,
      "step": 30444
    },
    {
      "epoch": 7.121623201964683,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 30445
    },
    {
      "epoch": 7.121857092737692,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 30446
    },
    {
      "epoch": 7.1220909835107005,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 30447
    },
    {
      "epoch": 7.122324874283709,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 30448
    },
    {
      "epoch": 7.122558765056718,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 30449
    },
    {
      "epoch": 7.122792655829728,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 30450
    },
    {
      "epoch": 7.123026546602737,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 30451
    },
    {
      "epoch": 7.1232604373757455,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7622,
      "step": 30452
    },
    {
      "epoch": 7.123494328148754,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 30453
    },
    {
      "epoch": 7.123728218921763,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.589,
      "step": 30454
    },
    {
      "epoch": 7.123962109694773,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4712,
      "step": 30455
    },
    {
      "epoch": 7.124196000467782,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 30456
    },
    {
      "epoch": 7.124429891240791,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4687,
      "step": 30457
    },
    {
      "epoch": 7.1246637820137995,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6162,
      "step": 30458
    },
    {
      "epoch": 7.124897672786808,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7657,
      "step": 30459
    },
    {
      "epoch": 7.125131563559817,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6155,
      "step": 30460
    },
    {
      "epoch": 7.125365454332827,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 30461
    },
    {
      "epoch": 7.125599345105836,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6912,
      "step": 30462
    },
    {
      "epoch": 7.1258332358788445,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 30463
    },
    {
      "epoch": 7.126067126651853,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6413,
      "step": 30464
    },
    {
      "epoch": 7.126301017424862,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6921,
      "step": 30465
    },
    {
      "epoch": 7.126534908197872,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5276,
      "step": 30466
    },
    {
      "epoch": 7.126768798970881,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 30467
    },
    {
      "epoch": 7.12700268974389,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 30468
    },
    {
      "epoch": 7.1272365805168985,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5866,
      "step": 30469
    },
    {
      "epoch": 7.127470471289907,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 30470
    },
    {
      "epoch": 7.127704362062917,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 30471
    },
    {
      "epoch": 7.127938252835926,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 30472
    },
    {
      "epoch": 7.128172143608935,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 30473
    },
    {
      "epoch": 7.1284060343819435,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 30474
    },
    {
      "epoch": 7.128639925154952,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 30475
    },
    {
      "epoch": 7.128873815927962,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 30476
    },
    {
      "epoch": 7.129107706700971,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6201,
      "step": 30477
    },
    {
      "epoch": 7.12934159747398,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 30478
    },
    {
      "epoch": 7.129575488246989,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 30479
    },
    {
      "epoch": 7.1298093790199975,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 30480
    },
    {
      "epoch": 7.130043269793006,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 30481
    },
    {
      "epoch": 7.130277160566016,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4642,
      "step": 30482
    },
    {
      "epoch": 7.130511051339025,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8203,
      "step": 30483
    },
    {
      "epoch": 7.130744942112034,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6066,
      "step": 30484
    },
    {
      "epoch": 7.1309788328850425,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6373,
      "step": 30485
    },
    {
      "epoch": 7.131212723658051,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 30486
    },
    {
      "epoch": 7.131446614431061,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 30487
    },
    {
      "epoch": 7.13168050520407,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 30488
    },
    {
      "epoch": 7.131914395977079,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6908,
      "step": 30489
    },
    {
      "epoch": 7.132148286750088,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6541,
      "step": 30490
    },
    {
      "epoch": 7.1323821775230964,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 30491
    },
    {
      "epoch": 7.132616068296105,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 30492
    },
    {
      "epoch": 7.132849959069115,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 30493
    },
    {
      "epoch": 7.133083849842124,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 30494
    },
    {
      "epoch": 7.133317740615133,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 30495
    },
    {
      "epoch": 7.1335516313881415,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.4904,
      "step": 30496
    },
    {
      "epoch": 7.13378552216115,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 30497
    },
    {
      "epoch": 7.13401941293416,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8219,
      "step": 30498
    },
    {
      "epoch": 7.134253303707169,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8674,
      "step": 30499
    },
    {
      "epoch": 7.134487194480178,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6647,
      "step": 30500
    },
    {
      "epoch": 7.134487194480178,
      "eval_runtime": 4.6141,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 30500
    },
    {
      "epoch": 7.134721085253187,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 30501
    },
    {
      "epoch": 7.134954976026195,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3568,
      "step": 30502
    },
    {
      "epoch": 7.135188866799205,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6501,
      "step": 30503
    },
    {
      "epoch": 7.135422757572214,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5738,
      "step": 30504
    },
    {
      "epoch": 7.135656648345223,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 30505
    },
    {
      "epoch": 7.135890539118232,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 30506
    },
    {
      "epoch": 7.1361244298912405,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7161,
      "step": 30507
    },
    {
      "epoch": 7.13635832066425,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 30508
    },
    {
      "epoch": 7.136592211437259,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4278,
      "step": 30509
    },
    {
      "epoch": 7.136826102210268,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 30510
    },
    {
      "epoch": 7.137059992983277,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5231,
      "step": 30511
    },
    {
      "epoch": 7.137293883756286,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 30512
    },
    {
      "epoch": 7.137527774529294,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.6268,
      "step": 30513
    },
    {
      "epoch": 7.137761665302304,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 30514
    },
    {
      "epoch": 7.137995556075313,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 30515
    },
    {
      "epoch": 7.138229446848322,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 30516
    },
    {
      "epoch": 7.138463337621331,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6447,
      "step": 30517
    },
    {
      "epoch": 7.1386972283943395,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 30518
    },
    {
      "epoch": 7.138931119167349,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5905,
      "step": 30519
    },
    {
      "epoch": 7.139165009940358,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 30520
    },
    {
      "epoch": 7.139398900713367,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 30521
    },
    {
      "epoch": 7.139632791486376,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.5507,
      "step": 30522
    },
    {
      "epoch": 7.139866682259385,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5755,
      "step": 30523
    },
    {
      "epoch": 7.140100573032393,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6067,
      "step": 30524
    },
    {
      "epoch": 7.140334463805403,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 30525
    },
    {
      "epoch": 7.140568354578412,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 30526
    },
    {
      "epoch": 7.140802245351421,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 30527
    },
    {
      "epoch": 7.14103613612443,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 30528
    },
    {
      "epoch": 7.1412700268974385,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 30529
    },
    {
      "epoch": 7.141503917670448,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5282,
      "step": 30530
    },
    {
      "epoch": 7.141737808443457,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 30531
    },
    {
      "epoch": 7.141971699216466,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6584,
      "step": 30532
    },
    {
      "epoch": 7.142205589989475,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 30533
    },
    {
      "epoch": 7.142439480762484,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6573,
      "step": 30534
    },
    {
      "epoch": 7.142673371535493,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 30535
    },
    {
      "epoch": 7.142907262308502,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7019,
      "step": 30536
    },
    {
      "epoch": 7.143141153081511,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 30537
    },
    {
      "epoch": 7.14337504385452,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5887,
      "step": 30538
    },
    {
      "epoch": 7.143608934627529,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8201,
      "step": 30539
    },
    {
      "epoch": 7.143842825400538,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.6533,
      "step": 30540
    },
    {
      "epoch": 7.144076716173547,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 30541
    },
    {
      "epoch": 7.144310606946556,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6798,
      "step": 30542
    },
    {
      "epoch": 7.144544497719565,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6946,
      "step": 30543
    },
    {
      "epoch": 7.144778388492574,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6259,
      "step": 30544
    },
    {
      "epoch": 7.145012279265583,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 30545
    },
    {
      "epoch": 7.145246170038592,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7507,
      "step": 30546
    },
    {
      "epoch": 7.145480060811601,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 30547
    },
    {
      "epoch": 7.14571395158461,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.4861,
      "step": 30548
    },
    {
      "epoch": 7.145947842357619,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 30549
    },
    {
      "epoch": 7.146181733130628,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6525,
      "step": 30550
    },
    {
      "epoch": 7.146415623903637,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 30551
    },
    {
      "epoch": 7.146649514676646,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6,
      "step": 30552
    },
    {
      "epoch": 7.146883405449655,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 30553
    },
    {
      "epoch": 7.147117296222664,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7078,
      "step": 30554
    },
    {
      "epoch": 7.147351186995673,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7728,
      "step": 30555
    },
    {
      "epoch": 7.147585077768682,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 30556
    },
    {
      "epoch": 7.147818968541691,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.743,
      "step": 30557
    },
    {
      "epoch": 7.1480528593147,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 30558
    },
    {
      "epoch": 7.148286750087709,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 30559
    },
    {
      "epoch": 7.148520640860718,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 30560
    },
    {
      "epoch": 7.148754531633727,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 30561
    },
    {
      "epoch": 7.148988422406736,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4362,
      "step": 30562
    },
    {
      "epoch": 7.149222313179745,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 30563
    },
    {
      "epoch": 7.149456203952754,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6307,
      "step": 30564
    },
    {
      "epoch": 7.149690094725763,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 30565
    },
    {
      "epoch": 7.149923985498772,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9912,
      "step": 30566
    },
    {
      "epoch": 7.1501578762717815,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 30567
    },
    {
      "epoch": 7.15039176704479,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 30568
    },
    {
      "epoch": 7.150625657817799,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 30569
    },
    {
      "epoch": 7.150859548590808,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 30570
    },
    {
      "epoch": 7.151093439363817,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7232,
      "step": 30571
    },
    {
      "epoch": 7.1513273301368265,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 30572
    },
    {
      "epoch": 7.151561220909835,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 30573
    },
    {
      "epoch": 7.151795111682844,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 30574
    },
    {
      "epoch": 7.152029002455853,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 30575
    },
    {
      "epoch": 7.152262893228862,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5089,
      "step": 30576
    },
    {
      "epoch": 7.152496784001871,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5622,
      "step": 30577
    },
    {
      "epoch": 7.1527306747748804,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 30578
    },
    {
      "epoch": 7.152964565547889,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5658,
      "step": 30579
    },
    {
      "epoch": 7.153198456320898,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 30580
    },
    {
      "epoch": 7.153432347093907,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5093,
      "step": 30581
    },
    {
      "epoch": 7.153666237866916,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8524,
      "step": 30582
    },
    {
      "epoch": 7.1539001286399255,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 30583
    },
    {
      "epoch": 7.154134019412934,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 30584
    },
    {
      "epoch": 7.154367910185943,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 30585
    },
    {
      "epoch": 7.154601800958952,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7239,
      "step": 30586
    },
    {
      "epoch": 7.154835691731961,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5937,
      "step": 30587
    },
    {
      "epoch": 7.155069582504971,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 30588
    },
    {
      "epoch": 7.155303473277979,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 30589
    },
    {
      "epoch": 7.155537364050988,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5346,
      "step": 30590
    },
    {
      "epoch": 7.155771254823997,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6233,
      "step": 30591
    },
    {
      "epoch": 7.156005145597006,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 30592
    },
    {
      "epoch": 7.156239036370015,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 30593
    },
    {
      "epoch": 7.1564729271430245,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 30594
    },
    {
      "epoch": 7.156706817916033,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.638,
      "step": 30595
    },
    {
      "epoch": 7.156940708689042,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 30596
    },
    {
      "epoch": 7.157174599462051,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 30597
    },
    {
      "epoch": 7.15740849023506,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5979,
      "step": 30598
    },
    {
      "epoch": 7.15764238100807,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6408,
      "step": 30599
    },
    {
      "epoch": 7.157876271781078,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.622,
      "step": 30600
    },
    {
      "epoch": 7.157876271781078,
      "eval_runtime": 4.6121,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 30600
    },
    {
      "epoch": 7.158110162554087,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6084,
      "step": 30601
    },
    {
      "epoch": 7.158344053327096,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 30602
    },
    {
      "epoch": 7.158577944100105,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 30603
    },
    {
      "epoch": 7.158811834873115,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7615,
      "step": 30604
    },
    {
      "epoch": 7.1590457256461235,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7385,
      "step": 30605
    },
    {
      "epoch": 7.159279616419132,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 30606
    },
    {
      "epoch": 7.159513507192141,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 30607
    },
    {
      "epoch": 7.15974739796515,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 30608
    },
    {
      "epoch": 7.159981288738159,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 30609
    },
    {
      "epoch": 7.160215179511169,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 30610
    },
    {
      "epoch": 7.160449070284177,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 30611
    },
    {
      "epoch": 7.160682961057186,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8308,
      "step": 30612
    },
    {
      "epoch": 7.160916851830195,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 30613
    },
    {
      "epoch": 7.161150742603204,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6697,
      "step": 30614
    },
    {
      "epoch": 7.161384633376214,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 30615
    },
    {
      "epoch": 7.1616185241492225,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8776,
      "step": 30616
    },
    {
      "epoch": 7.161852414922231,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 30617
    },
    {
      "epoch": 7.16208630569524,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 30618
    },
    {
      "epoch": 7.162320196468249,
      "grad_norm": 2.546875,
      "learning_rate": 3e-05,
      "loss": 1.4767,
      "step": 30619
    },
    {
      "epoch": 7.162554087241259,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 30620
    },
    {
      "epoch": 7.162787978014268,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 30621
    },
    {
      "epoch": 7.163021868787276,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 30622
    },
    {
      "epoch": 7.163255759560285,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 30623
    },
    {
      "epoch": 7.163489650333294,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6241,
      "step": 30624
    },
    {
      "epoch": 7.163723541106303,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.3538,
      "step": 30625
    },
    {
      "epoch": 7.163957431879313,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 30626
    },
    {
      "epoch": 7.1641913226523215,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 30627
    },
    {
      "epoch": 7.16442521342533,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 30628
    },
    {
      "epoch": 7.164659104198339,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6397,
      "step": 30629
    },
    {
      "epoch": 7.164892994971348,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8027,
      "step": 30630
    },
    {
      "epoch": 7.165126885744358,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 30631
    },
    {
      "epoch": 7.165360776517367,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 30632
    },
    {
      "epoch": 7.165594667290375,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7502,
      "step": 30633
    },
    {
      "epoch": 7.165828558063384,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8276,
      "step": 30634
    },
    {
      "epoch": 7.166062448836393,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5558,
      "step": 30635
    },
    {
      "epoch": 7.166296339609403,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5341,
      "step": 30636
    },
    {
      "epoch": 7.166530230382412,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.743,
      "step": 30637
    },
    {
      "epoch": 7.1667641211554205,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 30638
    },
    {
      "epoch": 7.166998011928429,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.702,
      "step": 30639
    },
    {
      "epoch": 7.167231902701438,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 30640
    },
    {
      "epoch": 7.167465793474447,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7219,
      "step": 30641
    },
    {
      "epoch": 7.167699684247457,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6373,
      "step": 30642
    },
    {
      "epoch": 7.167933575020466,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 30643
    },
    {
      "epoch": 7.168167465793474,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7653,
      "step": 30644
    },
    {
      "epoch": 7.168401356566483,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 30645
    },
    {
      "epoch": 7.168635247339492,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.621,
      "step": 30646
    },
    {
      "epoch": 7.168869138112502,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 30647
    },
    {
      "epoch": 7.169103028885511,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 30648
    },
    {
      "epoch": 7.1693369196585195,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8434,
      "step": 30649
    },
    {
      "epoch": 7.169570810431528,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6699,
      "step": 30650
    },
    {
      "epoch": 7.169804701204537,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 30651
    },
    {
      "epoch": 7.170038591977547,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5268,
      "step": 30652
    },
    {
      "epoch": 7.170272482750556,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6258,
      "step": 30653
    },
    {
      "epoch": 7.170506373523565,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 30654
    },
    {
      "epoch": 7.170740264296573,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6112,
      "step": 30655
    },
    {
      "epoch": 7.170974155069582,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 30656
    },
    {
      "epoch": 7.171208045842591,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 30657
    },
    {
      "epoch": 7.171441936615601,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 30658
    },
    {
      "epoch": 7.17167582738861,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.866,
      "step": 30659
    },
    {
      "epoch": 7.1719097181616185,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 30660
    },
    {
      "epoch": 7.172143608934627,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.663,
      "step": 30661
    },
    {
      "epoch": 7.172377499707636,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 30662
    },
    {
      "epoch": 7.172611390480646,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 30663
    },
    {
      "epoch": 7.172845281253655,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5439,
      "step": 30664
    },
    {
      "epoch": 7.173079172026664,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5865,
      "step": 30665
    },
    {
      "epoch": 7.173313062799672,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 30666
    },
    {
      "epoch": 7.173546953572681,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 30667
    },
    {
      "epoch": 7.173780844345691,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 30668
    },
    {
      "epoch": 7.1740147351187,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 30669
    },
    {
      "epoch": 7.174248625891709,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6064,
      "step": 30670
    },
    {
      "epoch": 7.1744825166647175,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5699,
      "step": 30671
    },
    {
      "epoch": 7.174716407437726,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 30672
    },
    {
      "epoch": 7.174950298210735,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7306,
      "step": 30673
    },
    {
      "epoch": 7.175184188983745,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 30674
    },
    {
      "epoch": 7.175418079756754,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 30675
    },
    {
      "epoch": 7.175651970529763,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 30676
    },
    {
      "epoch": 7.175885861302771,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 30677
    },
    {
      "epoch": 7.17611975207578,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 30678
    },
    {
      "epoch": 7.17635364284879,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 30679
    },
    {
      "epoch": 7.176587533621799,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 30680
    },
    {
      "epoch": 7.176821424394808,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 30681
    },
    {
      "epoch": 7.1770553151678165,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 30682
    },
    {
      "epoch": 7.177289205940825,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.6471,
      "step": 30683
    },
    {
      "epoch": 7.177523096713835,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6102,
      "step": 30684
    },
    {
      "epoch": 7.177756987486844,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 30685
    },
    {
      "epoch": 7.177990878259853,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5198,
      "step": 30686
    },
    {
      "epoch": 7.1782247690328616,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7657,
      "step": 30687
    },
    {
      "epoch": 7.17845865980587,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 30688
    },
    {
      "epoch": 7.17869255057888,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 30689
    },
    {
      "epoch": 7.178926441351889,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.592,
      "step": 30690
    },
    {
      "epoch": 7.179160332124898,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8174,
      "step": 30691
    },
    {
      "epoch": 7.179394222897907,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6548,
      "step": 30692
    },
    {
      "epoch": 7.1796281136709155,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 30693
    },
    {
      "epoch": 7.179862004443924,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5242,
      "step": 30694
    },
    {
      "epoch": 7.180095895216934,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 30695
    },
    {
      "epoch": 7.180329785989943,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6014,
      "step": 30696
    },
    {
      "epoch": 7.180563676762952,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 30697
    },
    {
      "epoch": 7.1807975675359605,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6126,
      "step": 30698
    },
    {
      "epoch": 7.181031458308969,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 30699
    },
    {
      "epoch": 7.181265349081979,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5415,
      "step": 30700
    },
    {
      "epoch": 7.181265349081979,
      "eval_runtime": 4.6514,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 30700
    },
    {
      "epoch": 7.181499239854988,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 30701
    },
    {
      "epoch": 7.181733130627997,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 30702
    },
    {
      "epoch": 7.181967021401006,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5655,
      "step": 30703
    },
    {
      "epoch": 7.1822009121740145,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4943,
      "step": 30704
    },
    {
      "epoch": 7.182434802947023,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 30705
    },
    {
      "epoch": 7.182668693720033,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 30706
    },
    {
      "epoch": 7.182902584493042,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0348,
      "step": 30707
    },
    {
      "epoch": 7.183136475266051,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 30708
    },
    {
      "epoch": 7.1833703660390595,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 30709
    },
    {
      "epoch": 7.183604256812068,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 30710
    },
    {
      "epoch": 7.183838147585078,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7548,
      "step": 30711
    },
    {
      "epoch": 7.184072038358087,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 30712
    },
    {
      "epoch": 7.184305929131096,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 30713
    },
    {
      "epoch": 7.184539819904105,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 30714
    },
    {
      "epoch": 7.1847737106771135,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 30715
    },
    {
      "epoch": 7.185007601450123,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 30716
    },
    {
      "epoch": 7.185241492223132,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5142,
      "step": 30717
    },
    {
      "epoch": 7.185475382996141,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 30718
    },
    {
      "epoch": 7.18570927376915,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6579,
      "step": 30719
    },
    {
      "epoch": 7.1859431645421585,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7362,
      "step": 30720
    },
    {
      "epoch": 7.186177055315168,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6747,
      "step": 30721
    },
    {
      "epoch": 7.186410946088177,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4271,
      "step": 30722
    },
    {
      "epoch": 7.186644836861186,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 30723
    },
    {
      "epoch": 7.186878727634195,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6232,
      "step": 30724
    },
    {
      "epoch": 7.187112618407204,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 30725
    },
    {
      "epoch": 7.1873465091802125,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 30726
    },
    {
      "epoch": 7.187580399953222,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.5331,
      "step": 30727
    },
    {
      "epoch": 7.187814290726231,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 30728
    },
    {
      "epoch": 7.18804818149924,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6681,
      "step": 30729
    },
    {
      "epoch": 7.188282072272249,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 30730
    },
    {
      "epoch": 7.1885159630452575,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 1.6797,
      "step": 30731
    },
    {
      "epoch": 7.188749853818267,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 30732
    },
    {
      "epoch": 7.188983744591276,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6583,
      "step": 30733
    },
    {
      "epoch": 7.189217635364285,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6788,
      "step": 30734
    },
    {
      "epoch": 7.189451526137294,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 30735
    },
    {
      "epoch": 7.189685416910303,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.6879,
      "step": 30736
    },
    {
      "epoch": 7.1899193076833114,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5584,
      "step": 30737
    },
    {
      "epoch": 7.190153198456321,
      "grad_norm": 7.65625,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 30738
    },
    {
      "epoch": 7.19038708922933,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5895,
      "step": 30739
    },
    {
      "epoch": 7.190620980002339,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 30740
    },
    {
      "epoch": 7.190854870775348,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 30741
    },
    {
      "epoch": 7.1910887615483565,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6086,
      "step": 30742
    },
    {
      "epoch": 7.191322652321366,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.651,
      "step": 30743
    },
    {
      "epoch": 7.191556543094375,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 30744
    },
    {
      "epoch": 7.191790433867384,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 30745
    },
    {
      "epoch": 7.192024324640393,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 30746
    },
    {
      "epoch": 7.192258215413402,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6274,
      "step": 30747
    },
    {
      "epoch": 7.192492106186411,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 30748
    },
    {
      "epoch": 7.19272599695942,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 30749
    },
    {
      "epoch": 7.192959887732429,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 30750
    },
    {
      "epoch": 7.193193778505438,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 30751
    },
    {
      "epoch": 7.193427669278447,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7472,
      "step": 30752
    },
    {
      "epoch": 7.193661560051456,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 30753
    },
    {
      "epoch": 7.193895450824465,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4135,
      "step": 30754
    },
    {
      "epoch": 7.194129341597474,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7239,
      "step": 30755
    },
    {
      "epoch": 7.194363232370483,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 30756
    },
    {
      "epoch": 7.194597123143492,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 30757
    },
    {
      "epoch": 7.194831013916501,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 30758
    },
    {
      "epoch": 7.19506490468951,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 30759
    },
    {
      "epoch": 7.195298795462519,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 30760
    },
    {
      "epoch": 7.195532686235528,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 30761
    },
    {
      "epoch": 7.195766577008537,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 30762
    },
    {
      "epoch": 7.196000467781546,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5056,
      "step": 30763
    },
    {
      "epoch": 7.196234358554555,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 30764
    },
    {
      "epoch": 7.196468249327564,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 30765
    },
    {
      "epoch": 7.196702140100573,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6747,
      "step": 30766
    },
    {
      "epoch": 7.196936030873582,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6837,
      "step": 30767
    },
    {
      "epoch": 7.197169921646591,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5687,
      "step": 30768
    },
    {
      "epoch": 7.1974038124196,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6527,
      "step": 30769
    },
    {
      "epoch": 7.197637703192609,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5897,
      "step": 30770
    },
    {
      "epoch": 7.197871593965618,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8653,
      "step": 30771
    },
    {
      "epoch": 7.198105484738627,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 30772
    },
    {
      "epoch": 7.198339375511636,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 30773
    },
    {
      "epoch": 7.198573266284645,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5443,
      "step": 30774
    },
    {
      "epoch": 7.198807157057654,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7306,
      "step": 30775
    },
    {
      "epoch": 7.199041047830663,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 30776
    },
    {
      "epoch": 7.199274938603672,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 30777
    },
    {
      "epoch": 7.199508829376681,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 30778
    },
    {
      "epoch": 7.19974272014969,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7236,
      "step": 30779
    },
    {
      "epoch": 7.1999766109226995,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.581,
      "step": 30780
    },
    {
      "epoch": 7.200210501695708,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5242,
      "step": 30781
    },
    {
      "epoch": 7.200444392468717,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 30782
    },
    {
      "epoch": 7.200678283241726,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4311,
      "step": 30783
    },
    {
      "epoch": 7.200912174014735,
      "grad_norm": 9.25,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 30784
    },
    {
      "epoch": 7.2011460647877445,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 30785
    },
    {
      "epoch": 7.201379955560753,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 30786
    },
    {
      "epoch": 7.201613846333762,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 30787
    },
    {
      "epoch": 7.201847737106771,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 30788
    },
    {
      "epoch": 7.20208162787978,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5476,
      "step": 30789
    },
    {
      "epoch": 7.202315518652789,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 30790
    },
    {
      "epoch": 7.2025494094257985,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 30791
    },
    {
      "epoch": 7.202783300198807,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6305,
      "step": 30792
    },
    {
      "epoch": 7.203017190971816,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.423,
      "step": 30793
    },
    {
      "epoch": 7.203251081744825,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 30794
    },
    {
      "epoch": 7.203484972517834,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 30795
    },
    {
      "epoch": 7.2037188632908435,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 30796
    },
    {
      "epoch": 7.203952754063852,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 30797
    },
    {
      "epoch": 7.204186644836861,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 30798
    },
    {
      "epoch": 7.20442053560987,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5442,
      "step": 30799
    },
    {
      "epoch": 7.204654426382879,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6826,
      "step": 30800
    },
    {
      "epoch": 7.204654426382879,
      "eval_runtime": 4.5875,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 30800
    },
    {
      "epoch": 7.204888317155889,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 30801
    },
    {
      "epoch": 7.2051222079288975,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 30802
    },
    {
      "epoch": 7.205356098701906,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 30803
    },
    {
      "epoch": 7.205589989474915,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 30804
    },
    {
      "epoch": 7.205823880247924,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 30805
    },
    {
      "epoch": 7.206057771020933,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 30806
    },
    {
      "epoch": 7.2062916617939425,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7597,
      "step": 30807
    },
    {
      "epoch": 7.206525552566951,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5944,
      "step": 30808
    },
    {
      "epoch": 7.20675944333996,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 30809
    },
    {
      "epoch": 7.206993334112969,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 30810
    },
    {
      "epoch": 7.207227224885978,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 30811
    },
    {
      "epoch": 7.207461115658988,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6682,
      "step": 30812
    },
    {
      "epoch": 7.2076950064319965,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 30813
    },
    {
      "epoch": 7.207928897205005,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 30814
    },
    {
      "epoch": 7.208162787978014,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 30815
    },
    {
      "epoch": 7.208396678751023,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 30816
    },
    {
      "epoch": 7.208630569524033,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 30817
    },
    {
      "epoch": 7.2088644602970415,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 30818
    },
    {
      "epoch": 7.20909835107005,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 30819
    },
    {
      "epoch": 7.209332241843059,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 30820
    },
    {
      "epoch": 7.209566132616068,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 30821
    },
    {
      "epoch": 7.209800023389077,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 30822
    },
    {
      "epoch": 7.210033914162087,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 30823
    },
    {
      "epoch": 7.2102678049350954,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6623,
      "step": 30824
    },
    {
      "epoch": 7.210501695708104,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 30825
    },
    {
      "epoch": 7.210735586481113,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 30826
    },
    {
      "epoch": 7.210969477254122,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7037,
      "step": 30827
    },
    {
      "epoch": 7.211203368027132,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 30828
    },
    {
      "epoch": 7.2114372588001405,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 30829
    },
    {
      "epoch": 7.211671149573149,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 30830
    },
    {
      "epoch": 7.211905040346158,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8223,
      "step": 30831
    },
    {
      "epoch": 7.212138931119167,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 30832
    },
    {
      "epoch": 7.212372821892177,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 30833
    },
    {
      "epoch": 7.212606712665186,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 30834
    },
    {
      "epoch": 7.212840603438194,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 30835
    },
    {
      "epoch": 7.213074494211203,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 30836
    },
    {
      "epoch": 7.213308384984212,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 30837
    },
    {
      "epoch": 7.213542275757221,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 30838
    },
    {
      "epoch": 7.213776166530231,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 30839
    },
    {
      "epoch": 7.2140100573032395,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6326,
      "step": 30840
    },
    {
      "epoch": 7.214243948076248,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 30841
    },
    {
      "epoch": 7.214477838849257,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 30842
    },
    {
      "epoch": 7.214711729622266,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7203,
      "step": 30843
    },
    {
      "epoch": 7.214945620395276,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0791,
      "step": 30844
    },
    {
      "epoch": 7.215179511168285,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9622,
      "step": 30845
    },
    {
      "epoch": 7.215413401941293,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 30846
    },
    {
      "epoch": 7.215647292714302,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.4355,
      "step": 30847
    },
    {
      "epoch": 7.215881183487311,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 30848
    },
    {
      "epoch": 7.216115074260321,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6677,
      "step": 30849
    },
    {
      "epoch": 7.21634896503333,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 30850
    },
    {
      "epoch": 7.2165828558063385,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6444,
      "step": 30851
    },
    {
      "epoch": 7.216816746579347,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 30852
    },
    {
      "epoch": 7.217050637352356,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5468,
      "step": 30853
    },
    {
      "epoch": 7.217284528125365,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 30854
    },
    {
      "epoch": 7.217518418898375,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6682,
      "step": 30855
    },
    {
      "epoch": 7.217752309671384,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.467,
      "step": 30856
    },
    {
      "epoch": 7.217986200444392,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 30857
    },
    {
      "epoch": 7.218220091217401,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 30858
    },
    {
      "epoch": 7.21845398199041,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8629,
      "step": 30859
    },
    {
      "epoch": 7.21868787276342,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.6772,
      "step": 30860
    },
    {
      "epoch": 7.218921763536429,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.4581,
      "step": 30861
    },
    {
      "epoch": 7.2191556543094375,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5945,
      "step": 30862
    },
    {
      "epoch": 7.219389545082446,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 30863
    },
    {
      "epoch": 7.219623435855455,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7107,
      "step": 30864
    },
    {
      "epoch": 7.219857326628465,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6275,
      "step": 30865
    },
    {
      "epoch": 7.220091217401474,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 30866
    },
    {
      "epoch": 7.220325108174483,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6491,
      "step": 30867
    },
    {
      "epoch": 7.220558998947491,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 30868
    },
    {
      "epoch": 7.2207928897205,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 30869
    },
    {
      "epoch": 7.22102678049351,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 30870
    },
    {
      "epoch": 7.221260671266519,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 30871
    },
    {
      "epoch": 7.221494562039528,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5229,
      "step": 30872
    },
    {
      "epoch": 7.2217284528125365,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.713,
      "step": 30873
    },
    {
      "epoch": 7.221962343585545,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 30874
    },
    {
      "epoch": 7.222196234358554,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 30875
    },
    {
      "epoch": 7.222430125131564,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7769,
      "step": 30876
    },
    {
      "epoch": 7.222664015904573,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 30877
    },
    {
      "epoch": 7.222897906677582,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6327,
      "step": 30878
    },
    {
      "epoch": 7.22313179745059,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 30879
    },
    {
      "epoch": 7.223365688223599,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 30880
    },
    {
      "epoch": 7.223599578996609,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6878,
      "step": 30881
    },
    {
      "epoch": 7.223833469769618,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 30882
    },
    {
      "epoch": 7.224067360542627,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 30883
    },
    {
      "epoch": 7.2243012513156355,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 30884
    },
    {
      "epoch": 7.224535142088644,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.5738,
      "step": 30885
    },
    {
      "epoch": 7.224769032861653,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 30886
    },
    {
      "epoch": 7.225002923634663,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 30887
    },
    {
      "epoch": 7.225236814407672,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 30888
    },
    {
      "epoch": 7.225470705180681,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6935,
      "step": 30889
    },
    {
      "epoch": 7.225704595953689,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4669,
      "step": 30890
    },
    {
      "epoch": 7.225938486726698,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 30891
    },
    {
      "epoch": 7.226172377499708,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.4678,
      "step": 30892
    },
    {
      "epoch": 7.226406268272717,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7311,
      "step": 30893
    },
    {
      "epoch": 7.226640159045726,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6442,
      "step": 30894
    },
    {
      "epoch": 7.2268740498187345,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 30895
    },
    {
      "epoch": 7.227107940591743,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7107,
      "step": 30896
    },
    {
      "epoch": 7.227341831364753,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5132,
      "step": 30897
    },
    {
      "epoch": 7.227575722137762,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.3997,
      "step": 30898
    },
    {
      "epoch": 7.227809612910771,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 30899
    },
    {
      "epoch": 7.22804350368378,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 30900
    },
    {
      "epoch": 7.22804350368378,
      "eval_runtime": 4.5699,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 30900
    },
    {
      "epoch": 7.228277394456788,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1605,
      "step": 30901
    },
    {
      "epoch": 7.228511285229798,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 30902
    },
    {
      "epoch": 7.228745176002807,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8707,
      "step": 30903
    },
    {
      "epoch": 7.228979066775816,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 30904
    },
    {
      "epoch": 7.229212957548825,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.4698,
      "step": 30905
    },
    {
      "epoch": 7.2294468483218335,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 30906
    },
    {
      "epoch": 7.229680739094842,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 30907
    },
    {
      "epoch": 7.229914629867852,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 30908
    },
    {
      "epoch": 7.230148520640861,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 30909
    },
    {
      "epoch": 7.23038241141387,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 30910
    },
    {
      "epoch": 7.230616302186879,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 30911
    },
    {
      "epoch": 7.230850192959887,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 30912
    },
    {
      "epoch": 7.231084083732897,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 30913
    },
    {
      "epoch": 7.231317974505906,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6156,
      "step": 30914
    },
    {
      "epoch": 7.231551865278915,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.631,
      "step": 30915
    },
    {
      "epoch": 7.231785756051924,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 30916
    },
    {
      "epoch": 7.2320196468249325,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 30917
    },
    {
      "epoch": 7.232253537597941,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 30918
    },
    {
      "epoch": 7.232487428370951,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 30919
    },
    {
      "epoch": 7.23272131914396,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.4479,
      "step": 30920
    },
    {
      "epoch": 7.232955209916969,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 30921
    },
    {
      "epoch": 7.233189100689978,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 30922
    },
    {
      "epoch": 7.233422991462986,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 30923
    },
    {
      "epoch": 7.233656882235996,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 30924
    },
    {
      "epoch": 7.233890773009005,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 30925
    },
    {
      "epoch": 7.234124663782014,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7932,
      "step": 30926
    },
    {
      "epoch": 7.234358554555023,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5974,
      "step": 30927
    },
    {
      "epoch": 7.2345924453280315,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6516,
      "step": 30928
    },
    {
      "epoch": 7.234826336101041,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 30929
    },
    {
      "epoch": 7.23506022687405,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4845,
      "step": 30930
    },
    {
      "epoch": 7.235294117647059,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 30931
    },
    {
      "epoch": 7.235528008420068,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 30932
    },
    {
      "epoch": 7.2357618991930766,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 30933
    },
    {
      "epoch": 7.235995789966086,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 30934
    },
    {
      "epoch": 7.236229680739095,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6007,
      "step": 30935
    },
    {
      "epoch": 7.236463571512104,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 30936
    },
    {
      "epoch": 7.236697462285113,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7273,
      "step": 30937
    },
    {
      "epoch": 7.236931353058122,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 30938
    },
    {
      "epoch": 7.2371652438311305,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 30939
    },
    {
      "epoch": 7.23739913460414,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 30940
    },
    {
      "epoch": 7.237633025377149,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.4569,
      "step": 30941
    },
    {
      "epoch": 7.237866916150158,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6585,
      "step": 30942
    },
    {
      "epoch": 7.238100806923167,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 30943
    },
    {
      "epoch": 7.2383346976961755,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5591,
      "step": 30944
    },
    {
      "epoch": 7.238568588469185,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.5813,
      "step": 30945
    },
    {
      "epoch": 7.238802479242194,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7607,
      "step": 30946
    },
    {
      "epoch": 7.239036370015203,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.4768,
      "step": 30947
    },
    {
      "epoch": 7.239270260788212,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 30948
    },
    {
      "epoch": 7.239504151561221,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6312,
      "step": 30949
    },
    {
      "epoch": 7.2397380423342295,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 30950
    },
    {
      "epoch": 7.239971933107239,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 30951
    },
    {
      "epoch": 7.240205823880248,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 30952
    },
    {
      "epoch": 7.240439714653257,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 30953
    },
    {
      "epoch": 7.240673605426266,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.748,
      "step": 30954
    },
    {
      "epoch": 7.2409074961992745,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.488,
      "step": 30955
    },
    {
      "epoch": 7.241141386972284,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 30956
    },
    {
      "epoch": 7.241375277745293,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6067,
      "step": 30957
    },
    {
      "epoch": 7.241609168518302,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 30958
    },
    {
      "epoch": 7.241843059291311,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 30959
    },
    {
      "epoch": 7.24207695006432,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 30960
    },
    {
      "epoch": 7.242310840837329,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 30961
    },
    {
      "epoch": 7.242544731610338,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 30962
    },
    {
      "epoch": 7.242778622383347,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 30963
    },
    {
      "epoch": 7.243012513156356,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6452,
      "step": 30964
    },
    {
      "epoch": 7.243246403929365,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 30965
    },
    {
      "epoch": 7.243480294702374,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 30966
    },
    {
      "epoch": 7.243714185475383,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 30967
    },
    {
      "epoch": 7.243948076248392,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 30968
    },
    {
      "epoch": 7.244181967021401,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.4767,
      "step": 30969
    },
    {
      "epoch": 7.24441585779441,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 30970
    },
    {
      "epoch": 7.244649748567419,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6887,
      "step": 30971
    },
    {
      "epoch": 7.244883639340428,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5874,
      "step": 30972
    },
    {
      "epoch": 7.245117530113437,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7253,
      "step": 30973
    },
    {
      "epoch": 7.245351420886446,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 30974
    },
    {
      "epoch": 7.245585311659455,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.4204,
      "step": 30975
    },
    {
      "epoch": 7.245819202432464,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5654,
      "step": 30976
    },
    {
      "epoch": 7.246053093205473,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 30977
    },
    {
      "epoch": 7.246286983978482,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4115,
      "step": 30978
    },
    {
      "epoch": 7.246520874751491,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 30979
    },
    {
      "epoch": 7.2467547655245,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8508,
      "step": 30980
    },
    {
      "epoch": 7.246988656297509,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5747,
      "step": 30981
    },
    {
      "epoch": 7.2472225470705185,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 30982
    },
    {
      "epoch": 7.247456437843527,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6419,
      "step": 30983
    },
    {
      "epoch": 7.247690328616536,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 30984
    },
    {
      "epoch": 7.247924219389545,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.598,
      "step": 30985
    },
    {
      "epoch": 7.248158110162554,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 30986
    },
    {
      "epoch": 7.248392000935563,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 30987
    },
    {
      "epoch": 7.248625891708572,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.4302,
      "step": 30988
    },
    {
      "epoch": 7.248859782481581,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6783,
      "step": 30989
    },
    {
      "epoch": 7.24909367325459,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 30990
    },
    {
      "epoch": 7.249327564027599,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9314,
      "step": 30991
    },
    {
      "epoch": 7.249561454800608,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 30992
    },
    {
      "epoch": 7.2497953455736175,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 30993
    },
    {
      "epoch": 7.250029236346626,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 30994
    },
    {
      "epoch": 7.250263127119635,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 30995
    },
    {
      "epoch": 7.250497017892644,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 30996
    },
    {
      "epoch": 7.250730908665653,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 30997
    },
    {
      "epoch": 7.250964799438663,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 30998
    },
    {
      "epoch": 7.251198690211671,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.3142,
      "step": 30999
    },
    {
      "epoch": 7.25143258098468,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 31000
    },
    {
      "epoch": 7.25143258098468,
      "eval_runtime": 4.6259,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 31000
    },
    {
      "epoch": 7.251666471757689,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 31001
    },
    {
      "epoch": 7.251900362530698,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 31002
    },
    {
      "epoch": 7.252134253303707,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 31003
    },
    {
      "epoch": 7.2523681440767165,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7045,
      "step": 31004
    },
    {
      "epoch": 7.252602034849725,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6576,
      "step": 31005
    },
    {
      "epoch": 7.252835925622734,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 31006
    },
    {
      "epoch": 7.253069816395743,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 31007
    },
    {
      "epoch": 7.253303707168752,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 31008
    },
    {
      "epoch": 7.253537597941762,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 31009
    },
    {
      "epoch": 7.25377148871477,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 31010
    },
    {
      "epoch": 7.254005379487779,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6452,
      "step": 31011
    },
    {
      "epoch": 7.254239270260788,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 31012
    },
    {
      "epoch": 7.254473161033797,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5535,
      "step": 31013
    },
    {
      "epoch": 7.254707051806806,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6159,
      "step": 31014
    },
    {
      "epoch": 7.2549409425798155,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7439,
      "step": 31015
    },
    {
      "epoch": 7.255174833352824,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 31016
    },
    {
      "epoch": 7.255408724125833,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 31017
    },
    {
      "epoch": 7.255642614898842,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 31018
    },
    {
      "epoch": 7.255876505671852,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 31019
    },
    {
      "epoch": 7.2561103964448606,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6277,
      "step": 31020
    },
    {
      "epoch": 7.256344287217869,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 31021
    },
    {
      "epoch": 7.256578177990878,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6571,
      "step": 31022
    },
    {
      "epoch": 7.256812068763887,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.5148,
      "step": 31023
    },
    {
      "epoch": 7.257045959536896,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 31024
    },
    {
      "epoch": 7.257279850309906,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 31025
    },
    {
      "epoch": 7.2575137410829145,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 31026
    },
    {
      "epoch": 7.257747631855923,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 31027
    },
    {
      "epoch": 7.257981522628932,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5926,
      "step": 31028
    },
    {
      "epoch": 7.258215413401941,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.66,
      "step": 31029
    },
    {
      "epoch": 7.258449304174951,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 31030
    },
    {
      "epoch": 7.2586831949479595,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 31031
    },
    {
      "epoch": 7.258917085720968,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 31032
    },
    {
      "epoch": 7.259150976493977,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5077,
      "step": 31033
    },
    {
      "epoch": 7.259384867266986,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 31034
    },
    {
      "epoch": 7.259618758039995,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 31035
    },
    {
      "epoch": 7.259852648813005,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 31036
    },
    {
      "epoch": 7.2600865395860135,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.3576,
      "step": 31037
    },
    {
      "epoch": 7.260320430359022,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 31038
    },
    {
      "epoch": 7.260554321132031,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.4509,
      "step": 31039
    },
    {
      "epoch": 7.26078821190504,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 31040
    },
    {
      "epoch": 7.26102210267805,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 31041
    },
    {
      "epoch": 7.2612559934510585,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 31042
    },
    {
      "epoch": 7.261489884224067,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6299,
      "step": 31043
    },
    {
      "epoch": 7.261723774997076,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 31044
    },
    {
      "epoch": 7.261957665770085,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 31045
    },
    {
      "epoch": 7.262191556543095,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 31046
    },
    {
      "epoch": 7.262425447316104,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7134,
      "step": 31047
    },
    {
      "epoch": 7.2626593380891125,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6732,
      "step": 31048
    },
    {
      "epoch": 7.262893228862121,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6516,
      "step": 31049
    },
    {
      "epoch": 7.26312711963513,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4478,
      "step": 31050
    },
    {
      "epoch": 7.26336101040814,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 31051
    },
    {
      "epoch": 7.263594901181149,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7281,
      "step": 31052
    },
    {
      "epoch": 7.2638287919541575,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6818,
      "step": 31053
    },
    {
      "epoch": 7.264062682727166,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4361,
      "step": 31054
    },
    {
      "epoch": 7.264296573500175,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6156,
      "step": 31055
    },
    {
      "epoch": 7.264530464273184,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 31056
    },
    {
      "epoch": 7.264764355046194,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 31057
    },
    {
      "epoch": 7.264998245819203,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 31058
    },
    {
      "epoch": 7.2652321365922115,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6028,
      "step": 31059
    },
    {
      "epoch": 7.26546602736522,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 31060
    },
    {
      "epoch": 7.265699918138229,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 31061
    },
    {
      "epoch": 7.265933808911239,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 31062
    },
    {
      "epoch": 7.266167699684248,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5233,
      "step": 31063
    },
    {
      "epoch": 7.2664015904572565,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 31064
    },
    {
      "epoch": 7.266635481230265,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 31065
    },
    {
      "epoch": 7.266869372003274,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 31066
    },
    {
      "epoch": 7.267103262776283,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.4309,
      "step": 31067
    },
    {
      "epoch": 7.267337153549293,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.5638,
      "step": 31068
    },
    {
      "epoch": 7.267571044322302,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4817,
      "step": 31069
    },
    {
      "epoch": 7.2678049350953104,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6816,
      "step": 31070
    },
    {
      "epoch": 7.268038825868319,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 31071
    },
    {
      "epoch": 7.268272716641328,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 31072
    },
    {
      "epoch": 7.268506607414338,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 31073
    },
    {
      "epoch": 7.268740498187347,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 31074
    },
    {
      "epoch": 7.2689743889603555,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6677,
      "step": 31075
    },
    {
      "epoch": 7.269208279733364,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 31076
    },
    {
      "epoch": 7.269442170506373,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 31077
    },
    {
      "epoch": 7.269676061279383,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 31078
    },
    {
      "epoch": 7.269909952052392,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6156,
      "step": 31079
    },
    {
      "epoch": 7.270143842825401,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 31080
    },
    {
      "epoch": 7.270377733598409,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.769,
      "step": 31081
    },
    {
      "epoch": 7.270611624371418,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 31082
    },
    {
      "epoch": 7.270845515144428,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 31083
    },
    {
      "epoch": 7.271079405917437,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5354,
      "step": 31084
    },
    {
      "epoch": 7.271313296690446,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 31085
    },
    {
      "epoch": 7.2715471874634545,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6045,
      "step": 31086
    },
    {
      "epoch": 7.271781078236463,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 31087
    },
    {
      "epoch": 7.272014969009472,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4487,
      "step": 31088
    },
    {
      "epoch": 7.272248859782482,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6408,
      "step": 31089
    },
    {
      "epoch": 7.272482750555491,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 31090
    },
    {
      "epoch": 7.2727166413285,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 31091
    },
    {
      "epoch": 7.272950532101508,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 31092
    },
    {
      "epoch": 7.273184422874517,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 31093
    },
    {
      "epoch": 7.273418313647527,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6067,
      "step": 31094
    },
    {
      "epoch": 7.273652204420536,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7718,
      "step": 31095
    },
    {
      "epoch": 7.273886095193545,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4342,
      "step": 31096
    },
    {
      "epoch": 7.2741199859665535,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 31097
    },
    {
      "epoch": 7.274353876739562,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 31098
    },
    {
      "epoch": 7.274587767512571,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7531,
      "step": 31099
    },
    {
      "epoch": 7.274821658285581,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.7114,
      "step": 31100
    },
    {
      "epoch": 7.274821658285581,
      "eval_runtime": 4.623,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 31100
    },
    {
      "epoch": 7.27505554905859,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 31101
    },
    {
      "epoch": 7.275289439831599,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6568,
      "step": 31102
    },
    {
      "epoch": 7.275523330604607,
      "grad_norm": 9.0,
      "learning_rate": 3e-05,
      "loss": 2.33,
      "step": 31103
    },
    {
      "epoch": 7.275757221377616,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 31104
    },
    {
      "epoch": 7.275991112150626,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 31105
    },
    {
      "epoch": 7.276225002923635,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 31106
    },
    {
      "epoch": 7.276458893696644,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4935,
      "step": 31107
    },
    {
      "epoch": 7.2766927844696525,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.636,
      "step": 31108
    },
    {
      "epoch": 7.276926675242661,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 31109
    },
    {
      "epoch": 7.277160566015671,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 31110
    },
    {
      "epoch": 7.27739445678868,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 31111
    },
    {
      "epoch": 7.277628347561689,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6881,
      "step": 31112
    },
    {
      "epoch": 7.277862238334698,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.513,
      "step": 31113
    },
    {
      "epoch": 7.278096129107706,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 31114
    },
    {
      "epoch": 7.278330019880716,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 31115
    },
    {
      "epoch": 7.278563910653725,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5242,
      "step": 31116
    },
    {
      "epoch": 7.278797801426734,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 31117
    },
    {
      "epoch": 7.279031692199743,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6048,
      "step": 31118
    },
    {
      "epoch": 7.2792655829727515,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 31119
    },
    {
      "epoch": 7.27949947374576,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 31120
    },
    {
      "epoch": 7.27973336451877,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5413,
      "step": 31121
    },
    {
      "epoch": 7.279967255291779,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 31122
    },
    {
      "epoch": 7.280201146064788,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 31123
    },
    {
      "epoch": 7.280435036837797,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 31124
    },
    {
      "epoch": 7.280668927610805,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.3574,
      "step": 31125
    },
    {
      "epoch": 7.280902818383815,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 31126
    },
    {
      "epoch": 7.281136709156824,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 31127
    },
    {
      "epoch": 7.281370599929833,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 31128
    },
    {
      "epoch": 7.281604490702842,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6394,
      "step": 31129
    },
    {
      "epoch": 7.2818383814758505,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7628,
      "step": 31130
    },
    {
      "epoch": 7.282072272248859,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 31131
    },
    {
      "epoch": 7.282306163021869,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 31132
    },
    {
      "epoch": 7.282540053794878,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6185,
      "step": 31133
    },
    {
      "epoch": 7.282773944567887,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 31134
    },
    {
      "epoch": 7.283007835340896,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 31135
    },
    {
      "epoch": 7.283241726113904,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 31136
    },
    {
      "epoch": 7.283475616886914,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 31137
    },
    {
      "epoch": 7.283709507659923,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 31138
    },
    {
      "epoch": 7.283943398432932,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 31139
    },
    {
      "epoch": 7.284177289205941,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.5688,
      "step": 31140
    },
    {
      "epoch": 7.2844111799789495,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 31141
    },
    {
      "epoch": 7.284645070751959,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.3522,
      "step": 31142
    },
    {
      "epoch": 7.284878961524968,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.658,
      "step": 31143
    },
    {
      "epoch": 7.285112852297977,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6256,
      "step": 31144
    },
    {
      "epoch": 7.285346743070986,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7655,
      "step": 31145
    },
    {
      "epoch": 7.285580633843995,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 31146
    },
    {
      "epoch": 7.285814524617004,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 31147
    },
    {
      "epoch": 7.286048415390013,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 31148
    },
    {
      "epoch": 7.286282306163022,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6372,
      "step": 31149
    },
    {
      "epoch": 7.286516196936031,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 31150
    },
    {
      "epoch": 7.28675008770904,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6485,
      "step": 31151
    },
    {
      "epoch": 7.2869839784820485,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 31152
    },
    {
      "epoch": 7.287217869255058,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 31153
    },
    {
      "epoch": 7.287451760028067,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 31154
    },
    {
      "epoch": 7.287685650801076,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 31155
    },
    {
      "epoch": 7.287919541574085,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 31156
    },
    {
      "epoch": 7.288153432347094,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 31157
    },
    {
      "epoch": 7.288387323120103,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 31158
    },
    {
      "epoch": 7.288621213893112,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 31159
    },
    {
      "epoch": 7.288855104666121,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.641,
      "step": 31160
    },
    {
      "epoch": 7.28908899543913,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.4809,
      "step": 31161
    },
    {
      "epoch": 7.289322886212139,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 31162
    },
    {
      "epoch": 7.2895567769851475,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 31163
    },
    {
      "epoch": 7.289790667758157,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6622,
      "step": 31164
    },
    {
      "epoch": 7.290024558531166,
      "grad_norm": 11.0,
      "learning_rate": 3e-05,
      "loss": 2.4612,
      "step": 31165
    },
    {
      "epoch": 7.290258449304175,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.525,
      "step": 31166
    },
    {
      "epoch": 7.290492340077184,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7987,
      "step": 31167
    },
    {
      "epoch": 7.290726230850193,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.5345,
      "step": 31168
    },
    {
      "epoch": 7.290960121623202,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 31169
    },
    {
      "epoch": 7.291194012396211,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8029,
      "step": 31170
    },
    {
      "epoch": 7.29142790316922,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 31171
    },
    {
      "epoch": 7.291661793942229,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 31172
    },
    {
      "epoch": 7.291895684715238,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6876,
      "step": 31173
    },
    {
      "epoch": 7.292129575488247,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 31174
    },
    {
      "epoch": 7.292363466261256,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 31175
    },
    {
      "epoch": 7.292597357034265,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.7001,
      "step": 31176
    },
    {
      "epoch": 7.292831247807274,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 31177
    },
    {
      "epoch": 7.293065138580283,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 31178
    },
    {
      "epoch": 7.293299029353292,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 31179
    },
    {
      "epoch": 7.293532920126301,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 31180
    },
    {
      "epoch": 7.29376681089931,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 31181
    },
    {
      "epoch": 7.294000701672319,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 31182
    },
    {
      "epoch": 7.294234592445328,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 31183
    },
    {
      "epoch": 7.294468483218337,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 31184
    },
    {
      "epoch": 7.294702373991346,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6423,
      "step": 31185
    },
    {
      "epoch": 7.294936264764355,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 31186
    },
    {
      "epoch": 7.295170155537364,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6014,
      "step": 31187
    },
    {
      "epoch": 7.295404046310373,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 31188
    },
    {
      "epoch": 7.295637937083382,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 31189
    },
    {
      "epoch": 7.295871827856391,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 31190
    },
    {
      "epoch": 7.2961057186294,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5966,
      "step": 31191
    },
    {
      "epoch": 7.296339609402409,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 31192
    },
    {
      "epoch": 7.296573500175418,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.3071,
      "step": 31193
    },
    {
      "epoch": 7.296807390948427,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8289,
      "step": 31194
    },
    {
      "epoch": 7.297041281721436,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 31195
    },
    {
      "epoch": 7.297275172494445,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6672,
      "step": 31196
    },
    {
      "epoch": 7.297509063267454,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 31197
    },
    {
      "epoch": 7.297742954040463,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.758,
      "step": 31198
    },
    {
      "epoch": 7.297976844813472,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 31199
    },
    {
      "epoch": 7.298210735586481,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6506,
      "step": 31200
    },
    {
      "epoch": 7.298210735586481,
      "eval_runtime": 4.6379,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 31200
    },
    {
      "epoch": 7.29844462635949,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 31201
    },
    {
      "epoch": 7.298678517132499,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6975,
      "step": 31202
    },
    {
      "epoch": 7.298912407905508,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5025,
      "step": 31203
    },
    {
      "epoch": 7.299146298678517,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.377,
      "step": 31204
    },
    {
      "epoch": 7.299380189451526,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 31205
    },
    {
      "epoch": 7.2996140802245355,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 31206
    },
    {
      "epoch": 7.299847970997544,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 31207
    },
    {
      "epoch": 7.300081861770553,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 31208
    },
    {
      "epoch": 7.300315752543562,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 31209
    },
    {
      "epoch": 7.300549643316571,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6561,
      "step": 31210
    },
    {
      "epoch": 7.300783534089581,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5666,
      "step": 31211
    },
    {
      "epoch": 7.301017424862589,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 31212
    },
    {
      "epoch": 7.301251315635598,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.741,
      "step": 31213
    },
    {
      "epoch": 7.301485206408607,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 31214
    },
    {
      "epoch": 7.301719097181616,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 31215
    },
    {
      "epoch": 7.301952987954625,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 31216
    },
    {
      "epoch": 7.3021868787276345,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 31217
    },
    {
      "epoch": 7.302420769500643,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5052,
      "step": 31218
    },
    {
      "epoch": 7.302654660273652,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 31219
    },
    {
      "epoch": 7.302888551046661,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6015,
      "step": 31220
    },
    {
      "epoch": 7.30312244181967,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 31221
    },
    {
      "epoch": 7.30335633259268,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 31222
    },
    {
      "epoch": 7.303590223365688,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5659,
      "step": 31223
    },
    {
      "epoch": 7.303824114138697,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 31224
    },
    {
      "epoch": 7.304058004911706,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6812,
      "step": 31225
    },
    {
      "epoch": 7.304291895684715,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 31226
    },
    {
      "epoch": 7.304525786457725,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 31227
    },
    {
      "epoch": 7.3047596772307335,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 31228
    },
    {
      "epoch": 7.304993568003742,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 31229
    },
    {
      "epoch": 7.305227458776751,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 31230
    },
    {
      "epoch": 7.30546134954976,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 31231
    },
    {
      "epoch": 7.30569524032277,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 31232
    },
    {
      "epoch": 7.305929131095779,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 31233
    },
    {
      "epoch": 7.306163021868787,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 31234
    },
    {
      "epoch": 7.306396912641796,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 31235
    },
    {
      "epoch": 7.306630803414805,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 31236
    },
    {
      "epoch": 7.306864694187814,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 31237
    },
    {
      "epoch": 7.307098584960824,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 31238
    },
    {
      "epoch": 7.3073324757338325,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7083,
      "step": 31239
    },
    {
      "epoch": 7.307566366506841,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3428,
      "step": 31240
    },
    {
      "epoch": 7.30780025727985,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5744,
      "step": 31241
    },
    {
      "epoch": 7.308034148052859,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 31242
    },
    {
      "epoch": 7.308268038825869,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6546,
      "step": 31243
    },
    {
      "epoch": 7.308501929598878,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 31244
    },
    {
      "epoch": 7.308735820371886,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7008,
      "step": 31245
    },
    {
      "epoch": 7.308969711144895,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 31246
    },
    {
      "epoch": 7.309203601917904,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 31247
    },
    {
      "epoch": 7.309437492690913,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.553,
      "step": 31248
    },
    {
      "epoch": 7.309671383463923,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.4181,
      "step": 31249
    },
    {
      "epoch": 7.3099052742369315,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7232,
      "step": 31250
    },
    {
      "epoch": 7.31013916500994,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 31251
    },
    {
      "epoch": 7.310373055782949,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7426,
      "step": 31252
    },
    {
      "epoch": 7.310606946555958,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 31253
    },
    {
      "epoch": 7.310840837328968,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 31254
    },
    {
      "epoch": 7.311074728101977,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 31255
    },
    {
      "epoch": 7.311308618874985,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 31256
    },
    {
      "epoch": 7.311542509647994,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 31257
    },
    {
      "epoch": 7.311776400421003,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6978,
      "step": 31258
    },
    {
      "epoch": 7.312010291194013,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 31259
    },
    {
      "epoch": 7.312244181967022,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.4658,
      "step": 31260
    },
    {
      "epoch": 7.3124780727400305,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 31261
    },
    {
      "epoch": 7.312711963513039,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 31262
    },
    {
      "epoch": 7.312945854286048,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7946,
      "step": 31263
    },
    {
      "epoch": 7.313179745059058,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 31264
    },
    {
      "epoch": 7.313413635832067,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7771,
      "step": 31265
    },
    {
      "epoch": 7.3136475266050756,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.4975,
      "step": 31266
    },
    {
      "epoch": 7.313881417378084,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 31267
    },
    {
      "epoch": 7.314115308151093,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 31268
    },
    {
      "epoch": 7.314349198924102,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5733,
      "step": 31269
    },
    {
      "epoch": 7.314583089697112,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 31270
    },
    {
      "epoch": 7.314816980470121,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 31271
    },
    {
      "epoch": 7.3150508712431295,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5455,
      "step": 31272
    },
    {
      "epoch": 7.315284762016138,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 31273
    },
    {
      "epoch": 7.315518652789147,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6013,
      "step": 31274
    },
    {
      "epoch": 7.315752543562157,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 31275
    },
    {
      "epoch": 7.315986434335166,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 31276
    },
    {
      "epoch": 7.3162203251081745,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 31277
    },
    {
      "epoch": 7.316454215881183,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 31278
    },
    {
      "epoch": 7.316688106654192,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 31279
    },
    {
      "epoch": 7.316921997427201,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 31280
    },
    {
      "epoch": 7.317155888200211,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6108,
      "step": 31281
    },
    {
      "epoch": 7.31738977897322,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 31282
    },
    {
      "epoch": 7.3176236697462285,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 31283
    },
    {
      "epoch": 7.317857560519237,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 31284
    },
    {
      "epoch": 7.318091451292246,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 31285
    },
    {
      "epoch": 7.318325342065256,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 31286
    },
    {
      "epoch": 7.318559232838265,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 31287
    },
    {
      "epoch": 7.3187931236112735,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.595,
      "step": 31288
    },
    {
      "epoch": 7.319027014384282,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 31289
    },
    {
      "epoch": 7.319260905157291,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 31290
    },
    {
      "epoch": 7.319494795930301,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 31291
    },
    {
      "epoch": 7.31972868670331,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.6347,
      "step": 31292
    },
    {
      "epoch": 7.319962577476319,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 31293
    },
    {
      "epoch": 7.3201964682493275,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6393,
      "step": 31294
    },
    {
      "epoch": 7.320430359022336,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 31295
    },
    {
      "epoch": 7.320664249795346,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 31296
    },
    {
      "epoch": 7.320898140568355,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 31297
    },
    {
      "epoch": 7.321132031341364,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5432,
      "step": 31298
    },
    {
      "epoch": 7.3213659221143725,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5833,
      "step": 31299
    },
    {
      "epoch": 7.321599812887381,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 31300
    },
    {
      "epoch": 7.321599812887381,
      "eval_runtime": 4.6675,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 31300
    },
    {
      "epoch": 7.32183370366039,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 31301
    },
    {
      "epoch": 7.3220675944334,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 31302
    },
    {
      "epoch": 7.322301485206409,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7212,
      "step": 31303
    },
    {
      "epoch": 7.322535375979418,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.513,
      "step": 31304
    },
    {
      "epoch": 7.3227692667524265,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 31305
    },
    {
      "epoch": 7.323003157525435,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5358,
      "step": 31306
    },
    {
      "epoch": 7.323237048298445,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6501,
      "step": 31307
    },
    {
      "epoch": 7.323470939071454,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 31308
    },
    {
      "epoch": 7.323704829844463,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 31309
    },
    {
      "epoch": 7.3239387206174715,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6432,
      "step": 31310
    },
    {
      "epoch": 7.32417261139048,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5245,
      "step": 31311
    },
    {
      "epoch": 7.324406502163489,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 31312
    },
    {
      "epoch": 7.324640392936499,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 31313
    },
    {
      "epoch": 7.324874283709508,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6432,
      "step": 31314
    },
    {
      "epoch": 7.325108174482517,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 31315
    },
    {
      "epoch": 7.3253420652555254,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 31316
    },
    {
      "epoch": 7.325575956028534,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 31317
    },
    {
      "epoch": 7.325809846801544,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8315,
      "step": 31318
    },
    {
      "epoch": 7.326043737574553,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 31319
    },
    {
      "epoch": 7.326277628347562,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 31320
    },
    {
      "epoch": 7.3265115191205705,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 31321
    },
    {
      "epoch": 7.326745409893579,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 31322
    },
    {
      "epoch": 7.326979300666589,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 31323
    },
    {
      "epoch": 7.327213191439598,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4637,
      "step": 31324
    },
    {
      "epoch": 7.327447082212607,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6346,
      "step": 31325
    },
    {
      "epoch": 7.327680972985616,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 31326
    },
    {
      "epoch": 7.327914863758624,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 31327
    },
    {
      "epoch": 7.328148754531634,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 31328
    },
    {
      "epoch": 7.328382645304643,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 31329
    },
    {
      "epoch": 7.328616536077652,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 31330
    },
    {
      "epoch": 7.328850426850661,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7104,
      "step": 31331
    },
    {
      "epoch": 7.3290843176236695,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 31332
    },
    {
      "epoch": 7.329318208396678,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5931,
      "step": 31333
    },
    {
      "epoch": 7.329552099169688,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 31334
    },
    {
      "epoch": 7.329785989942697,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7463,
      "step": 31335
    },
    {
      "epoch": 7.330019880715706,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 31336
    },
    {
      "epoch": 7.330253771488715,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.565,
      "step": 31337
    },
    {
      "epoch": 7.330487662261723,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 31338
    },
    {
      "epoch": 7.330721553034733,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 31339
    },
    {
      "epoch": 7.330955443807742,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 31340
    },
    {
      "epoch": 7.331189334580751,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 31341
    },
    {
      "epoch": 7.33142322535376,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.662,
      "step": 31342
    },
    {
      "epoch": 7.3316571161267685,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 31343
    },
    {
      "epoch": 7.331891006899777,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 31344
    },
    {
      "epoch": 7.332124897672787,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 31345
    },
    {
      "epoch": 7.332358788445796,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8166,
      "step": 31346
    },
    {
      "epoch": 7.332592679218805,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.5221,
      "step": 31347
    },
    {
      "epoch": 7.332826569991814,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 31348
    },
    {
      "epoch": 7.333060460764822,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 31349
    },
    {
      "epoch": 7.333294351537832,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 31350
    },
    {
      "epoch": 7.333528242310841,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 31351
    },
    {
      "epoch": 7.33376213308385,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 31352
    },
    {
      "epoch": 7.333996023856859,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 31353
    },
    {
      "epoch": 7.3342299146298675,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.3994,
      "step": 31354
    },
    {
      "epoch": 7.334463805402877,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 31355
    },
    {
      "epoch": 7.334697696175886,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7622,
      "step": 31356
    },
    {
      "epoch": 7.334931586948895,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 31357
    },
    {
      "epoch": 7.335165477721904,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 31358
    },
    {
      "epoch": 7.335399368494913,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6098,
      "step": 31359
    },
    {
      "epoch": 7.335633259267922,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5145,
      "step": 31360
    },
    {
      "epoch": 7.335867150040931,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 31361
    },
    {
      "epoch": 7.33610104081394,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6307,
      "step": 31362
    },
    {
      "epoch": 7.336334931586949,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7134,
      "step": 31363
    },
    {
      "epoch": 7.336568822359958,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 31364
    },
    {
      "epoch": 7.3368027131329665,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6001,
      "step": 31365
    },
    {
      "epoch": 7.337036603905976,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.4108,
      "step": 31366
    },
    {
      "epoch": 7.337270494678985,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 31367
    },
    {
      "epoch": 7.337504385451994,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 31368
    },
    {
      "epoch": 7.337738276225003,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 31369
    },
    {
      "epoch": 7.337972166998012,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6437,
      "step": 31370
    },
    {
      "epoch": 7.338206057771021,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 31371
    },
    {
      "epoch": 7.33843994854403,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5451,
      "step": 31372
    },
    {
      "epoch": 7.338673839317039,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 31373
    },
    {
      "epoch": 7.338907730090048,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5074,
      "step": 31374
    },
    {
      "epoch": 7.339141620863057,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5881,
      "step": 31375
    },
    {
      "epoch": 7.3393755116360655,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 31376
    },
    {
      "epoch": 7.339609402409075,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6616,
      "step": 31377
    },
    {
      "epoch": 7.339843293182084,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 31378
    },
    {
      "epoch": 7.340077183955093,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6191,
      "step": 31379
    },
    {
      "epoch": 7.340311074728102,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 31380
    },
    {
      "epoch": 7.340544965501111,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 31381
    },
    {
      "epoch": 7.34077885627412,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 31382
    },
    {
      "epoch": 7.341012747047129,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 31383
    },
    {
      "epoch": 7.341246637820138,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 31384
    },
    {
      "epoch": 7.341480528593147,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6785,
      "step": 31385
    },
    {
      "epoch": 7.341714419366156,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6957,
      "step": 31386
    },
    {
      "epoch": 7.341948310139165,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 31387
    },
    {
      "epoch": 7.342182200912174,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 31388
    },
    {
      "epoch": 7.342416091685183,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 31389
    },
    {
      "epoch": 7.342649982458192,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 31390
    },
    {
      "epoch": 7.342883873231201,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6135,
      "step": 31391
    },
    {
      "epoch": 7.3431177640042105,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 31392
    },
    {
      "epoch": 7.343351654777219,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7701,
      "step": 31393
    },
    {
      "epoch": 7.343585545550228,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7261,
      "step": 31394
    },
    {
      "epoch": 7.343819436323237,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 31395
    },
    {
      "epoch": 7.344053327096246,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6289,
      "step": 31396
    },
    {
      "epoch": 7.344287217869255,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7206,
      "step": 31397
    },
    {
      "epoch": 7.344521108642264,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8296,
      "step": 31398
    },
    {
      "epoch": 7.344754999415273,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.659,
      "step": 31399
    },
    {
      "epoch": 7.344988890188282,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 31400
    },
    {
      "epoch": 7.344988890188282,
      "eval_runtime": 4.6466,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 31400
    },
    {
      "epoch": 7.345222780961291,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 31401
    },
    {
      "epoch": 7.3454566717343,
      "grad_norm": 7.3125,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 31402
    },
    {
      "epoch": 7.3456905625073095,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.509,
      "step": 31403
    },
    {
      "epoch": 7.345924453280318,
      "grad_norm": 2.515625,
      "learning_rate": 3e-05,
      "loss": 1.3886,
      "step": 31404
    },
    {
      "epoch": 7.346158344053327,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 31405
    },
    {
      "epoch": 7.346392234826336,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7472,
      "step": 31406
    },
    {
      "epoch": 7.346626125599345,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 31407
    },
    {
      "epoch": 7.346860016372354,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.6683,
      "step": 31408
    },
    {
      "epoch": 7.347093907145363,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.5219,
      "step": 31409
    },
    {
      "epoch": 7.347327797918372,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 31410
    },
    {
      "epoch": 7.347561688691381,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6947,
      "step": 31411
    },
    {
      "epoch": 7.34779557946439,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 31412
    },
    {
      "epoch": 7.348029470237399,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7643,
      "step": 31413
    },
    {
      "epoch": 7.3482633610104084,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 31414
    },
    {
      "epoch": 7.348497251783417,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 31415
    },
    {
      "epoch": 7.348731142556426,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 31416
    },
    {
      "epoch": 7.348965033329435,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 31417
    },
    {
      "epoch": 7.349198924102444,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6109,
      "step": 31418
    },
    {
      "epoch": 7.3494328148754535,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5296,
      "step": 31419
    },
    {
      "epoch": 7.349666705648462,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7448,
      "step": 31420
    },
    {
      "epoch": 7.349900596421471,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 31421
    },
    {
      "epoch": 7.35013448719448,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5836,
      "step": 31422
    },
    {
      "epoch": 7.350368377967489,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5866,
      "step": 31423
    },
    {
      "epoch": 7.350602268740499,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 31424
    },
    {
      "epoch": 7.350836159513507,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 31425
    },
    {
      "epoch": 7.351070050286516,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 31426
    },
    {
      "epoch": 7.351303941059525,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 31427
    },
    {
      "epoch": 7.351537831832534,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 31428
    },
    {
      "epoch": 7.351771722605543,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6625,
      "step": 31429
    },
    {
      "epoch": 7.3520056133785525,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 31430
    },
    {
      "epoch": 7.352239504151561,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 31431
    },
    {
      "epoch": 7.35247339492457,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 31432
    },
    {
      "epoch": 7.352707285697579,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 31433
    },
    {
      "epoch": 7.352941176470588,
      "grad_norm": 7.46875,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 31434
    },
    {
      "epoch": 7.353175067243598,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 31435
    },
    {
      "epoch": 7.353408958016606,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 31436
    },
    {
      "epoch": 7.353642848789615,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 31437
    },
    {
      "epoch": 7.353876739562624,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7923,
      "step": 31438
    },
    {
      "epoch": 7.354110630335633,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 31439
    },
    {
      "epoch": 7.354344521108643,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.583,
      "step": 31440
    },
    {
      "epoch": 7.3545784118816515,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7185,
      "step": 31441
    },
    {
      "epoch": 7.35481230265466,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 31442
    },
    {
      "epoch": 7.355046193427669,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5645,
      "step": 31443
    },
    {
      "epoch": 7.355280084200678,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4483,
      "step": 31444
    },
    {
      "epoch": 7.355513974973688,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4709,
      "step": 31445
    },
    {
      "epoch": 7.355747865746697,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 31446
    },
    {
      "epoch": 7.355981756519705,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 31447
    },
    {
      "epoch": 7.356215647292714,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 31448
    },
    {
      "epoch": 7.356449538065723,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6081,
      "step": 31449
    },
    {
      "epoch": 7.356683428838732,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 31450
    },
    {
      "epoch": 7.356917319611742,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 31451
    },
    {
      "epoch": 7.3571512103847505,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 31452
    },
    {
      "epoch": 7.357385101157759,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7619,
      "step": 31453
    },
    {
      "epoch": 7.357618991930768,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8503,
      "step": 31454
    },
    {
      "epoch": 7.357852882703777,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6786,
      "step": 31455
    },
    {
      "epoch": 7.358086773476787,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 31456
    },
    {
      "epoch": 7.358320664249796,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 31457
    },
    {
      "epoch": 7.358554555022804,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.3141,
      "step": 31458
    },
    {
      "epoch": 7.358788445795813,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6871,
      "step": 31459
    },
    {
      "epoch": 7.359022336568822,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 31460
    },
    {
      "epoch": 7.359256227341831,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 31461
    },
    {
      "epoch": 7.359490118114841,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5203,
      "step": 31462
    },
    {
      "epoch": 7.3597240088878495,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 31463
    },
    {
      "epoch": 7.359957899660858,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6378,
      "step": 31464
    },
    {
      "epoch": 7.360191790433867,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4513,
      "step": 31465
    },
    {
      "epoch": 7.360425681206876,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5452,
      "step": 31466
    },
    {
      "epoch": 7.360659571979886,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 31467
    },
    {
      "epoch": 7.360893462752895,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 31468
    },
    {
      "epoch": 7.361127353525903,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5118,
      "step": 31469
    },
    {
      "epoch": 7.361361244298912,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 31470
    },
    {
      "epoch": 7.361595135071921,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 31471
    },
    {
      "epoch": 7.361829025844931,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4746,
      "step": 31472
    },
    {
      "epoch": 7.36206291661794,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 31473
    },
    {
      "epoch": 7.3622968073909485,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2274,
      "step": 31474
    },
    {
      "epoch": 7.362530698163957,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 31475
    },
    {
      "epoch": 7.362764588936966,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 31476
    },
    {
      "epoch": 7.362998479709976,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 31477
    },
    {
      "epoch": 7.363232370482985,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6734,
      "step": 31478
    },
    {
      "epoch": 7.363466261255994,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 31479
    },
    {
      "epoch": 7.363700152029002,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 31480
    },
    {
      "epoch": 7.363934042802011,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6951,
      "step": 31481
    },
    {
      "epoch": 7.36416793357502,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6133,
      "step": 31482
    },
    {
      "epoch": 7.36440182434803,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6218,
      "step": 31483
    },
    {
      "epoch": 7.364635715121039,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 31484
    },
    {
      "epoch": 7.3648696058940475,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6871,
      "step": 31485
    },
    {
      "epoch": 7.365103496667056,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 31486
    },
    {
      "epoch": 7.365337387440065,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.611,
      "step": 31487
    },
    {
      "epoch": 7.365571278213075,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 31488
    },
    {
      "epoch": 7.365805168986084,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1445,
      "step": 31489
    },
    {
      "epoch": 7.366039059759093,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5154,
      "step": 31490
    },
    {
      "epoch": 7.366272950532101,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 31491
    },
    {
      "epoch": 7.36650684130511,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 31492
    },
    {
      "epoch": 7.366740732078119,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5715,
      "step": 31493
    },
    {
      "epoch": 7.366974622851129,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 31494
    },
    {
      "epoch": 7.367208513624138,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.4339,
      "step": 31495
    },
    {
      "epoch": 7.3674424043971465,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 31496
    },
    {
      "epoch": 7.367676295170155,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 31497
    },
    {
      "epoch": 7.367910185943164,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6184,
      "step": 31498
    },
    {
      "epoch": 7.368144076716174,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 31499
    },
    {
      "epoch": 7.368377967489183,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 31500
    },
    {
      "epoch": 7.368377967489183,
      "eval_runtime": 4.6103,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 31500
    },
    {
      "epoch": 7.368611858262192,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 31501
    },
    {
      "epoch": 7.3688457490352,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 31502
    },
    {
      "epoch": 7.369079639808209,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6307,
      "step": 31503
    },
    {
      "epoch": 7.369313530581219,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7897,
      "step": 31504
    },
    {
      "epoch": 7.369547421354228,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 31505
    },
    {
      "epoch": 7.369781312127237,
      "grad_norm": 7.28125,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 31506
    },
    {
      "epoch": 7.3700152029002455,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 31507
    },
    {
      "epoch": 7.370249093673254,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 31508
    },
    {
      "epoch": 7.370482984446264,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 31509
    },
    {
      "epoch": 7.370716875219273,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 31510
    },
    {
      "epoch": 7.370950765992282,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7201,
      "step": 31511
    },
    {
      "epoch": 7.3711846567652906,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7085,
      "step": 31512
    },
    {
      "epoch": 7.371418547538299,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 31513
    },
    {
      "epoch": 7.371652438311308,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 31514
    },
    {
      "epoch": 7.371886329084318,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2511,
      "step": 31515
    },
    {
      "epoch": 7.372120219857327,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 31516
    },
    {
      "epoch": 7.372354110630336,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 31517
    },
    {
      "epoch": 7.3725880014033445,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 31518
    },
    {
      "epoch": 7.372821892176353,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 31519
    },
    {
      "epoch": 7.373055782949363,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 31520
    },
    {
      "epoch": 7.373289673722372,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 31521
    },
    {
      "epoch": 7.373523564495381,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7858,
      "step": 31522
    },
    {
      "epoch": 7.3737574552683895,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6118,
      "step": 31523
    },
    {
      "epoch": 7.373991346041398,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 31524
    },
    {
      "epoch": 7.374225236814407,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 31525
    },
    {
      "epoch": 7.374459127587417,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7073,
      "step": 31526
    },
    {
      "epoch": 7.374693018360426,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 31527
    },
    {
      "epoch": 7.374926909133435,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 31528
    },
    {
      "epoch": 7.3751607999064435,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 31529
    },
    {
      "epoch": 7.375394690679452,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 31530
    },
    {
      "epoch": 7.375628581452462,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 31531
    },
    {
      "epoch": 7.375862472225471,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6189,
      "step": 31532
    },
    {
      "epoch": 7.37609636299848,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6852,
      "step": 31533
    },
    {
      "epoch": 7.3763302537714885,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 31534
    },
    {
      "epoch": 7.376564144544497,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 31535
    },
    {
      "epoch": 7.376798035317507,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 31536
    },
    {
      "epoch": 7.377031926090516,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 31537
    },
    {
      "epoch": 7.377265816863525,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6452,
      "step": 31538
    },
    {
      "epoch": 7.377499707636534,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7868,
      "step": 31539
    },
    {
      "epoch": 7.3777335984095425,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7437,
      "step": 31540
    },
    {
      "epoch": 7.377967489182552,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.54,
      "step": 31541
    },
    {
      "epoch": 7.378201379955561,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6732,
      "step": 31542
    },
    {
      "epoch": 7.37843527072857,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 31543
    },
    {
      "epoch": 7.378669161501579,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 31544
    },
    {
      "epoch": 7.3789030522745875,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5543,
      "step": 31545
    },
    {
      "epoch": 7.379136943047596,
      "grad_norm": 7.3125,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 31546
    },
    {
      "epoch": 7.379370833820606,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8187,
      "step": 31547
    },
    {
      "epoch": 7.379604724593615,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 31548
    },
    {
      "epoch": 7.379838615366624,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5558,
      "step": 31549
    },
    {
      "epoch": 7.380072506139633,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 31550
    },
    {
      "epoch": 7.3803063969126415,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 31551
    },
    {
      "epoch": 7.380540287685651,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 31552
    },
    {
      "epoch": 7.38077417845866,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 31553
    },
    {
      "epoch": 7.381008069231669,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 31554
    },
    {
      "epoch": 7.381241960004678,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 31555
    },
    {
      "epoch": 7.3814758507776865,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5342,
      "step": 31556
    },
    {
      "epoch": 7.381709741550695,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 31557
    },
    {
      "epoch": 7.381943632323705,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 31558
    },
    {
      "epoch": 7.382177523096714,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7524,
      "step": 31559
    },
    {
      "epoch": 7.382411413869723,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7239,
      "step": 31560
    },
    {
      "epoch": 7.382645304642732,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6004,
      "step": 31561
    },
    {
      "epoch": 7.3828791954157404,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 31562
    },
    {
      "epoch": 7.38311308618875,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 31563
    },
    {
      "epoch": 7.383346976961759,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6453,
      "step": 31564
    },
    {
      "epoch": 7.383580867734768,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7543,
      "step": 31565
    },
    {
      "epoch": 7.383814758507777,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 31566
    },
    {
      "epoch": 7.3840486492807855,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 31567
    },
    {
      "epoch": 7.384282540053795,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 31568
    },
    {
      "epoch": 7.384516430826804,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.827,
      "step": 31569
    },
    {
      "epoch": 7.384750321599813,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 31570
    },
    {
      "epoch": 7.384984212372822,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 31571
    },
    {
      "epoch": 7.385218103145831,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 31572
    },
    {
      "epoch": 7.38545199391884,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6559,
      "step": 31573
    },
    {
      "epoch": 7.385685884691849,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 31574
    },
    {
      "epoch": 7.385919775464858,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 31575
    },
    {
      "epoch": 7.386153666237867,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7117,
      "step": 31576
    },
    {
      "epoch": 7.386387557010876,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 31577
    },
    {
      "epoch": 7.3866214477838845,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 31578
    },
    {
      "epoch": 7.386855338556894,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 31579
    },
    {
      "epoch": 7.387089229329903,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.2902,
      "step": 31580
    },
    {
      "epoch": 7.387323120102912,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 31581
    },
    {
      "epoch": 7.387557010875921,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5379,
      "step": 31582
    },
    {
      "epoch": 7.38779090164893,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.532,
      "step": 31583
    },
    {
      "epoch": 7.388024792421939,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6274,
      "step": 31584
    },
    {
      "epoch": 7.388258683194948,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 31585
    },
    {
      "epoch": 7.388492573967957,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5226,
      "step": 31586
    },
    {
      "epoch": 7.388726464740966,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 31587
    },
    {
      "epoch": 7.388960355513975,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 31588
    },
    {
      "epoch": 7.3891942462869835,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 31589
    },
    {
      "epoch": 7.389428137059993,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8434,
      "step": 31590
    },
    {
      "epoch": 7.389662027833002,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 31591
    },
    {
      "epoch": 7.389895918606011,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5055,
      "step": 31592
    },
    {
      "epoch": 7.39012980937902,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 31593
    },
    {
      "epoch": 7.390363700152029,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5692,
      "step": 31594
    },
    {
      "epoch": 7.390597590925038,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.4928,
      "step": 31595
    },
    {
      "epoch": 7.390831481698047,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 31596
    },
    {
      "epoch": 7.391065372471056,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.5591,
      "step": 31597
    },
    {
      "epoch": 7.391299263244065,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6552,
      "step": 31598
    },
    {
      "epoch": 7.391533154017074,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.3035,
      "step": 31599
    },
    {
      "epoch": 7.391767044790083,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 31600
    },
    {
      "epoch": 7.391767044790083,
      "eval_runtime": 4.684,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 31600
    },
    {
      "epoch": 7.392000935563092,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5974,
      "step": 31601
    },
    {
      "epoch": 7.392234826336101,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 31602
    },
    {
      "epoch": 7.39246871710911,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 31603
    },
    {
      "epoch": 7.392702607882119,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5751,
      "step": 31604
    },
    {
      "epoch": 7.3929364986551285,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6581,
      "step": 31605
    },
    {
      "epoch": 7.393170389428137,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7073,
      "step": 31606
    },
    {
      "epoch": 7.393404280201146,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 31607
    },
    {
      "epoch": 7.393638170974155,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 31608
    },
    {
      "epoch": 7.393872061747164,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 31609
    },
    {
      "epoch": 7.394105952520173,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 31610
    },
    {
      "epoch": 7.394339843293182,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 31611
    },
    {
      "epoch": 7.394573734066191,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 31612
    },
    {
      "epoch": 7.3948076248392,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 31613
    },
    {
      "epoch": 7.395041515612209,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 31614
    },
    {
      "epoch": 7.395275406385218,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 31615
    },
    {
      "epoch": 7.3955092971582275,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6592,
      "step": 31616
    },
    {
      "epoch": 7.395743187931236,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 31617
    },
    {
      "epoch": 7.395977078704245,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6504,
      "step": 31618
    },
    {
      "epoch": 7.396210969477254,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6102,
      "step": 31619
    },
    {
      "epoch": 7.396444860250263,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6338,
      "step": 31620
    },
    {
      "epoch": 7.3966787510232725,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 31621
    },
    {
      "epoch": 7.396912641796281,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 31622
    },
    {
      "epoch": 7.39714653256929,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6881,
      "step": 31623
    },
    {
      "epoch": 7.397380423342299,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 31624
    },
    {
      "epoch": 7.397614314115308,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 31625
    },
    {
      "epoch": 7.397848204888318,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 31626
    },
    {
      "epoch": 7.3980820956613265,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 31627
    },
    {
      "epoch": 7.398315986434335,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7152,
      "step": 31628
    },
    {
      "epoch": 7.398549877207344,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 31629
    },
    {
      "epoch": 7.398783767980353,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5295,
      "step": 31630
    },
    {
      "epoch": 7.399017658753362,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 31631
    },
    {
      "epoch": 7.3992515495263715,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6949,
      "step": 31632
    },
    {
      "epoch": 7.39948544029938,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 31633
    },
    {
      "epoch": 7.399719331072389,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 31634
    },
    {
      "epoch": 7.399953221845398,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 31635
    },
    {
      "epoch": 7.400187112618407,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 31636
    },
    {
      "epoch": 7.400421003391417,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6739,
      "step": 31637
    },
    {
      "epoch": 7.4006548941644255,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 31638
    },
    {
      "epoch": 7.400888784937434,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 31639
    },
    {
      "epoch": 7.401122675710443,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5887,
      "step": 31640
    },
    {
      "epoch": 7.401356566483452,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 1.5598,
      "step": 31641
    },
    {
      "epoch": 7.401590457256461,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 31642
    },
    {
      "epoch": 7.4018243480294705,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 31643
    },
    {
      "epoch": 7.402058238802479,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 31644
    },
    {
      "epoch": 7.402292129575488,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 31645
    },
    {
      "epoch": 7.402526020348497,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8025,
      "step": 31646
    },
    {
      "epoch": 7.402759911121506,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 31647
    },
    {
      "epoch": 7.402993801894516,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 31648
    },
    {
      "epoch": 7.4032276926675245,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 31649
    },
    {
      "epoch": 7.403461583440533,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 31650
    },
    {
      "epoch": 7.403695474213542,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.694,
      "step": 31651
    },
    {
      "epoch": 7.403929364986551,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 31652
    },
    {
      "epoch": 7.404163255759561,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 31653
    },
    {
      "epoch": 7.4043971465325695,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 31654
    },
    {
      "epoch": 7.404631037305578,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 31655
    },
    {
      "epoch": 7.404864928078587,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 31656
    },
    {
      "epoch": 7.405098818851596,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 31657
    },
    {
      "epoch": 7.405332709624606,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 31658
    },
    {
      "epoch": 7.405566600397615,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 31659
    },
    {
      "epoch": 7.4058004911706234,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 31660
    },
    {
      "epoch": 7.406034381943632,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.525,
      "step": 31661
    },
    {
      "epoch": 7.406268272716641,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8501,
      "step": 31662
    },
    {
      "epoch": 7.40650216348965,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6901,
      "step": 31663
    },
    {
      "epoch": 7.40673605426266,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 31664
    },
    {
      "epoch": 7.4069699450356685,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8369,
      "step": 31665
    },
    {
      "epoch": 7.407203835808677,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 31666
    },
    {
      "epoch": 7.407437726581686,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7331,
      "step": 31667
    },
    {
      "epoch": 7.407671617354695,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 31668
    },
    {
      "epoch": 7.407905508127705,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 31669
    },
    {
      "epoch": 7.408139398900714,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 31670
    },
    {
      "epoch": 7.408373289673722,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 31671
    },
    {
      "epoch": 7.408607180446731,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 31672
    },
    {
      "epoch": 7.40884107121974,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 31673
    },
    {
      "epoch": 7.409074961992749,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.6655,
      "step": 31674
    },
    {
      "epoch": 7.409308852765759,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6852,
      "step": 31675
    },
    {
      "epoch": 7.4095427435387675,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 31676
    },
    {
      "epoch": 7.409776634311776,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 31677
    },
    {
      "epoch": 7.410010525084785,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 31678
    },
    {
      "epoch": 7.410244415857794,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 31679
    },
    {
      "epoch": 7.410478306630804,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 31680
    },
    {
      "epoch": 7.410712197403813,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 31681
    },
    {
      "epoch": 7.410946088176821,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7005,
      "step": 31682
    },
    {
      "epoch": 7.41117997894983,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 31683
    },
    {
      "epoch": 7.411413869722839,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6668,
      "step": 31684
    },
    {
      "epoch": 7.411647760495849,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7663,
      "step": 31685
    },
    {
      "epoch": 7.411881651268858,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 31686
    },
    {
      "epoch": 7.4121155420418665,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 31687
    },
    {
      "epoch": 7.412349432814875,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6586,
      "step": 31688
    },
    {
      "epoch": 7.412583323587884,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 31689
    },
    {
      "epoch": 7.412817214360894,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 31690
    },
    {
      "epoch": 7.413051105133903,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 31691
    },
    {
      "epoch": 7.413284995906912,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 31692
    },
    {
      "epoch": 7.41351888667992,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 31693
    },
    {
      "epoch": 7.413752777452929,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8714,
      "step": 31694
    },
    {
      "epoch": 7.413986668225938,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 31695
    },
    {
      "epoch": 7.414220558998948,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6416,
      "step": 31696
    },
    {
      "epoch": 7.414454449771957,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 31697
    },
    {
      "epoch": 7.4146883405449655,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 31698
    },
    {
      "epoch": 7.414922231317974,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7242,
      "step": 31699
    },
    {
      "epoch": 7.415156122090983,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7191,
      "step": 31700
    },
    {
      "epoch": 7.415156122090983,
      "eval_runtime": 4.5992,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 31700
    },
    {
      "epoch": 7.415390012863993,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 31701
    },
    {
      "epoch": 7.415623903637002,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4082,
      "step": 31702
    },
    {
      "epoch": 7.415857794410011,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 31703
    },
    {
      "epoch": 7.416091685183019,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 31704
    },
    {
      "epoch": 7.416325575956028,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6561,
      "step": 31705
    },
    {
      "epoch": 7.416559466729037,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 31706
    },
    {
      "epoch": 7.416793357502047,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 31707
    },
    {
      "epoch": 7.417027248275056,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 31708
    },
    {
      "epoch": 7.4172611390480645,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5944,
      "step": 31709
    },
    {
      "epoch": 7.417495029821073,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6764,
      "step": 31710
    },
    {
      "epoch": 7.417728920594082,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 31711
    },
    {
      "epoch": 7.417962811367092,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6502,
      "step": 31712
    },
    {
      "epoch": 7.418196702140101,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 31713
    },
    {
      "epoch": 7.41843059291311,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 31714
    },
    {
      "epoch": 7.418664483686118,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 31715
    },
    {
      "epoch": 7.418898374459127,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7305,
      "step": 31716
    },
    {
      "epoch": 7.419132265232137,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 31717
    },
    {
      "epoch": 7.419366156005146,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7317,
      "step": 31718
    },
    {
      "epoch": 7.419600046778155,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 31719
    },
    {
      "epoch": 7.4198339375511635,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 31720
    },
    {
      "epoch": 7.420067828324172,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 31721
    },
    {
      "epoch": 7.420301719097182,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 31722
    },
    {
      "epoch": 7.420535609870191,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4421,
      "step": 31723
    },
    {
      "epoch": 7.4207695006432,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 31724
    },
    {
      "epoch": 7.421003391416209,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6578,
      "step": 31725
    },
    {
      "epoch": 7.421237282189217,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7653,
      "step": 31726
    },
    {
      "epoch": 7.421471172962226,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5872,
      "step": 31727
    },
    {
      "epoch": 7.421705063735236,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 31728
    },
    {
      "epoch": 7.421938954508245,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7283,
      "step": 31729
    },
    {
      "epoch": 7.422172845281254,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 31730
    },
    {
      "epoch": 7.4224067360542625,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5683,
      "step": 31731
    },
    {
      "epoch": 7.422640626827271,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6667,
      "step": 31732
    },
    {
      "epoch": 7.422874517600281,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.3451,
      "step": 31733
    },
    {
      "epoch": 7.42310840837329,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6732,
      "step": 31734
    },
    {
      "epoch": 7.423342299146299,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2855,
      "step": 31735
    },
    {
      "epoch": 7.423576189919308,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4609,
      "step": 31736
    },
    {
      "epoch": 7.423810080692316,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8093,
      "step": 31737
    },
    {
      "epoch": 7.424043971465325,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 31738
    },
    {
      "epoch": 7.424277862238335,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7048,
      "step": 31739
    },
    {
      "epoch": 7.424511753011344,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5298,
      "step": 31740
    },
    {
      "epoch": 7.424745643784353,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 31741
    },
    {
      "epoch": 7.4249795345573615,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 31742
    },
    {
      "epoch": 7.42521342533037,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 31743
    },
    {
      "epoch": 7.42544731610338,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 31744
    },
    {
      "epoch": 7.425681206876389,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.688,
      "step": 31745
    },
    {
      "epoch": 7.425915097649398,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5742,
      "step": 31746
    },
    {
      "epoch": 7.426148988422407,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 31747
    },
    {
      "epoch": 7.426382879195415,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 31748
    },
    {
      "epoch": 7.426616769968425,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 31749
    },
    {
      "epoch": 7.426850660741434,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 31750
    },
    {
      "epoch": 7.427084551514443,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 31751
    },
    {
      "epoch": 7.427318442287452,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 31752
    },
    {
      "epoch": 7.4275523330604605,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 31753
    },
    {
      "epoch": 7.42778622383347,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 31754
    },
    {
      "epoch": 7.428020114606479,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.4502,
      "step": 31755
    },
    {
      "epoch": 7.428254005379488,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6195,
      "step": 31756
    },
    {
      "epoch": 7.428487896152497,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6415,
      "step": 31757
    },
    {
      "epoch": 7.4287217869255056,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6647,
      "step": 31758
    },
    {
      "epoch": 7.428955677698514,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 31759
    },
    {
      "epoch": 7.429189568471524,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 31760
    },
    {
      "epoch": 7.429423459244533,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 31761
    },
    {
      "epoch": 7.429657350017542,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 31762
    },
    {
      "epoch": 7.429891240790551,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 31763
    },
    {
      "epoch": 7.4301251315635595,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 31764
    },
    {
      "epoch": 7.430359022336569,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7511,
      "step": 31765
    },
    {
      "epoch": 7.430592913109578,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 31766
    },
    {
      "epoch": 7.430826803882587,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2823,
      "step": 31767
    },
    {
      "epoch": 7.431060694655596,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3677,
      "step": 31768
    },
    {
      "epoch": 7.4312945854286045,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 31769
    },
    {
      "epoch": 7.431528476201613,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6646,
      "step": 31770
    },
    {
      "epoch": 7.431762366974623,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 31771
    },
    {
      "epoch": 7.431996257747632,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 31772
    },
    {
      "epoch": 7.432230148520641,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6029,
      "step": 31773
    },
    {
      "epoch": 7.43246403929365,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7202,
      "step": 31774
    },
    {
      "epoch": 7.4326979300666585,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 31775
    },
    {
      "epoch": 7.432931820839668,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.705,
      "step": 31776
    },
    {
      "epoch": 7.433165711612677,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7141,
      "step": 31777
    },
    {
      "epoch": 7.433399602385686,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4801,
      "step": 31778
    },
    {
      "epoch": 7.433633493158695,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 31779
    },
    {
      "epoch": 7.4338673839317035,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 31780
    },
    {
      "epoch": 7.434101274704713,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7165,
      "step": 31781
    },
    {
      "epoch": 7.434335165477722,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.595,
      "step": 31782
    },
    {
      "epoch": 7.434569056250731,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 31783
    },
    {
      "epoch": 7.43480294702374,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 31784
    },
    {
      "epoch": 7.435036837796749,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 31785
    },
    {
      "epoch": 7.435270728569758,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 31786
    },
    {
      "epoch": 7.435504619342767,
      "grad_norm": 7.9375,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 31787
    },
    {
      "epoch": 7.435738510115776,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6761,
      "step": 31788
    },
    {
      "epoch": 7.435972400888785,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 31789
    },
    {
      "epoch": 7.436206291661794,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.575,
      "step": 31790
    },
    {
      "epoch": 7.4364401824348025,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7045,
      "step": 31791
    },
    {
      "epoch": 7.436674073207812,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8108,
      "step": 31792
    },
    {
      "epoch": 7.436907963980821,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 31793
    },
    {
      "epoch": 7.43714185475383,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.573,
      "step": 31794
    },
    {
      "epoch": 7.437375745526839,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 31795
    },
    {
      "epoch": 7.437609636299848,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5372,
      "step": 31796
    },
    {
      "epoch": 7.437843527072857,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5468,
      "step": 31797
    },
    {
      "epoch": 7.438077417845866,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 31798
    },
    {
      "epoch": 7.438311308618875,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6958,
      "step": 31799
    },
    {
      "epoch": 7.438545199391884,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5114,
      "step": 31800
    },
    {
      "epoch": 7.438545199391884,
      "eval_runtime": 4.5836,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 31800
    },
    {
      "epoch": 7.438779090164893,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 31801
    },
    {
      "epoch": 7.4390129809379015,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0967,
      "step": 31802
    },
    {
      "epoch": 7.439246871710911,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6132,
      "step": 31803
    },
    {
      "epoch": 7.43948076248392,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 31804
    },
    {
      "epoch": 7.439714653256929,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6,
      "step": 31805
    },
    {
      "epoch": 7.439948544029938,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6502,
      "step": 31806
    },
    {
      "epoch": 7.440182434802947,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5296,
      "step": 31807
    },
    {
      "epoch": 7.440416325575956,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6819,
      "step": 31808
    },
    {
      "epoch": 7.440650216348965,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 31809
    },
    {
      "epoch": 7.440884107121974,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 31810
    },
    {
      "epoch": 7.441117997894983,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 31811
    },
    {
      "epoch": 7.441351888667992,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 31812
    },
    {
      "epoch": 7.441585779441001,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 31813
    },
    {
      "epoch": 7.44181967021401,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9937,
      "step": 31814
    },
    {
      "epoch": 7.442053560987019,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 31815
    },
    {
      "epoch": 7.442287451760028,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1757,
      "step": 31816
    },
    {
      "epoch": 7.442521342533037,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 31817
    },
    {
      "epoch": 7.4427552333060465,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4643,
      "step": 31818
    },
    {
      "epoch": 7.442989124079055,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 31819
    },
    {
      "epoch": 7.443223014852064,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6517,
      "step": 31820
    },
    {
      "epoch": 7.443456905625073,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7868,
      "step": 31821
    },
    {
      "epoch": 7.443690796398082,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 31822
    },
    {
      "epoch": 7.443924687171091,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 31823
    },
    {
      "epoch": 7.4441585779441,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 31824
    },
    {
      "epoch": 7.444392468717109,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 31825
    },
    {
      "epoch": 7.444626359490118,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 31826
    },
    {
      "epoch": 7.444860250263127,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.597,
      "step": 31827
    },
    {
      "epoch": 7.445094141036136,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 31828
    },
    {
      "epoch": 7.4453280318091455,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 31829
    },
    {
      "epoch": 7.445561922582154,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 31830
    },
    {
      "epoch": 7.445795813355163,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6516,
      "step": 31831
    },
    {
      "epoch": 7.446029704128172,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7283,
      "step": 31832
    },
    {
      "epoch": 7.446263594901181,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6928,
      "step": 31833
    },
    {
      "epoch": 7.446497485674191,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6847,
      "step": 31834
    },
    {
      "epoch": 7.446731376447199,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6118,
      "step": 31835
    },
    {
      "epoch": 7.446965267220208,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 31836
    },
    {
      "epoch": 7.447199157993217,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7155,
      "step": 31837
    },
    {
      "epoch": 7.447433048766226,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6314,
      "step": 31838
    },
    {
      "epoch": 7.447666939539236,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 31839
    },
    {
      "epoch": 7.4479008303122445,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 31840
    },
    {
      "epoch": 7.448134721085253,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 31841
    },
    {
      "epoch": 7.448368611858262,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 31842
    },
    {
      "epoch": 7.448602502631271,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6177,
      "step": 31843
    },
    {
      "epoch": 7.44883639340428,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 31844
    },
    {
      "epoch": 7.44907028417729,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 31845
    },
    {
      "epoch": 7.449304174950298,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 31846
    },
    {
      "epoch": 7.449538065723307,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5588,
      "step": 31847
    },
    {
      "epoch": 7.449771956496316,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 31848
    },
    {
      "epoch": 7.450005847269325,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 31849
    },
    {
      "epoch": 7.450239738042335,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 31850
    },
    {
      "epoch": 7.4504736288153435,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6723,
      "step": 31851
    },
    {
      "epoch": 7.450707519588352,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.561,
      "step": 31852
    },
    {
      "epoch": 7.450941410361361,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 31853
    },
    {
      "epoch": 7.45117530113437,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.656,
      "step": 31854
    },
    {
      "epoch": 7.451409191907379,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.3529,
      "step": 31855
    },
    {
      "epoch": 7.4516430826803886,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 31856
    },
    {
      "epoch": 7.451876973453397,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 31857
    },
    {
      "epoch": 7.452110864226406,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 31858
    },
    {
      "epoch": 7.452344754999415,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 31859
    },
    {
      "epoch": 7.452578645772424,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6054,
      "step": 31860
    },
    {
      "epoch": 7.452812536545434,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 31861
    },
    {
      "epoch": 7.4530464273184425,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 31862
    },
    {
      "epoch": 7.453280318091451,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 31863
    },
    {
      "epoch": 7.45351420886446,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6885,
      "step": 31864
    },
    {
      "epoch": 7.453748099637469,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5977,
      "step": 31865
    },
    {
      "epoch": 7.453981990410479,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 31866
    },
    {
      "epoch": 7.4542158811834875,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5286,
      "step": 31867
    },
    {
      "epoch": 7.454449771956496,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7117,
      "step": 31868
    },
    {
      "epoch": 7.454683662729505,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 31869
    },
    {
      "epoch": 7.454917553502514,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.6933,
      "step": 31870
    },
    {
      "epoch": 7.455151444275524,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.5134,
      "step": 31871
    },
    {
      "epoch": 7.455385335048533,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8141,
      "step": 31872
    },
    {
      "epoch": 7.4556192258215415,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 31873
    },
    {
      "epoch": 7.45585311659455,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 31874
    },
    {
      "epoch": 7.456087007367559,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 31875
    },
    {
      "epoch": 7.456320898140568,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 31876
    },
    {
      "epoch": 7.456554788913578,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 31877
    },
    {
      "epoch": 7.4567886796865865,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6958,
      "step": 31878
    },
    {
      "epoch": 7.457022570459595,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 31879
    },
    {
      "epoch": 7.457256461232604,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6421,
      "step": 31880
    },
    {
      "epoch": 7.457490352005613,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 31881
    },
    {
      "epoch": 7.457724242778623,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6258,
      "step": 31882
    },
    {
      "epoch": 7.457958133551632,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7298,
      "step": 31883
    },
    {
      "epoch": 7.4581920243246405,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6029,
      "step": 31884
    },
    {
      "epoch": 7.458425915097649,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 31885
    },
    {
      "epoch": 7.458659805870658,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8903,
      "step": 31886
    },
    {
      "epoch": 7.458893696643667,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 31887
    },
    {
      "epoch": 7.459127587416677,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6933,
      "step": 31888
    },
    {
      "epoch": 7.4593614781896855,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.5193,
      "step": 31889
    },
    {
      "epoch": 7.459595368962694,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7854,
      "step": 31890
    },
    {
      "epoch": 7.459829259735703,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6505,
      "step": 31891
    },
    {
      "epoch": 7.460063150508712,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4755,
      "step": 31892
    },
    {
      "epoch": 7.460297041281722,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 31893
    },
    {
      "epoch": 7.460530932054731,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5862,
      "step": 31894
    },
    {
      "epoch": 7.4607648228277395,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 31895
    },
    {
      "epoch": 7.460998713600748,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 31896
    },
    {
      "epoch": 7.461232604373757,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 31897
    },
    {
      "epoch": 7.461466495146767,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 31898
    },
    {
      "epoch": 7.461700385919776,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5057,
      "step": 31899
    },
    {
      "epoch": 7.4619342766927845,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 1.3977,
      "step": 31900
    },
    {
      "epoch": 7.4619342766927845,
      "eval_runtime": 4.589,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 31900
    },
    {
      "epoch": 7.462168167465793,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7589,
      "step": 31901
    },
    {
      "epoch": 7.462402058238802,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6634,
      "step": 31902
    },
    {
      "epoch": 7.462635949011812,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 31903
    },
    {
      "epoch": 7.462869839784821,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8187,
      "step": 31904
    },
    {
      "epoch": 7.46310373055783,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6961,
      "step": 31905
    },
    {
      "epoch": 7.4633376213308384,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 31906
    },
    {
      "epoch": 7.463571512103847,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 31907
    },
    {
      "epoch": 7.463805402876856,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 31908
    },
    {
      "epoch": 7.464039293649866,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5737,
      "step": 31909
    },
    {
      "epoch": 7.464273184422875,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8196,
      "step": 31910
    },
    {
      "epoch": 7.4645070751958835,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 31911
    },
    {
      "epoch": 7.464740965968892,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 31912
    },
    {
      "epoch": 7.464974856741901,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6858,
      "step": 31913
    },
    {
      "epoch": 7.465208747514911,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 31914
    },
    {
      "epoch": 7.46544263828792,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8276,
      "step": 31915
    },
    {
      "epoch": 7.465676529060929,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 31916
    },
    {
      "epoch": 7.465910419833937,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 31917
    },
    {
      "epoch": 7.466144310606946,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 31918
    },
    {
      "epoch": 7.466378201379955,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 31919
    },
    {
      "epoch": 7.466612092152965,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 31920
    },
    {
      "epoch": 7.466845982925974,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 31921
    },
    {
      "epoch": 7.4670798736989825,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 31922
    },
    {
      "epoch": 7.467313764471991,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 31923
    },
    {
      "epoch": 7.467547655245,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 31924
    },
    {
      "epoch": 7.46778154601801,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 31925
    },
    {
      "epoch": 7.468015436791019,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 31926
    },
    {
      "epoch": 7.468249327564028,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 31927
    },
    {
      "epoch": 7.468483218337036,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 31928
    },
    {
      "epoch": 7.468717109110045,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7359,
      "step": 31929
    },
    {
      "epoch": 7.468950999883055,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6829,
      "step": 31930
    },
    {
      "epoch": 7.469184890656064,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 31931
    },
    {
      "epoch": 7.469418781429073,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6033,
      "step": 31932
    },
    {
      "epoch": 7.4696526722020815,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6118,
      "step": 31933
    },
    {
      "epoch": 7.46988656297509,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 31934
    },
    {
      "epoch": 7.4701204537481,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5542,
      "step": 31935
    },
    {
      "epoch": 7.470354344521109,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6949,
      "step": 31936
    },
    {
      "epoch": 7.470588235294118,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8653,
      "step": 31937
    },
    {
      "epoch": 7.470822126067127,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.6957,
      "step": 31938
    },
    {
      "epoch": 7.471056016840135,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 31939
    },
    {
      "epoch": 7.471289907613144,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 31940
    },
    {
      "epoch": 7.471523798386154,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 31941
    },
    {
      "epoch": 7.471757689159163,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 31942
    },
    {
      "epoch": 7.471991579932172,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8347,
      "step": 31943
    },
    {
      "epoch": 7.4722254707051805,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 31944
    },
    {
      "epoch": 7.472459361478189,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 31945
    },
    {
      "epoch": 7.472693252251199,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6251,
      "step": 31946
    },
    {
      "epoch": 7.472927143024208,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 31947
    },
    {
      "epoch": 7.473161033797217,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 31948
    },
    {
      "epoch": 7.473394924570226,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6235,
      "step": 31949
    },
    {
      "epoch": 7.473628815343234,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3956,
      "step": 31950
    },
    {
      "epoch": 7.473862706116243,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.6033,
      "step": 31951
    },
    {
      "epoch": 7.474096596889253,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 31952
    },
    {
      "epoch": 7.474330487662262,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5413,
      "step": 31953
    },
    {
      "epoch": 7.474564378435271,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 31954
    },
    {
      "epoch": 7.4747982692082795,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 31955
    },
    {
      "epoch": 7.475032159981288,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 31956
    },
    {
      "epoch": 7.475266050754298,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6253,
      "step": 31957
    },
    {
      "epoch": 7.475499941527307,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 31958
    },
    {
      "epoch": 7.475733832300316,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 31959
    },
    {
      "epoch": 7.475967723073325,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9503,
      "step": 31960
    },
    {
      "epoch": 7.476201613846333,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 31961
    },
    {
      "epoch": 7.476435504619343,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6111,
      "step": 31962
    },
    {
      "epoch": 7.476669395392352,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5458,
      "step": 31963
    },
    {
      "epoch": 7.476903286165361,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 31964
    },
    {
      "epoch": 7.47713717693837,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 31965
    },
    {
      "epoch": 7.4773710677113785,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8305,
      "step": 31966
    },
    {
      "epoch": 7.477604958484388,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 31967
    },
    {
      "epoch": 7.477838849257397,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 31968
    },
    {
      "epoch": 7.478072740030406,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 31969
    },
    {
      "epoch": 7.478306630803415,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 31970
    },
    {
      "epoch": 7.478540521576424,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8043,
      "step": 31971
    },
    {
      "epoch": 7.478774412349432,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7939,
      "step": 31972
    },
    {
      "epoch": 7.479008303122442,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 31973
    },
    {
      "epoch": 7.479242193895451,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 31974
    },
    {
      "epoch": 7.47947608466846,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 31975
    },
    {
      "epoch": 7.479709975441469,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 31976
    },
    {
      "epoch": 7.4799438662144775,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 31977
    },
    {
      "epoch": 7.480177756987487,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6506,
      "step": 31978
    },
    {
      "epoch": 7.480411647760496,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4348,
      "step": 31979
    },
    {
      "epoch": 7.480645538533505,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 31980
    },
    {
      "epoch": 7.480879429306514,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.6505,
      "step": 31981
    },
    {
      "epoch": 7.481113320079523,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 31982
    },
    {
      "epoch": 7.481347210852531,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 31983
    },
    {
      "epoch": 7.481581101625541,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6266,
      "step": 31984
    },
    {
      "epoch": 7.48181499239855,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 31985
    },
    {
      "epoch": 7.482048883171559,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5771,
      "step": 31986
    },
    {
      "epoch": 7.482282773944568,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 31987
    },
    {
      "epoch": 7.4825166647175765,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 31988
    },
    {
      "epoch": 7.482750555490586,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.392,
      "step": 31989
    },
    {
      "epoch": 7.482984446263595,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 31990
    },
    {
      "epoch": 7.483218337036604,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7406,
      "step": 31991
    },
    {
      "epoch": 7.483452227809613,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.4918,
      "step": 31992
    },
    {
      "epoch": 7.483686118582622,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.701,
      "step": 31993
    },
    {
      "epoch": 7.483920009355631,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5216,
      "step": 31994
    },
    {
      "epoch": 7.48415390012864,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.658,
      "step": 31995
    },
    {
      "epoch": 7.484387790901649,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 31996
    },
    {
      "epoch": 7.484621681674658,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 31997
    },
    {
      "epoch": 7.484855572447667,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 31998
    },
    {
      "epoch": 7.485089463220676,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 31999
    },
    {
      "epoch": 7.485323353993685,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5432,
      "step": 32000
    },
    {
      "epoch": 7.485323353993685,
      "eval_runtime": 4.6053,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 32000
    },
    {
      "epoch": 7.485557244766694,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7207,
      "step": 32001
    },
    {
      "epoch": 7.485791135539703,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7678,
      "step": 32002
    },
    {
      "epoch": 7.486025026312712,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 32003
    },
    {
      "epoch": 7.4862589170857206,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 32004
    },
    {
      "epoch": 7.48649280785873,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 32005
    },
    {
      "epoch": 7.486726698631739,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 32006
    },
    {
      "epoch": 7.486960589404748,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5041,
      "step": 32007
    },
    {
      "epoch": 7.487194480177757,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 32008
    },
    {
      "epoch": 7.487428370950766,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5082,
      "step": 32009
    },
    {
      "epoch": 7.487662261723775,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6262,
      "step": 32010
    },
    {
      "epoch": 7.487896152496784,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5931,
      "step": 32011
    },
    {
      "epoch": 7.488130043269793,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 32012
    },
    {
      "epoch": 7.488363934042802,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 32013
    },
    {
      "epoch": 7.488597824815811,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 32014
    },
    {
      "epoch": 7.48883171558882,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 32015
    },
    {
      "epoch": 7.489065606361829,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 32016
    },
    {
      "epoch": 7.489299497134838,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5205,
      "step": 32017
    },
    {
      "epoch": 7.489533387907847,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 32018
    },
    {
      "epoch": 7.489767278680856,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 32019
    },
    {
      "epoch": 7.4900011694538655,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7981,
      "step": 32020
    },
    {
      "epoch": 7.490235060226874,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 32021
    },
    {
      "epoch": 7.490468950999883,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6271,
      "step": 32022
    },
    {
      "epoch": 7.490702841772892,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 32023
    },
    {
      "epoch": 7.490936732545901,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.689,
      "step": 32024
    },
    {
      "epoch": 7.49117062331891,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5653,
      "step": 32025
    },
    {
      "epoch": 7.491404514091919,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5881,
      "step": 32026
    },
    {
      "epoch": 7.491638404864928,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 32027
    },
    {
      "epoch": 7.491872295637937,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 32028
    },
    {
      "epoch": 7.492106186410946,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 32029
    },
    {
      "epoch": 7.492340077183955,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6086,
      "step": 32030
    },
    {
      "epoch": 7.4925739679569645,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 32031
    },
    {
      "epoch": 7.492807858729973,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 32032
    },
    {
      "epoch": 7.493041749502982,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.667,
      "step": 32033
    },
    {
      "epoch": 7.493275640275991,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 32034
    },
    {
      "epoch": 7.493509531049,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 32035
    },
    {
      "epoch": 7.493743421822009,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 32036
    },
    {
      "epoch": 7.493977312595018,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 32037
    },
    {
      "epoch": 7.494211203368027,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 32038
    },
    {
      "epoch": 7.494445094141036,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6688,
      "step": 32039
    },
    {
      "epoch": 7.494678984914045,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 32040
    },
    {
      "epoch": 7.494912875687054,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 32041
    },
    {
      "epoch": 7.4951467664600635,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 32042
    },
    {
      "epoch": 7.495380657233072,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 32043
    },
    {
      "epoch": 7.495614548006081,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 32044
    },
    {
      "epoch": 7.49584843877909,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 32045
    },
    {
      "epoch": 7.496082329552099,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6583,
      "step": 32046
    },
    {
      "epoch": 7.496316220325109,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5936,
      "step": 32047
    },
    {
      "epoch": 7.496550111098117,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7281,
      "step": 32048
    },
    {
      "epoch": 7.496784001871126,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 32049
    },
    {
      "epoch": 7.497017892644135,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 32050
    },
    {
      "epoch": 7.497251783417144,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 32051
    },
    {
      "epoch": 7.497485674190154,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 32052
    },
    {
      "epoch": 7.4977195649631625,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 32053
    },
    {
      "epoch": 7.497953455736171,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 32054
    },
    {
      "epoch": 7.49818734650918,
      "grad_norm": 14.0,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 32055
    },
    {
      "epoch": 7.498421237282189,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0902,
      "step": 32056
    },
    {
      "epoch": 7.498655128055198,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 32057
    },
    {
      "epoch": 7.498889018828208,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4225,
      "step": 32058
    },
    {
      "epoch": 7.499122909601216,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 32059
    },
    {
      "epoch": 7.499356800374225,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9503,
      "step": 32060
    },
    {
      "epoch": 7.499590691147234,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.4113,
      "step": 32061
    },
    {
      "epoch": 7.499824581920243,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7188,
      "step": 32062
    },
    {
      "epoch": 7.500058472693253,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 32063
    },
    {
      "epoch": 7.5002923634662615,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6266,
      "step": 32064
    },
    {
      "epoch": 7.50052625423927,
      "grad_norm": 7.5625,
      "learning_rate": 3e-05,
      "loss": 1.668,
      "step": 32065
    },
    {
      "epoch": 7.500760145012279,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 32066
    },
    {
      "epoch": 7.500994035785288,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6176,
      "step": 32067
    },
    {
      "epoch": 7.501227926558297,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 32068
    },
    {
      "epoch": 7.501461817331307,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 32069
    },
    {
      "epoch": 7.501695708104315,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 32070
    },
    {
      "epoch": 7.501929598877324,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6821,
      "step": 32071
    },
    {
      "epoch": 7.502163489650333,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 32072
    },
    {
      "epoch": 7.502397380423342,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 32073
    },
    {
      "epoch": 7.502631271196352,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 32074
    },
    {
      "epoch": 7.5028651619693605,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 32075
    },
    {
      "epoch": 7.503099052742369,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 32076
    },
    {
      "epoch": 7.503332943515378,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6559,
      "step": 32077
    },
    {
      "epoch": 7.503566834288387,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6457,
      "step": 32078
    },
    {
      "epoch": 7.503800725061396,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7637,
      "step": 32079
    },
    {
      "epoch": 7.504034615834406,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 32080
    },
    {
      "epoch": 7.504268506607414,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 32081
    },
    {
      "epoch": 7.504502397380423,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 32082
    },
    {
      "epoch": 7.504736288153432,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 32083
    },
    {
      "epoch": 7.504970178926442,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.587,
      "step": 32084
    },
    {
      "epoch": 7.505204069699451,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6683,
      "step": 32085
    },
    {
      "epoch": 7.5054379604724595,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 32086
    },
    {
      "epoch": 7.505671851245468,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 32087
    },
    {
      "epoch": 7.505905742018477,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 32088
    },
    {
      "epoch": 7.506139632791486,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 32089
    },
    {
      "epoch": 7.506373523564496,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 32090
    },
    {
      "epoch": 7.506607414337505,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 32091
    },
    {
      "epoch": 7.506841305110513,
      "grad_norm": 13.5625,
      "learning_rate": 3e-05,
      "loss": 1.8166,
      "step": 32092
    },
    {
      "epoch": 7.507075195883522,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4827,
      "step": 32093
    },
    {
      "epoch": 7.507309086656531,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 32094
    },
    {
      "epoch": 7.507542977429541,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6539,
      "step": 32095
    },
    {
      "epoch": 7.50777686820255,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6885,
      "step": 32096
    },
    {
      "epoch": 7.5080107589755585,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 32097
    },
    {
      "epoch": 7.508244649748567,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6739,
      "step": 32098
    },
    {
      "epoch": 7.508478540521576,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 32099
    },
    {
      "epoch": 7.508712431294585,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 32100
    },
    {
      "epoch": 7.508712431294585,
      "eval_runtime": 4.6009,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 32100
    },
    {
      "epoch": 7.508946322067595,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 32101
    },
    {
      "epoch": 7.5091802128406036,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 32102
    },
    {
      "epoch": 7.509414103613612,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5383,
      "step": 32103
    },
    {
      "epoch": 7.509647994386621,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 32104
    },
    {
      "epoch": 7.50988188515963,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 32105
    },
    {
      "epoch": 7.51011577593264,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6821,
      "step": 32106
    },
    {
      "epoch": 7.510349666705649,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5588,
      "step": 32107
    },
    {
      "epoch": 7.5105835574786575,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 32108
    },
    {
      "epoch": 7.510817448251666,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 32109
    },
    {
      "epoch": 7.511051339024675,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 32110
    },
    {
      "epoch": 7.511285229797684,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 32111
    },
    {
      "epoch": 7.511519120570694,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6128,
      "step": 32112
    },
    {
      "epoch": 7.5117530113437025,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4911,
      "step": 32113
    },
    {
      "epoch": 7.511986902116711,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 32114
    },
    {
      "epoch": 7.51222079288972,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7494,
      "step": 32115
    },
    {
      "epoch": 7.51245468366273,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6391,
      "step": 32116
    },
    {
      "epoch": 7.512688574435739,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 32117
    },
    {
      "epoch": 7.512922465208748,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 32118
    },
    {
      "epoch": 7.5131563559817565,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.7321,
      "step": 32119
    },
    {
      "epoch": 7.513390246754765,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 32120
    },
    {
      "epoch": 7.513624137527774,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.6577,
      "step": 32121
    },
    {
      "epoch": 7.513858028300784,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7449,
      "step": 32122
    },
    {
      "epoch": 7.514091919073793,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 32123
    },
    {
      "epoch": 7.5143258098468015,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5643,
      "step": 32124
    },
    {
      "epoch": 7.51455970061981,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 32125
    },
    {
      "epoch": 7.514793591392819,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 32126
    },
    {
      "epoch": 7.515027482165829,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 32127
    },
    {
      "epoch": 7.515261372938838,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 32128
    },
    {
      "epoch": 7.515495263711847,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.5293,
      "step": 32129
    },
    {
      "epoch": 7.5157291544848555,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 32130
    },
    {
      "epoch": 7.515963045257864,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 32131
    },
    {
      "epoch": 7.516196936030873,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.7558,
      "step": 32132
    },
    {
      "epoch": 7.516430826803883,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6552,
      "step": 32133
    },
    {
      "epoch": 7.516664717576892,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 32134
    },
    {
      "epoch": 7.5168986083499005,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 32135
    },
    {
      "epoch": 7.517132499122909,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.5165,
      "step": 32136
    },
    {
      "epoch": 7.517366389895919,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 32137
    },
    {
      "epoch": 7.517600280668928,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5671,
      "step": 32138
    },
    {
      "epoch": 7.517834171441937,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 32139
    },
    {
      "epoch": 7.518068062214946,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8089,
      "step": 32140
    },
    {
      "epoch": 7.5183019529879545,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 32141
    },
    {
      "epoch": 7.518535843760963,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5759,
      "step": 32142
    },
    {
      "epoch": 7.518769734533973,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 32143
    },
    {
      "epoch": 7.519003625306982,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6524,
      "step": 32144
    },
    {
      "epoch": 7.519237516079991,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 32145
    },
    {
      "epoch": 7.5194714068529995,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6635,
      "step": 32146
    },
    {
      "epoch": 7.519705297626008,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 32147
    },
    {
      "epoch": 7.519939188399018,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 32148
    },
    {
      "epoch": 7.520173079172027,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 32149
    },
    {
      "epoch": 7.520406969945036,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 32150
    },
    {
      "epoch": 7.520640860718045,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 32151
    },
    {
      "epoch": 7.5208747514910534,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7632,
      "step": 32152
    },
    {
      "epoch": 7.521108642264062,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6589,
      "step": 32153
    },
    {
      "epoch": 7.521342533037072,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6993,
      "step": 32154
    },
    {
      "epoch": 7.521576423810081,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7041,
      "step": 32155
    },
    {
      "epoch": 7.52181031458309,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5384,
      "step": 32156
    },
    {
      "epoch": 7.5220442053560985,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 32157
    },
    {
      "epoch": 7.522278096129107,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 32158
    },
    {
      "epoch": 7.522511986902117,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 32159
    },
    {
      "epoch": 7.522745877675126,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 32160
    },
    {
      "epoch": 7.522979768448135,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 32161
    },
    {
      "epoch": 7.523213659221144,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.3447,
      "step": 32162
    },
    {
      "epoch": 7.523447549994152,
      "grad_norm": 8.4375,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 32163
    },
    {
      "epoch": 7.523681440767161,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 32164
    },
    {
      "epoch": 7.523915331540171,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 32165
    },
    {
      "epoch": 7.52414922231318,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 32166
    },
    {
      "epoch": 7.524383113086189,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.572,
      "step": 32167
    },
    {
      "epoch": 7.5246170038591975,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6582,
      "step": 32168
    },
    {
      "epoch": 7.524850894632207,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 32169
    },
    {
      "epoch": 7.525084785405216,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 32170
    },
    {
      "epoch": 7.525318676178225,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5389,
      "step": 32171
    },
    {
      "epoch": 7.525552566951234,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 32172
    },
    {
      "epoch": 7.525786457724243,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5587,
      "step": 32173
    },
    {
      "epoch": 7.526020348497251,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6445,
      "step": 32174
    },
    {
      "epoch": 7.526254239270261,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6806,
      "step": 32175
    },
    {
      "epoch": 7.52648813004327,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5283,
      "step": 32176
    },
    {
      "epoch": 7.526722020816279,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6803,
      "step": 32177
    },
    {
      "epoch": 7.526955911589288,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5777,
      "step": 32178
    },
    {
      "epoch": 7.5271898023622965,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6581,
      "step": 32179
    },
    {
      "epoch": 7.527423693135306,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.4746,
      "step": 32180
    },
    {
      "epoch": 7.527657583908315,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4709,
      "step": 32181
    },
    {
      "epoch": 7.527891474681324,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 32182
    },
    {
      "epoch": 7.528125365454333,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 32183
    },
    {
      "epoch": 7.528359256227342,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7764,
      "step": 32184
    },
    {
      "epoch": 7.52859314700035,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4286,
      "step": 32185
    },
    {
      "epoch": 7.52882703777336,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.4877,
      "step": 32186
    },
    {
      "epoch": 7.529060928546369,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6404,
      "step": 32187
    },
    {
      "epoch": 7.529294819319378,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7534,
      "step": 32188
    },
    {
      "epoch": 7.529528710092387,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 32189
    },
    {
      "epoch": 7.5297626008653955,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6437,
      "step": 32190
    },
    {
      "epoch": 7.529996491638405,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 32191
    },
    {
      "epoch": 7.530230382411414,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 32192
    },
    {
      "epoch": 7.530464273184423,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7862,
      "step": 32193
    },
    {
      "epoch": 7.530698163957432,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 32194
    },
    {
      "epoch": 7.530932054730441,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 32195
    },
    {
      "epoch": 7.531165945503449,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 32196
    },
    {
      "epoch": 7.531399836276459,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7117,
      "step": 32197
    },
    {
      "epoch": 7.531633727049468,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 32198
    },
    {
      "epoch": 7.531867617822477,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 32199
    },
    {
      "epoch": 7.532101508595486,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 32200
    },
    {
      "epoch": 7.532101508595486,
      "eval_runtime": 4.6205,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 32200
    },
    {
      "epoch": 7.532335399368495,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 32201
    },
    {
      "epoch": 7.532569290141504,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7971,
      "step": 32202
    },
    {
      "epoch": 7.532803180914513,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 32203
    },
    {
      "epoch": 7.533037071687522,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7039,
      "step": 32204
    },
    {
      "epoch": 7.533270962460531,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6335,
      "step": 32205
    },
    {
      "epoch": 7.53350485323354,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 32206
    },
    {
      "epoch": 7.533738744006549,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 32207
    },
    {
      "epoch": 7.533972634779558,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 32208
    },
    {
      "epoch": 7.534206525552567,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.3709,
      "step": 32209
    },
    {
      "epoch": 7.534440416325576,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6716,
      "step": 32210
    },
    {
      "epoch": 7.534674307098585,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7049,
      "step": 32211
    },
    {
      "epoch": 7.534908197871594,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 32212
    },
    {
      "epoch": 7.535142088644603,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 32213
    },
    {
      "epoch": 7.535375979417612,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 32214
    },
    {
      "epoch": 7.535609870190621,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 32215
    },
    {
      "epoch": 7.53584376096363,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 32216
    },
    {
      "epoch": 7.536077651736639,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8075,
      "step": 32217
    },
    {
      "epoch": 7.536311542509648,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5329,
      "step": 32218
    },
    {
      "epoch": 7.536545433282657,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 32219
    },
    {
      "epoch": 7.536779324055666,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1204,
      "step": 32220
    },
    {
      "epoch": 7.537013214828675,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7022,
      "step": 32221
    },
    {
      "epoch": 7.537247105601684,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7257,
      "step": 32222
    },
    {
      "epoch": 7.537480996374693,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 32223
    },
    {
      "epoch": 7.537714887147702,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 32224
    },
    {
      "epoch": 7.537948777920711,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 32225
    },
    {
      "epoch": 7.53818266869372,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 32226
    },
    {
      "epoch": 7.538416559466729,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6271,
      "step": 32227
    },
    {
      "epoch": 7.538650450239738,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 32228
    },
    {
      "epoch": 7.538884341012747,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.758,
      "step": 32229
    },
    {
      "epoch": 7.539118231785756,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7488,
      "step": 32230
    },
    {
      "epoch": 7.539352122558765,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5105,
      "step": 32231
    },
    {
      "epoch": 7.539586013331774,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6202,
      "step": 32232
    },
    {
      "epoch": 7.5398199041047835,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 32233
    },
    {
      "epoch": 7.540053794877792,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 32234
    },
    {
      "epoch": 7.540287685650801,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6874,
      "step": 32235
    },
    {
      "epoch": 7.54052157642381,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 32236
    },
    {
      "epoch": 7.540755467196819,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 32237
    },
    {
      "epoch": 7.540989357969828,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 32238
    },
    {
      "epoch": 7.5412232487428374,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6477,
      "step": 32239
    },
    {
      "epoch": 7.541457139515846,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6234,
      "step": 32240
    },
    {
      "epoch": 7.541691030288855,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 32241
    },
    {
      "epoch": 7.541924921061864,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6282,
      "step": 32242
    },
    {
      "epoch": 7.542158811834873,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6747,
      "step": 32243
    },
    {
      "epoch": 7.5423927026078825,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 32244
    },
    {
      "epoch": 7.542626593380891,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 32245
    },
    {
      "epoch": 7.5428604841539,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 32246
    },
    {
      "epoch": 7.543094374926909,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 32247
    },
    {
      "epoch": 7.543328265699918,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8058,
      "step": 32248
    },
    {
      "epoch": 7.543562156472927,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 32249
    },
    {
      "epoch": 7.543796047245936,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 32250
    },
    {
      "epoch": 7.544029938018945,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 32251
    },
    {
      "epoch": 7.544263828791954,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7104,
      "step": 32252
    },
    {
      "epoch": 7.544497719564963,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6116,
      "step": 32253
    },
    {
      "epoch": 7.544731610337972,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6458,
      "step": 32254
    },
    {
      "epoch": 7.5449655011109815,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 32255
    },
    {
      "epoch": 7.54519939188399,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 32256
    },
    {
      "epoch": 7.545433282656999,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 32257
    },
    {
      "epoch": 7.545667173430008,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 32258
    },
    {
      "epoch": 7.545901064203017,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1255,
      "step": 32259
    },
    {
      "epoch": 7.546134954976026,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9946,
      "step": 32260
    },
    {
      "epoch": 7.546368845749035,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 32261
    },
    {
      "epoch": 7.546602736522044,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 32262
    },
    {
      "epoch": 7.546836627295053,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 32263
    },
    {
      "epoch": 7.547070518068062,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 32264
    },
    {
      "epoch": 7.547304408841072,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 32265
    },
    {
      "epoch": 7.5475382996140805,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 32266
    },
    {
      "epoch": 7.547772190387089,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.612,
      "step": 32267
    },
    {
      "epoch": 7.548006081160098,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 32268
    },
    {
      "epoch": 7.548239971933107,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 32269
    },
    {
      "epoch": 7.548473862706116,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 32270
    },
    {
      "epoch": 7.548707753479126,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 32271
    },
    {
      "epoch": 7.548941644252134,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 32272
    },
    {
      "epoch": 7.549175535025143,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.758,
      "step": 32273
    },
    {
      "epoch": 7.549409425798152,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 32274
    },
    {
      "epoch": 7.549643316571161,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 32275
    },
    {
      "epoch": 7.549877207344171,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 32276
    },
    {
      "epoch": 7.5501110981171795,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.578,
      "step": 32277
    },
    {
      "epoch": 7.550344988890188,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 32278
    },
    {
      "epoch": 7.550578879663197,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 32279
    },
    {
      "epoch": 7.550812770436206,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 32280
    },
    {
      "epoch": 7.551046661209215,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 32281
    },
    {
      "epoch": 7.551280551982225,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6292,
      "step": 32282
    },
    {
      "epoch": 7.551514442755233,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 32283
    },
    {
      "epoch": 7.551748333528242,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 32284
    },
    {
      "epoch": 7.551982224301251,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 32285
    },
    {
      "epoch": 7.55221611507426,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7749,
      "step": 32286
    },
    {
      "epoch": 7.55245000584727,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 32287
    },
    {
      "epoch": 7.5526838966202785,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6764,
      "step": 32288
    },
    {
      "epoch": 7.552917787393287,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5886,
      "step": 32289
    },
    {
      "epoch": 7.553151678166296,
      "grad_norm": 12.0,
      "learning_rate": 3e-05,
      "loss": 2.1805,
      "step": 32290
    },
    {
      "epoch": 7.553385568939305,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6443,
      "step": 32291
    },
    {
      "epoch": 7.553619459712314,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5437,
      "step": 32292
    },
    {
      "epoch": 7.553853350485324,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 32293
    },
    {
      "epoch": 7.554087241258332,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 32294
    },
    {
      "epoch": 7.554321132031341,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 32295
    },
    {
      "epoch": 7.55455502280435,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.736,
      "step": 32296
    },
    {
      "epoch": 7.55478891357736,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6406,
      "step": 32297
    },
    {
      "epoch": 7.555022804350369,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 32298
    },
    {
      "epoch": 7.5552566951233775,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 32299
    },
    {
      "epoch": 7.555490585896386,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 32300
    },
    {
      "epoch": 7.555490585896386,
      "eval_runtime": 4.6035,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 32300
    },
    {
      "epoch": 7.555724476669395,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5073,
      "step": 32301
    },
    {
      "epoch": 7.555958367442404,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5176,
      "step": 32302
    },
    {
      "epoch": 7.556192258215414,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5084,
      "step": 32303
    },
    {
      "epoch": 7.556426148988423,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6295,
      "step": 32304
    },
    {
      "epoch": 7.556660039761431,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6291,
      "step": 32305
    },
    {
      "epoch": 7.55689393053444,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6519,
      "step": 32306
    },
    {
      "epoch": 7.557127821307449,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 32307
    },
    {
      "epoch": 7.557361712080459,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5695,
      "step": 32308
    },
    {
      "epoch": 7.557595602853468,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 32309
    },
    {
      "epoch": 7.5578294936264765,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7392,
      "step": 32310
    },
    {
      "epoch": 7.558063384399485,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 32311
    },
    {
      "epoch": 7.558297275172494,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4619,
      "step": 32312
    },
    {
      "epoch": 7.558531165945503,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 32313
    },
    {
      "epoch": 7.558765056718513,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 32314
    },
    {
      "epoch": 7.558998947491522,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 32315
    },
    {
      "epoch": 7.55923283826453,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6901,
      "step": 32316
    },
    {
      "epoch": 7.559466729037539,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5373,
      "step": 32317
    },
    {
      "epoch": 7.559700619810549,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 32318
    },
    {
      "epoch": 7.559934510583558,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.3871,
      "step": 32319
    },
    {
      "epoch": 7.560168401356567,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 32320
    },
    {
      "epoch": 7.5604022921295755,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 32321
    },
    {
      "epoch": 7.560636182902584,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6143,
      "step": 32322
    },
    {
      "epoch": 7.560870073675593,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6823,
      "step": 32323
    },
    {
      "epoch": 7.561103964448603,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 32324
    },
    {
      "epoch": 7.561337855221612,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 32325
    },
    {
      "epoch": 7.561571745994621,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 32326
    },
    {
      "epoch": 7.561805636767629,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 32327
    },
    {
      "epoch": 7.562039527540638,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 32328
    },
    {
      "epoch": 7.562273418313648,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 32329
    },
    {
      "epoch": 7.562507309086657,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 32330
    },
    {
      "epoch": 7.562741199859666,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 32331
    },
    {
      "epoch": 7.5629750906326745,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6975,
      "step": 32332
    },
    {
      "epoch": 7.563208981405683,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 32333
    },
    {
      "epoch": 7.563442872178692,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 32334
    },
    {
      "epoch": 7.563676762951702,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6091,
      "step": 32335
    },
    {
      "epoch": 7.563910653724711,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6919,
      "step": 32336
    },
    {
      "epoch": 7.56414454449772,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 32337
    },
    {
      "epoch": 7.564378435270728,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5982,
      "step": 32338
    },
    {
      "epoch": 7.564612326043737,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6145,
      "step": 32339
    },
    {
      "epoch": 7.564846216816747,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5597,
      "step": 32340
    },
    {
      "epoch": 7.565080107589756,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 32341
    },
    {
      "epoch": 7.565313998362765,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 32342
    },
    {
      "epoch": 7.5655478891357735,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 32343
    },
    {
      "epoch": 7.565781779908782,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 32344
    },
    {
      "epoch": 7.566015670681791,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 32345
    },
    {
      "epoch": 7.566249561454801,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 32346
    },
    {
      "epoch": 7.56648345222781,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 32347
    },
    {
      "epoch": 7.5667173430008186,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 32348
    },
    {
      "epoch": 7.566951233773827,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 32349
    },
    {
      "epoch": 7.567185124546837,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 32350
    },
    {
      "epoch": 7.567419015319846,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 32351
    },
    {
      "epoch": 7.567652906092855,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 32352
    },
    {
      "epoch": 7.567886796865864,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5749,
      "step": 32353
    },
    {
      "epoch": 7.5681206876388725,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 32354
    },
    {
      "epoch": 7.568354578411881,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6624,
      "step": 32355
    },
    {
      "epoch": 7.568588469184891,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7883,
      "step": 32356
    },
    {
      "epoch": 7.5688223599579,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6135,
      "step": 32357
    },
    {
      "epoch": 7.569056250730909,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 32358
    },
    {
      "epoch": 7.5692901415039175,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 32359
    },
    {
      "epoch": 7.569524032276926,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.4466,
      "step": 32360
    },
    {
      "epoch": 7.569757923049936,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 32361
    },
    {
      "epoch": 7.569991813822945,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 32362
    },
    {
      "epoch": 7.570225704595954,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 32363
    },
    {
      "epoch": 7.570459595368963,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 32364
    },
    {
      "epoch": 7.5706934861419715,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 32365
    },
    {
      "epoch": 7.57092737691498,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 32366
    },
    {
      "epoch": 7.57116126768799,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 32367
    },
    {
      "epoch": 7.571395158460999,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 32368
    },
    {
      "epoch": 7.571629049234008,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6341,
      "step": 32369
    },
    {
      "epoch": 7.5718629400070165,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 32370
    },
    {
      "epoch": 7.572096830780025,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 32371
    },
    {
      "epoch": 7.572330721553035,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 32372
    },
    {
      "epoch": 7.572564612326044,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8318,
      "step": 32373
    },
    {
      "epoch": 7.572798503099053,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.653,
      "step": 32374
    },
    {
      "epoch": 7.573032393872062,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8409,
      "step": 32375
    },
    {
      "epoch": 7.5732662846450705,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6162,
      "step": 32376
    },
    {
      "epoch": 7.573500175418079,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 32377
    },
    {
      "epoch": 7.573734066191089,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 32378
    },
    {
      "epoch": 7.573967956964098,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8156,
      "step": 32379
    },
    {
      "epoch": 7.574201847737107,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 32380
    },
    {
      "epoch": 7.5744357385101155,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7615,
      "step": 32381
    },
    {
      "epoch": 7.574669629283125,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 32382
    },
    {
      "epoch": 7.574903520056134,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.613,
      "step": 32383
    },
    {
      "epoch": 7.575137410829143,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 32384
    },
    {
      "epoch": 7.575371301602152,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 32385
    },
    {
      "epoch": 7.575605192375161,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6339,
      "step": 32386
    },
    {
      "epoch": 7.5758390831481695,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 32387
    },
    {
      "epoch": 7.576072973921179,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 32388
    },
    {
      "epoch": 7.576306864694188,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6396,
      "step": 32389
    },
    {
      "epoch": 7.576540755467197,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 32390
    },
    {
      "epoch": 7.576774646240206,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 32391
    },
    {
      "epoch": 7.5770085370132145,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 32392
    },
    {
      "epoch": 7.577242427786224,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 32393
    },
    {
      "epoch": 7.577476318559233,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 32394
    },
    {
      "epoch": 7.577710209332242,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4795,
      "step": 32395
    },
    {
      "epoch": 7.577944100105251,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 32396
    },
    {
      "epoch": 7.57817799087826,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 32397
    },
    {
      "epoch": 7.5784118816512684,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 32398
    },
    {
      "epoch": 7.578645772424278,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6966,
      "step": 32399
    },
    {
      "epoch": 7.578879663197287,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.274,
      "step": 32400
    },
    {
      "epoch": 7.578879663197287,
      "eval_runtime": 4.6136,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 32400
    },
    {
      "epoch": 7.579113553970296,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 32401
    },
    {
      "epoch": 7.579347444743305,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 32402
    },
    {
      "epoch": 7.5795813355163135,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 32403
    },
    {
      "epoch": 7.579815226289323,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 32404
    },
    {
      "epoch": 7.580049117062332,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 32405
    },
    {
      "epoch": 7.580283007835341,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 32406
    },
    {
      "epoch": 7.58051689860835,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3573,
      "step": 32407
    },
    {
      "epoch": 7.580750789381359,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 32408
    },
    {
      "epoch": 7.580984680154367,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6398,
      "step": 32409
    },
    {
      "epoch": 7.581218570927377,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 32410
    },
    {
      "epoch": 7.581452461700386,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 32411
    },
    {
      "epoch": 7.581686352473395,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8644,
      "step": 32412
    },
    {
      "epoch": 7.581920243246404,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6335,
      "step": 32413
    },
    {
      "epoch": 7.582154134019413,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 32414
    },
    {
      "epoch": 7.582388024792422,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 32415
    },
    {
      "epoch": 7.582621915565431,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 32416
    },
    {
      "epoch": 7.58285580633844,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.625,
      "step": 32417
    },
    {
      "epoch": 7.583089697111449,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 32418
    },
    {
      "epoch": 7.583323587884458,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6534,
      "step": 32419
    },
    {
      "epoch": 7.583557478657467,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6236,
      "step": 32420
    },
    {
      "epoch": 7.583791369430476,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7117,
      "step": 32421
    },
    {
      "epoch": 7.584025260203485,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 32422
    },
    {
      "epoch": 7.584259150976494,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 32423
    },
    {
      "epoch": 7.584493041749503,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5934,
      "step": 32424
    },
    {
      "epoch": 7.584726932522512,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4811,
      "step": 32425
    },
    {
      "epoch": 7.584960823295521,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 32426
    },
    {
      "epoch": 7.58519471406853,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 32427
    },
    {
      "epoch": 7.585428604841539,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 32428
    },
    {
      "epoch": 7.585662495614548,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6408,
      "step": 32429
    },
    {
      "epoch": 7.585896386387557,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 32430
    },
    {
      "epoch": 7.586130277160566,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 32431
    },
    {
      "epoch": 7.586364167933575,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 32432
    },
    {
      "epoch": 7.586598058706584,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 32433
    },
    {
      "epoch": 7.586831949479593,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6055,
      "step": 32434
    },
    {
      "epoch": 7.587065840252602,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5474,
      "step": 32435
    },
    {
      "epoch": 7.587299731025611,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 32436
    },
    {
      "epoch": 7.58753362179862,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 32437
    },
    {
      "epoch": 7.587767512571629,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 32438
    },
    {
      "epoch": 7.588001403344638,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 32439
    },
    {
      "epoch": 7.588235294117647,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8219,
      "step": 32440
    },
    {
      "epoch": 7.588469184890656,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.4048,
      "step": 32441
    },
    {
      "epoch": 7.588703075663665,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6604,
      "step": 32442
    },
    {
      "epoch": 7.588936966436674,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9094,
      "step": 32443
    },
    {
      "epoch": 7.589170857209683,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 32444
    },
    {
      "epoch": 7.589404747982692,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.4652,
      "step": 32445
    },
    {
      "epoch": 7.5896386387557015,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.857,
      "step": 32446
    },
    {
      "epoch": 7.58987252952871,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 32447
    },
    {
      "epoch": 7.590106420301719,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 32448
    },
    {
      "epoch": 7.590340311074728,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 32449
    },
    {
      "epoch": 7.590574201847737,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 32450
    },
    {
      "epoch": 7.590808092620746,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 32451
    },
    {
      "epoch": 7.5910419833937555,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 32452
    },
    {
      "epoch": 7.591275874166764,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 32453
    },
    {
      "epoch": 7.591509764939773,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 32454
    },
    {
      "epoch": 7.591743655712782,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 32455
    },
    {
      "epoch": 7.591977546485791,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 32456
    },
    {
      "epoch": 7.5922114372588005,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 32457
    },
    {
      "epoch": 7.592445328031809,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5839,
      "step": 32458
    },
    {
      "epoch": 7.592679218804818,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 32459
    },
    {
      "epoch": 7.592913109577827,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4742,
      "step": 32460
    },
    {
      "epoch": 7.593147000350836,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 32461
    },
    {
      "epoch": 7.593380891123845,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5935,
      "step": 32462
    },
    {
      "epoch": 7.5936147818968545,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 32463
    },
    {
      "epoch": 7.593848672669863,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 32464
    },
    {
      "epoch": 7.594082563442872,
      "grad_norm": 7.25,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 32465
    },
    {
      "epoch": 7.594316454215881,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 32466
    },
    {
      "epoch": 7.59455034498889,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8535,
      "step": 32467
    },
    {
      "epoch": 7.5947842357618995,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 32468
    },
    {
      "epoch": 7.595018126534908,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 32469
    },
    {
      "epoch": 7.595252017307917,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.648,
      "step": 32470
    },
    {
      "epoch": 7.595485908080926,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 32471
    },
    {
      "epoch": 7.595719798853935,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 32472
    },
    {
      "epoch": 7.595953689626944,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8025,
      "step": 32473
    },
    {
      "epoch": 7.5961875803999535,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 32474
    },
    {
      "epoch": 7.596421471172962,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.668,
      "step": 32475
    },
    {
      "epoch": 7.596655361945971,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2332,
      "step": 32476
    },
    {
      "epoch": 7.59688925271898,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 32477
    },
    {
      "epoch": 7.59712314349199,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.64,
      "step": 32478
    },
    {
      "epoch": 7.5973570342649985,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6073,
      "step": 32479
    },
    {
      "epoch": 7.597590925038007,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6199,
      "step": 32480
    },
    {
      "epoch": 7.597824815811016,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 32481
    },
    {
      "epoch": 7.598058706584025,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 32482
    },
    {
      "epoch": 7.598292597357034,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.5734,
      "step": 32483
    },
    {
      "epoch": 7.598526488130044,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 32484
    },
    {
      "epoch": 7.5987603789030524,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 32485
    },
    {
      "epoch": 7.598994269676061,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6578,
      "step": 32486
    },
    {
      "epoch": 7.59922816044907,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 32487
    },
    {
      "epoch": 7.599462051222079,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 32488
    },
    {
      "epoch": 7.599695941995089,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 32489
    },
    {
      "epoch": 7.5999298327680975,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 32490
    },
    {
      "epoch": 7.600163723541106,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 32491
    },
    {
      "epoch": 7.600397614314115,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6675,
      "step": 32492
    },
    {
      "epoch": 7.600631505087124,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7587,
      "step": 32493
    },
    {
      "epoch": 7.600865395860133,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 32494
    },
    {
      "epoch": 7.601099286633143,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 32495
    },
    {
      "epoch": 7.601333177406151,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7782,
      "step": 32496
    },
    {
      "epoch": 7.60156706817916,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 32497
    },
    {
      "epoch": 7.601800958952169,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5282,
      "step": 32498
    },
    {
      "epoch": 7.602034849725178,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 32499
    },
    {
      "epoch": 7.602268740498188,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 32500
    },
    {
      "epoch": 7.602268740498188,
      "eval_runtime": 4.6281,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 32500
    },
    {
      "epoch": 7.6025026312711965,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 32501
    },
    {
      "epoch": 7.602736522044205,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5946,
      "step": 32502
    },
    {
      "epoch": 7.602970412817214,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 32503
    },
    {
      "epoch": 7.603204303590223,
      "grad_norm": 11.375,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 32504
    },
    {
      "epoch": 7.603438194363232,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 32505
    },
    {
      "epoch": 7.603672085136242,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 32506
    },
    {
      "epoch": 7.60390597590925,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5333,
      "step": 32507
    },
    {
      "epoch": 7.604139866682259,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 32508
    },
    {
      "epoch": 7.604373757455268,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 32509
    },
    {
      "epoch": 7.604607648228278,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 32510
    },
    {
      "epoch": 7.604841539001287,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 32511
    },
    {
      "epoch": 7.6050754297742955,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0059,
      "step": 32512
    },
    {
      "epoch": 7.605309320547304,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 32513
    },
    {
      "epoch": 7.605543211320313,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 32514
    },
    {
      "epoch": 7.605777102093322,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7234,
      "step": 32515
    },
    {
      "epoch": 7.606010992866332,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.5686,
      "step": 32516
    },
    {
      "epoch": 7.606244883639341,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 32517
    },
    {
      "epoch": 7.606478774412349,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5479,
      "step": 32518
    },
    {
      "epoch": 7.606712665185358,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5233,
      "step": 32519
    },
    {
      "epoch": 7.606946555958367,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 32520
    },
    {
      "epoch": 7.607180446731377,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 32521
    },
    {
      "epoch": 7.607414337504386,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5413,
      "step": 32522
    },
    {
      "epoch": 7.6076482282773945,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 32523
    },
    {
      "epoch": 7.607882119050403,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 32524
    },
    {
      "epoch": 7.608116009823412,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7907,
      "step": 32525
    },
    {
      "epoch": 7.608349900596421,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 32526
    },
    {
      "epoch": 7.608583791369431,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 32527
    },
    {
      "epoch": 7.60881768214244,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6636,
      "step": 32528
    },
    {
      "epoch": 7.609051572915448,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 32529
    },
    {
      "epoch": 7.609285463688457,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6447,
      "step": 32530
    },
    {
      "epoch": 7.609519354461467,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4882,
      "step": 32531
    },
    {
      "epoch": 7.609753245234476,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5754,
      "step": 32532
    },
    {
      "epoch": 7.609987136007485,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 32533
    },
    {
      "epoch": 7.6102210267804935,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.3403,
      "step": 32534
    },
    {
      "epoch": 7.610454917553502,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 32535
    },
    {
      "epoch": 7.610688808326511,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6693,
      "step": 32536
    },
    {
      "epoch": 7.610922699099521,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 32537
    },
    {
      "epoch": 7.61115658987253,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 32538
    },
    {
      "epoch": 7.611390480645539,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 32539
    },
    {
      "epoch": 7.611624371418547,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 32540
    },
    {
      "epoch": 7.611858262191556,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 32541
    },
    {
      "epoch": 7.612092152964566,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 32542
    },
    {
      "epoch": 7.612326043737575,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 32543
    },
    {
      "epoch": 7.612559934510584,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 32544
    },
    {
      "epoch": 7.6127938252835925,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 32545
    },
    {
      "epoch": 7.613027716056601,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 32546
    },
    {
      "epoch": 7.61326160682961,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 32547
    },
    {
      "epoch": 7.61349549760262,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 32548
    },
    {
      "epoch": 7.613729388375629,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 32549
    },
    {
      "epoch": 7.613963279148638,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 32550
    },
    {
      "epoch": 7.614197169921646,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5712,
      "step": 32551
    },
    {
      "epoch": 7.614431060694655,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 32552
    },
    {
      "epoch": 7.614664951467665,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6827,
      "step": 32553
    },
    {
      "epoch": 7.614898842240674,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7101,
      "step": 32554
    },
    {
      "epoch": 7.615132733013683,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7454,
      "step": 32555
    },
    {
      "epoch": 7.6153666237866915,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 32556
    },
    {
      "epoch": 7.6156005145597,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 32557
    },
    {
      "epoch": 7.615834405332709,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 32558
    },
    {
      "epoch": 7.616068296105719,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.4524,
      "step": 32559
    },
    {
      "epoch": 7.616302186878728,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 32560
    },
    {
      "epoch": 7.616536077651737,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 32561
    },
    {
      "epoch": 7.616769968424745,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 32562
    },
    {
      "epoch": 7.617003859197755,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5592,
      "step": 32563
    },
    {
      "epoch": 7.617237749970764,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6897,
      "step": 32564
    },
    {
      "epoch": 7.617471640743773,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 32565
    },
    {
      "epoch": 7.617705531516782,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7526,
      "step": 32566
    },
    {
      "epoch": 7.6179394222897905,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7671,
      "step": 32567
    },
    {
      "epoch": 7.618173313062799,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 32568
    },
    {
      "epoch": 7.618407203835809,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8132,
      "step": 32569
    },
    {
      "epoch": 7.618641094608818,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6972,
      "step": 32570
    },
    {
      "epoch": 7.618874985381827,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7093,
      "step": 32571
    },
    {
      "epoch": 7.619108876154836,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7548,
      "step": 32572
    },
    {
      "epoch": 7.619342766927844,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.5212,
      "step": 32573
    },
    {
      "epoch": 7.619576657700854,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 32574
    },
    {
      "epoch": 7.619810548473863,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6558,
      "step": 32575
    },
    {
      "epoch": 7.620044439246872,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 32576
    },
    {
      "epoch": 7.620278330019881,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 32577
    },
    {
      "epoch": 7.6205122207928895,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6505,
      "step": 32578
    },
    {
      "epoch": 7.620746111565898,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 32579
    },
    {
      "epoch": 7.620980002338908,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 32580
    },
    {
      "epoch": 7.621213893111917,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 32581
    },
    {
      "epoch": 7.621447783884926,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.4656,
      "step": 32582
    },
    {
      "epoch": 7.621681674657935,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6746,
      "step": 32583
    },
    {
      "epoch": 7.621915565430943,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6078,
      "step": 32584
    },
    {
      "epoch": 7.622149456203953,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7561,
      "step": 32585
    },
    {
      "epoch": 7.622383346976962,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 32586
    },
    {
      "epoch": 7.622617237749971,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 32587
    },
    {
      "epoch": 7.62285112852298,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.417,
      "step": 32588
    },
    {
      "epoch": 7.6230850192959885,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6239,
      "step": 32589
    },
    {
      "epoch": 7.623318910068997,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8318,
      "step": 32590
    },
    {
      "epoch": 7.623552800842007,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 32591
    },
    {
      "epoch": 7.623786691615016,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 32592
    },
    {
      "epoch": 7.624020582388025,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 32593
    },
    {
      "epoch": 7.6242544731610336,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7442,
      "step": 32594
    },
    {
      "epoch": 7.624488363934043,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 32595
    },
    {
      "epoch": 7.624722254707052,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.55,
      "step": 32596
    },
    {
      "epoch": 7.624956145480061,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5816,
      "step": 32597
    },
    {
      "epoch": 7.62519003625307,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 32598
    },
    {
      "epoch": 7.625423927026079,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 32599
    },
    {
      "epoch": 7.6256578177990875,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.2732,
      "step": 32600
    },
    {
      "epoch": 7.6256578177990875,
      "eval_runtime": 4.6462,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 32600
    },
    {
      "epoch": 7.625891708572097,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7802,
      "step": 32601
    },
    {
      "epoch": 7.626125599345106,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 32602
    },
    {
      "epoch": 7.626359490118115,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 32603
    },
    {
      "epoch": 7.626593380891124,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 32604
    },
    {
      "epoch": 7.6268272716641325,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 32605
    },
    {
      "epoch": 7.627061162437142,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5293,
      "step": 32606
    },
    {
      "epoch": 7.627295053210151,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 32607
    },
    {
      "epoch": 7.62752894398316,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 32608
    },
    {
      "epoch": 7.627762834756169,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 32609
    },
    {
      "epoch": 7.627996725529178,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 32610
    },
    {
      "epoch": 7.6282306163021865,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7064,
      "step": 32611
    },
    {
      "epoch": 7.628464507075196,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7741,
      "step": 32612
    },
    {
      "epoch": 7.628698397848205,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 32613
    },
    {
      "epoch": 7.628932288621214,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6774,
      "step": 32614
    },
    {
      "epoch": 7.629166179394223,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 32615
    },
    {
      "epoch": 7.6294000701672315,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7897,
      "step": 32616
    },
    {
      "epoch": 7.629633960940241,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 32617
    },
    {
      "epoch": 7.62986785171325,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 32618
    },
    {
      "epoch": 7.630101742486259,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5769,
      "step": 32619
    },
    {
      "epoch": 7.630335633259268,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 32620
    },
    {
      "epoch": 7.630569524032277,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6545,
      "step": 32621
    },
    {
      "epoch": 7.6308034148052855,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 32622
    },
    {
      "epoch": 7.631037305578295,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6445,
      "step": 32623
    },
    {
      "epoch": 7.631271196351304,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.5555,
      "step": 32624
    },
    {
      "epoch": 7.631505087124313,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9103,
      "step": 32625
    },
    {
      "epoch": 7.631738977897322,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5129,
      "step": 32626
    },
    {
      "epoch": 7.631972868670331,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 32627
    },
    {
      "epoch": 7.63220675944334,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5909,
      "step": 32628
    },
    {
      "epoch": 7.632440650216349,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 32629
    },
    {
      "epoch": 7.632674540989358,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7206,
      "step": 32630
    },
    {
      "epoch": 7.632908431762367,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5921,
      "step": 32631
    },
    {
      "epoch": 7.633142322535376,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7155,
      "step": 32632
    },
    {
      "epoch": 7.633376213308385,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.886,
      "step": 32633
    },
    {
      "epoch": 7.633610104081394,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.3521,
      "step": 32634
    },
    {
      "epoch": 7.633843994854403,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 32635
    },
    {
      "epoch": 7.634077885627412,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 32636
    },
    {
      "epoch": 7.634311776400421,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6041,
      "step": 32637
    },
    {
      "epoch": 7.63454566717343,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 32638
    },
    {
      "epoch": 7.634779557946439,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5861,
      "step": 32639
    },
    {
      "epoch": 7.635013448719448,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5267,
      "step": 32640
    },
    {
      "epoch": 7.635247339492457,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 32641
    },
    {
      "epoch": 7.635481230265466,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 32642
    },
    {
      "epoch": 7.635715121038475,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 32643
    },
    {
      "epoch": 7.635949011811484,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 32644
    },
    {
      "epoch": 7.636182902584493,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5737,
      "step": 32645
    },
    {
      "epoch": 7.636416793357502,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5266,
      "step": 32646
    },
    {
      "epoch": 7.636650684130511,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 32647
    },
    {
      "epoch": 7.63688457490352,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 32648
    },
    {
      "epoch": 7.637118465676529,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 32649
    },
    {
      "epoch": 7.637352356449538,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 32650
    },
    {
      "epoch": 7.637586247222547,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 32651
    },
    {
      "epoch": 7.637820137995556,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6239,
      "step": 32652
    },
    {
      "epoch": 7.638054028768565,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 32653
    },
    {
      "epoch": 7.638287919541574,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.622,
      "step": 32654
    },
    {
      "epoch": 7.638521810314583,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7089,
      "step": 32655
    },
    {
      "epoch": 7.638755701087592,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7737,
      "step": 32656
    },
    {
      "epoch": 7.638989591860601,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 32657
    },
    {
      "epoch": 7.63922348263361,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7504,
      "step": 32658
    },
    {
      "epoch": 7.63945737340662,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 32659
    },
    {
      "epoch": 7.639691264179628,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 32660
    },
    {
      "epoch": 7.639925154952637,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 32661
    },
    {
      "epoch": 7.640159045725646,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4994,
      "step": 32662
    },
    {
      "epoch": 7.640392936498655,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 32663
    },
    {
      "epoch": 7.640626827271664,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6768,
      "step": 32664
    },
    {
      "epoch": 7.6408607180446735,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 32665
    },
    {
      "epoch": 7.641094608817682,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 32666
    },
    {
      "epoch": 7.641328499590691,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 32667
    },
    {
      "epoch": 7.6415623903637,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 32668
    },
    {
      "epoch": 7.641796281136709,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 32669
    },
    {
      "epoch": 7.642030171909719,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9998,
      "step": 32670
    },
    {
      "epoch": 7.642264062682727,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6654,
      "step": 32671
    },
    {
      "epoch": 7.642497953455736,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 32672
    },
    {
      "epoch": 7.642731844228745,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 32673
    },
    {
      "epoch": 7.642965735001754,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6055,
      "step": 32674
    },
    {
      "epoch": 7.643199625774763,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 32675
    },
    {
      "epoch": 7.6434335165477725,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 32676
    },
    {
      "epoch": 7.643667407320781,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7113,
      "step": 32677
    },
    {
      "epoch": 7.64390129809379,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8382,
      "step": 32678
    },
    {
      "epoch": 7.644135188866799,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 32679
    },
    {
      "epoch": 7.644369079639808,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 32680
    },
    {
      "epoch": 7.6446029704128176,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 32681
    },
    {
      "epoch": 7.644836861185826,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4586,
      "step": 32682
    },
    {
      "epoch": 7.645070751958835,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.689,
      "step": 32683
    },
    {
      "epoch": 7.645304642731844,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5687,
      "step": 32684
    },
    {
      "epoch": 7.645538533504853,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6313,
      "step": 32685
    },
    {
      "epoch": 7.645772424277862,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 32686
    },
    {
      "epoch": 7.6460063150508715,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6806,
      "step": 32687
    },
    {
      "epoch": 7.64624020582388,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 32688
    },
    {
      "epoch": 7.646474096596889,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7982,
      "step": 32689
    },
    {
      "epoch": 7.646707987369898,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6194,
      "step": 32690
    },
    {
      "epoch": 7.646941878142908,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 32691
    },
    {
      "epoch": 7.6471757689159165,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 32692
    },
    {
      "epoch": 7.647409659688925,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 32693
    },
    {
      "epoch": 7.647643550461934,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 32694
    },
    {
      "epoch": 7.647877441234943,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 32695
    },
    {
      "epoch": 7.648111332007952,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7456,
      "step": 32696
    },
    {
      "epoch": 7.648345222780962,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 32697
    },
    {
      "epoch": 7.6485791135539705,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 32698
    },
    {
      "epoch": 7.648813004326979,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.576,
      "step": 32699
    },
    {
      "epoch": 7.649046895099988,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 32700
    },
    {
      "epoch": 7.649046895099988,
      "eval_runtime": 4.7235,
      "eval_samples_per_second": 0.212,
      "eval_steps_per_second": 0.212,
      "step": 32700
    },
    {
      "epoch": 7.649280785872997,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5652,
      "step": 32701
    },
    {
      "epoch": 7.649514676646007,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6331,
      "step": 32702
    },
    {
      "epoch": 7.6497485674190155,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 32703
    },
    {
      "epoch": 7.649982458192024,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5372,
      "step": 32704
    },
    {
      "epoch": 7.650216348965033,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 32705
    },
    {
      "epoch": 7.650450239738042,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 32706
    },
    {
      "epoch": 7.650684130511051,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 32707
    },
    {
      "epoch": 7.650918021284061,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9081,
      "step": 32708
    },
    {
      "epoch": 7.6511519120570695,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5009,
      "step": 32709
    },
    {
      "epoch": 7.651385802830078,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 32710
    },
    {
      "epoch": 7.651619693603087,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 32711
    },
    {
      "epoch": 7.651853584376097,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6903,
      "step": 32712
    },
    {
      "epoch": 7.652087475149106,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9301,
      "step": 32713
    },
    {
      "epoch": 7.6523213659221145,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6104,
      "step": 32714
    },
    {
      "epoch": 7.652555256695123,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 32715
    },
    {
      "epoch": 7.652789147468132,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 32716
    },
    {
      "epoch": 7.653023038241141,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 32717
    },
    {
      "epoch": 7.653256929014151,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 32718
    },
    {
      "epoch": 7.65349081978716,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6471,
      "step": 32719
    },
    {
      "epoch": 7.6537247105601685,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 32720
    },
    {
      "epoch": 7.653958601333177,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4673,
      "step": 32721
    },
    {
      "epoch": 7.654192492106186,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 32722
    },
    {
      "epoch": 7.654426382879196,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 32723
    },
    {
      "epoch": 7.654660273652205,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.492,
      "step": 32724
    },
    {
      "epoch": 7.6548941644252135,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 32725
    },
    {
      "epoch": 7.655128055198222,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 32726
    },
    {
      "epoch": 7.655361945971231,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 32727
    },
    {
      "epoch": 7.65559583674424,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.5757,
      "step": 32728
    },
    {
      "epoch": 7.65582972751725,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 32729
    },
    {
      "epoch": 7.656063618290259,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 32730
    },
    {
      "epoch": 7.6562975090632674,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7076,
      "step": 32731
    },
    {
      "epoch": 7.656531399836276,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5542,
      "step": 32732
    },
    {
      "epoch": 7.656765290609285,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 32733
    },
    {
      "epoch": 7.656999181382295,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 32734
    },
    {
      "epoch": 7.657233072155304,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5753,
      "step": 32735
    },
    {
      "epoch": 7.6574669629283125,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1038,
      "step": 32736
    },
    {
      "epoch": 7.657700853701321,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8177,
      "step": 32737
    },
    {
      "epoch": 7.65793474447433,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 32738
    },
    {
      "epoch": 7.658168635247339,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 32739
    },
    {
      "epoch": 7.658402526020349,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 32740
    },
    {
      "epoch": 7.658636416793358,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 32741
    },
    {
      "epoch": 7.658870307566366,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7624,
      "step": 32742
    },
    {
      "epoch": 7.659104198339375,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6846,
      "step": 32743
    },
    {
      "epoch": 7.659338089112385,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 32744
    },
    {
      "epoch": 7.659571979885394,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 32745
    },
    {
      "epoch": 7.659805870658403,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 32746
    },
    {
      "epoch": 7.6600397614314115,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 32747
    },
    {
      "epoch": 7.66027365220442,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6028,
      "step": 32748
    },
    {
      "epoch": 7.660507542977429,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 32749
    },
    {
      "epoch": 7.660741433750439,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 32750
    },
    {
      "epoch": 7.660975324523448,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6734,
      "step": 32751
    },
    {
      "epoch": 7.661209215296457,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 32752
    },
    {
      "epoch": 7.661443106069465,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8535,
      "step": 32753
    },
    {
      "epoch": 7.661676996842474,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 32754
    },
    {
      "epoch": 7.661910887615484,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5818,
      "step": 32755
    },
    {
      "epoch": 7.662144778388493,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4527,
      "step": 32756
    },
    {
      "epoch": 7.662378669161502,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 32757
    },
    {
      "epoch": 7.6626125599345105,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 32758
    },
    {
      "epoch": 7.662846450707519,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 32759
    },
    {
      "epoch": 7.663080341480528,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6022,
      "step": 32760
    },
    {
      "epoch": 7.663314232253538,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 32761
    },
    {
      "epoch": 7.663548123026547,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 32762
    },
    {
      "epoch": 7.663782013799556,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7173,
      "step": 32763
    },
    {
      "epoch": 7.664015904572564,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5983,
      "step": 32764
    },
    {
      "epoch": 7.664249795345573,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 32765
    },
    {
      "epoch": 7.664483686118583,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.5261,
      "step": 32766
    },
    {
      "epoch": 7.664717576891592,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 32767
    },
    {
      "epoch": 7.664951467664601,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 32768
    },
    {
      "epoch": 7.6651853584376095,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7242,
      "step": 32769
    },
    {
      "epoch": 7.665419249210618,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7997,
      "step": 32770
    },
    {
      "epoch": 7.665653139983627,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 32771
    },
    {
      "epoch": 7.665887030756637,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 32772
    },
    {
      "epoch": 7.666120921529646,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 32773
    },
    {
      "epoch": 7.666354812302655,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 32774
    },
    {
      "epoch": 7.666588703075663,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 32775
    },
    {
      "epoch": 7.666822593848673,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 32776
    },
    {
      "epoch": 7.667056484621682,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6523,
      "step": 32777
    },
    {
      "epoch": 7.667290375394691,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6872,
      "step": 32778
    },
    {
      "epoch": 7.6675242661677,
      "grad_norm": 7.34375,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 32779
    },
    {
      "epoch": 7.6677581569407085,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 32780
    },
    {
      "epoch": 7.667992047713717,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 32781
    },
    {
      "epoch": 7.668225938486727,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.5459,
      "step": 32782
    },
    {
      "epoch": 7.668459829259736,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.7586,
      "step": 32783
    },
    {
      "epoch": 7.668693720032745,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 32784
    },
    {
      "epoch": 7.668927610805754,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5471,
      "step": 32785
    },
    {
      "epoch": 7.669161501578762,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.5858,
      "step": 32786
    },
    {
      "epoch": 7.669395392351772,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 32787
    },
    {
      "epoch": 7.669629283124781,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 32788
    },
    {
      "epoch": 7.66986317389779,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6259,
      "step": 32789
    },
    {
      "epoch": 7.670097064670799,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 32790
    },
    {
      "epoch": 7.6703309554438075,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 32791
    },
    {
      "epoch": 7.670564846216816,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 32792
    },
    {
      "epoch": 7.670798736989826,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6242,
      "step": 32793
    },
    {
      "epoch": 7.671032627762835,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 32794
    },
    {
      "epoch": 7.671266518535844,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 32795
    },
    {
      "epoch": 7.671500409308853,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6345,
      "step": 32796
    },
    {
      "epoch": 7.671734300081861,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 32797
    },
    {
      "epoch": 7.671968190854871,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4939,
      "step": 32798
    },
    {
      "epoch": 7.67220208162788,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 32799
    },
    {
      "epoch": 7.672435972400889,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 32800
    },
    {
      "epoch": 7.672435972400889,
      "eval_runtime": 4.6253,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 32800
    },
    {
      "epoch": 7.672669863173898,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4246,
      "step": 32801
    },
    {
      "epoch": 7.6729037539469065,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 32802
    },
    {
      "epoch": 7.673137644719915,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6397,
      "step": 32803
    },
    {
      "epoch": 7.673371535492925,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 32804
    },
    {
      "epoch": 7.673605426265934,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5676,
      "step": 32805
    },
    {
      "epoch": 7.673839317038943,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 32806
    },
    {
      "epoch": 7.674073207811952,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.2635,
      "step": 32807
    },
    {
      "epoch": 7.674307098584961,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 32808
    },
    {
      "epoch": 7.67454098935797,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 32809
    },
    {
      "epoch": 7.674774880130979,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 32810
    },
    {
      "epoch": 7.675008770903988,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6065,
      "step": 32811
    },
    {
      "epoch": 7.675242661676997,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 32812
    },
    {
      "epoch": 7.6754765524500055,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 32813
    },
    {
      "epoch": 7.675710443223015,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 32814
    },
    {
      "epoch": 7.675944333996024,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 32815
    },
    {
      "epoch": 7.676178224769033,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 32816
    },
    {
      "epoch": 7.676412115542042,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 32817
    },
    {
      "epoch": 7.676646006315051,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 32818
    },
    {
      "epoch": 7.67687989708806,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 32819
    },
    {
      "epoch": 7.677113787861069,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 32820
    },
    {
      "epoch": 7.677347678634078,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 32821
    },
    {
      "epoch": 7.677581569407087,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 32822
    },
    {
      "epoch": 7.677815460180096,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5927,
      "step": 32823
    },
    {
      "epoch": 7.6780493509531045,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7234,
      "step": 32824
    },
    {
      "epoch": 7.678283241726114,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7353,
      "step": 32825
    },
    {
      "epoch": 7.678517132499123,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6365,
      "step": 32826
    },
    {
      "epoch": 7.678751023272132,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 32827
    },
    {
      "epoch": 7.678984914045141,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 32828
    },
    {
      "epoch": 7.67921880481815,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 32829
    },
    {
      "epoch": 7.679452695591159,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 32830
    },
    {
      "epoch": 7.679686586364168,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5272,
      "step": 32831
    },
    {
      "epoch": 7.679920477137177,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6729,
      "step": 32832
    },
    {
      "epoch": 7.680154367910186,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 32833
    },
    {
      "epoch": 7.680388258683195,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 32834
    },
    {
      "epoch": 7.6806221494562035,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 32835
    },
    {
      "epoch": 7.680856040229213,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5588,
      "step": 32836
    },
    {
      "epoch": 7.681089931002222,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 32837
    },
    {
      "epoch": 7.681323821775231,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7116,
      "step": 32838
    },
    {
      "epoch": 7.68155771254824,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 32839
    },
    {
      "epoch": 7.681791603321249,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 32840
    },
    {
      "epoch": 7.682025494094258,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 32841
    },
    {
      "epoch": 7.682259384867267,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 32842
    },
    {
      "epoch": 7.682493275640276,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 32843
    },
    {
      "epoch": 7.682727166413285,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 32844
    },
    {
      "epoch": 7.682961057186294,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 32845
    },
    {
      "epoch": 7.683194947959303,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 32846
    },
    {
      "epoch": 7.683428838732312,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 32847
    },
    {
      "epoch": 7.683662729505321,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 32848
    },
    {
      "epoch": 7.68389662027833,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.643,
      "step": 32849
    },
    {
      "epoch": 7.684130511051339,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6481,
      "step": 32850
    },
    {
      "epoch": 7.684364401824348,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5782,
      "step": 32851
    },
    {
      "epoch": 7.684598292597357,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 32852
    },
    {
      "epoch": 7.684832183370366,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7152,
      "step": 32853
    },
    {
      "epoch": 7.685066074143375,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 32854
    },
    {
      "epoch": 7.685299964916384,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 32855
    },
    {
      "epoch": 7.685533855689393,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 32856
    },
    {
      "epoch": 7.685767746462402,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 32857
    },
    {
      "epoch": 7.686001637235411,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5736,
      "step": 32858
    },
    {
      "epoch": 7.68623552800842,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 32859
    },
    {
      "epoch": 7.686469418781429,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 32860
    },
    {
      "epoch": 7.686703309554438,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5055,
      "step": 32861
    },
    {
      "epoch": 7.686937200327447,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9503,
      "step": 32862
    },
    {
      "epoch": 7.687171091100456,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 32863
    },
    {
      "epoch": 7.687404981873465,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 32864
    },
    {
      "epoch": 7.687638872646474,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5197,
      "step": 32865
    },
    {
      "epoch": 7.687872763419483,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 32866
    },
    {
      "epoch": 7.688106654192492,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 32867
    },
    {
      "epoch": 7.688340544965501,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6768,
      "step": 32868
    },
    {
      "epoch": 7.68857443573851,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 32869
    },
    {
      "epoch": 7.688808326511519,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5299,
      "step": 32870
    },
    {
      "epoch": 7.689042217284528,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6906,
      "step": 32871
    },
    {
      "epoch": 7.689276108057538,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.5109,
      "step": 32872
    },
    {
      "epoch": 7.689509998830546,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 32873
    },
    {
      "epoch": 7.689743889603555,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 32874
    },
    {
      "epoch": 7.689977780376564,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 32875
    },
    {
      "epoch": 7.690211671149573,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 32876
    },
    {
      "epoch": 7.690445561922582,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 32877
    },
    {
      "epoch": 7.6906794526955915,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 32878
    },
    {
      "epoch": 7.6909133434686,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 32879
    },
    {
      "epoch": 7.691147234241609,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 32880
    },
    {
      "epoch": 7.691381125014618,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 32881
    },
    {
      "epoch": 7.691615015787627,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 32882
    },
    {
      "epoch": 7.691848906560637,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7543,
      "step": 32883
    },
    {
      "epoch": 7.692082797333645,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7252,
      "step": 32884
    },
    {
      "epoch": 7.692316688106654,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5534,
      "step": 32885
    },
    {
      "epoch": 7.692550578879663,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 32886
    },
    {
      "epoch": 7.692784469652672,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 32887
    },
    {
      "epoch": 7.693018360425681,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.4737,
      "step": 32888
    },
    {
      "epoch": 7.6932522511986905,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 32889
    },
    {
      "epoch": 7.693486141971699,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5851,
      "step": 32890
    },
    {
      "epoch": 7.693720032744708,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8107,
      "step": 32891
    },
    {
      "epoch": 7.693953923517717,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.678,
      "step": 32892
    },
    {
      "epoch": 7.694187814290726,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 32893
    },
    {
      "epoch": 7.694421705063736,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6122,
      "step": 32894
    },
    {
      "epoch": 7.694655595836744,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5604,
      "step": 32895
    },
    {
      "epoch": 7.694889486609753,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 32896
    },
    {
      "epoch": 7.695123377382762,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 32897
    },
    {
      "epoch": 7.695357268155771,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5915,
      "step": 32898
    },
    {
      "epoch": 7.69559115892878,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.4837,
      "step": 32899
    },
    {
      "epoch": 7.6958250497017895,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 32900
    },
    {
      "epoch": 7.6958250497017895,
      "eval_runtime": 4.5933,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 32900
    },
    {
      "epoch": 7.696058940474798,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 32901
    },
    {
      "epoch": 7.696292831247807,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 32902
    },
    {
      "epoch": 7.696526722020816,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 32903
    },
    {
      "epoch": 7.696760612793826,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 32904
    },
    {
      "epoch": 7.696994503566835,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 32905
    },
    {
      "epoch": 7.697228394339843,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 32906
    },
    {
      "epoch": 7.697462285112852,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.4376,
      "step": 32907
    },
    {
      "epoch": 7.697696175885861,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 32908
    },
    {
      "epoch": 7.69793006665887,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 32909
    },
    {
      "epoch": 7.69816395743188,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 32910
    },
    {
      "epoch": 7.6983978482048885,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 32911
    },
    {
      "epoch": 7.698631738977897,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 32912
    },
    {
      "epoch": 7.698865629750906,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 32913
    },
    {
      "epoch": 7.699099520523915,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 32914
    },
    {
      "epoch": 7.699333411296925,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.5283,
      "step": 32915
    },
    {
      "epoch": 7.699567302069934,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 32916
    },
    {
      "epoch": 7.699801192842942,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 32917
    },
    {
      "epoch": 7.700035083615951,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 32918
    },
    {
      "epoch": 7.70026897438896,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 32919
    },
    {
      "epoch": 7.700502865161969,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 32920
    },
    {
      "epoch": 7.700736755934979,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 32921
    },
    {
      "epoch": 7.7009706467079875,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 32922
    },
    {
      "epoch": 7.701204537480996,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 32923
    },
    {
      "epoch": 7.701438428254005,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5977,
      "step": 32924
    },
    {
      "epoch": 7.701672319027015,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6492,
      "step": 32925
    },
    {
      "epoch": 7.701906209800024,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7204,
      "step": 32926
    },
    {
      "epoch": 7.7021401005730326,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 32927
    },
    {
      "epoch": 7.702373991346041,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 32928
    },
    {
      "epoch": 7.70260788211905,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7207,
      "step": 32929
    },
    {
      "epoch": 7.702841772892059,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 32930
    },
    {
      "epoch": 7.703075663665069,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 32931
    },
    {
      "epoch": 7.703309554438078,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 32932
    },
    {
      "epoch": 7.7035434452110865,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6072,
      "step": 32933
    },
    {
      "epoch": 7.703777335984095,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5238,
      "step": 32934
    },
    {
      "epoch": 7.704011226757104,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 32935
    },
    {
      "epoch": 7.704245117530114,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6591,
      "step": 32936
    },
    {
      "epoch": 7.704479008303123,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 32937
    },
    {
      "epoch": 7.7047128990761315,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 32938
    },
    {
      "epoch": 7.70494678984914,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 32939
    },
    {
      "epoch": 7.705180680622149,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5211,
      "step": 32940
    },
    {
      "epoch": 7.705414571395158,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.611,
      "step": 32941
    },
    {
      "epoch": 7.705648462168168,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.531,
      "step": 32942
    },
    {
      "epoch": 7.705882352941177,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5736,
      "step": 32943
    },
    {
      "epoch": 7.7061162437141855,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 32944
    },
    {
      "epoch": 7.706350134487194,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 32945
    },
    {
      "epoch": 7.706584025260203,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 32946
    },
    {
      "epoch": 7.706817916033213,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 32947
    },
    {
      "epoch": 7.707051806806222,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 32948
    },
    {
      "epoch": 7.7072856975792305,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 32949
    },
    {
      "epoch": 7.707519588352239,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 32950
    },
    {
      "epoch": 7.707753479125248,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5663,
      "step": 32951
    },
    {
      "epoch": 7.707987369898257,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 32952
    },
    {
      "epoch": 7.708221260671267,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5296,
      "step": 32953
    },
    {
      "epoch": 7.708455151444276,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8734,
      "step": 32954
    },
    {
      "epoch": 7.7086890422172845,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 32955
    },
    {
      "epoch": 7.708922932990293,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 32956
    },
    {
      "epoch": 7.709156823763303,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6423,
      "step": 32957
    },
    {
      "epoch": 7.709390714536312,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.626,
      "step": 32958
    },
    {
      "epoch": 7.709624605309321,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8751,
      "step": 32959
    },
    {
      "epoch": 7.7098584960823295,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5384,
      "step": 32960
    },
    {
      "epoch": 7.710092386855338,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8026,
      "step": 32961
    },
    {
      "epoch": 7.710326277628347,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 32962
    },
    {
      "epoch": 7.710560168401357,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 32963
    },
    {
      "epoch": 7.710794059174366,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 32964
    },
    {
      "epoch": 7.711027949947375,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7659,
      "step": 32965
    },
    {
      "epoch": 7.7112618407203835,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 32966
    },
    {
      "epoch": 7.711495731493392,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 32967
    },
    {
      "epoch": 7.711729622266402,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 32968
    },
    {
      "epoch": 7.711963513039411,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 32969
    },
    {
      "epoch": 7.71219740381242,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 32970
    },
    {
      "epoch": 7.7124312945854285,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7448,
      "step": 32971
    },
    {
      "epoch": 7.712665185358437,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.641,
      "step": 32972
    },
    {
      "epoch": 7.712899076131446,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 32973
    },
    {
      "epoch": 7.713132966904456,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 32974
    },
    {
      "epoch": 7.713366857677465,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 32975
    },
    {
      "epoch": 7.713600748450474,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 32976
    },
    {
      "epoch": 7.7138346392234824,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1385,
      "step": 32977
    },
    {
      "epoch": 7.714068529996491,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5934,
      "step": 32978
    },
    {
      "epoch": 7.714302420769501,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 32979
    },
    {
      "epoch": 7.71453631154251,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4818,
      "step": 32980
    },
    {
      "epoch": 7.714770202315519,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4185,
      "step": 32981
    },
    {
      "epoch": 7.7150040930885275,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 32982
    },
    {
      "epoch": 7.715237983861536,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 32983
    },
    {
      "epoch": 7.715471874634545,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 32984
    },
    {
      "epoch": 7.715705765407555,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 32985
    },
    {
      "epoch": 7.715939656180564,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 32986
    },
    {
      "epoch": 7.716173546953573,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 32987
    },
    {
      "epoch": 7.716407437726581,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 32988
    },
    {
      "epoch": 7.716641328499591,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7651,
      "step": 32989
    },
    {
      "epoch": 7.7168752192726,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5103,
      "step": 32990
    },
    {
      "epoch": 7.717109110045609,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 32991
    },
    {
      "epoch": 7.717343000818618,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 32992
    },
    {
      "epoch": 7.7175768915916265,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 32993
    },
    {
      "epoch": 7.717810782364635,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.506,
      "step": 32994
    },
    {
      "epoch": 7.718044673137645,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7561,
      "step": 32995
    },
    {
      "epoch": 7.718278563910654,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 32996
    },
    {
      "epoch": 7.718512454683663,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7215,
      "step": 32997
    },
    {
      "epoch": 7.718746345456672,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 32998
    },
    {
      "epoch": 7.71898023622968,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7114,
      "step": 32999
    },
    {
      "epoch": 7.71921412700269,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 33000
    },
    {
      "epoch": 7.71921412700269,
      "eval_runtime": 4.6006,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 33000
    },
    {
      "epoch": 7.719448017775699,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 33001
    },
    {
      "epoch": 7.719681908548708,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.4497,
      "step": 33002
    },
    {
      "epoch": 7.719915799321717,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 33003
    },
    {
      "epoch": 7.7201496900947255,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 33004
    },
    {
      "epoch": 7.720383580867734,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 33005
    },
    {
      "epoch": 7.720617471640744,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.4631,
      "step": 33006
    },
    {
      "epoch": 7.720851362413753,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5833,
      "step": 33007
    },
    {
      "epoch": 7.721085253186762,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5789,
      "step": 33008
    },
    {
      "epoch": 7.721319143959771,
      "grad_norm": 10.6875,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 33009
    },
    {
      "epoch": 7.721553034732779,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5058,
      "step": 33010
    },
    {
      "epoch": 7.721786925505789,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 33011
    },
    {
      "epoch": 7.722020816278798,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 33012
    },
    {
      "epoch": 7.722254707051807,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 33013
    },
    {
      "epoch": 7.722488597824816,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6756,
      "step": 33014
    },
    {
      "epoch": 7.7227224885978245,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.49,
      "step": 33015
    },
    {
      "epoch": 7.722956379370833,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.8303,
      "step": 33016
    },
    {
      "epoch": 7.723190270143843,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6531,
      "step": 33017
    },
    {
      "epoch": 7.723424160916852,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 33018
    },
    {
      "epoch": 7.723658051689861,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 33019
    },
    {
      "epoch": 7.72389194246287,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 33020
    },
    {
      "epoch": 7.724125833235879,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 33021
    },
    {
      "epoch": 7.724359724008888,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 33022
    },
    {
      "epoch": 7.724593614781897,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 33023
    },
    {
      "epoch": 7.724827505554906,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 33024
    },
    {
      "epoch": 7.725061396327915,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8007,
      "step": 33025
    },
    {
      "epoch": 7.7252952871009235,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5133,
      "step": 33026
    },
    {
      "epoch": 7.725529177873933,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6109,
      "step": 33027
    },
    {
      "epoch": 7.725763068646942,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 33028
    },
    {
      "epoch": 7.725996959419951,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 33029
    },
    {
      "epoch": 7.72623085019296,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6655,
      "step": 33030
    },
    {
      "epoch": 7.726464740965969,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5555,
      "step": 33031
    },
    {
      "epoch": 7.726698631738978,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 33032
    },
    {
      "epoch": 7.726932522511987,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.547,
      "step": 33033
    },
    {
      "epoch": 7.727166413284996,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 33034
    },
    {
      "epoch": 7.727400304058005,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7767,
      "step": 33035
    },
    {
      "epoch": 7.727634194831014,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6867,
      "step": 33036
    },
    {
      "epoch": 7.7278680856040225,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 33037
    },
    {
      "epoch": 7.728101976377032,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 33038
    },
    {
      "epoch": 7.728335867150041,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 33039
    },
    {
      "epoch": 7.72856975792305,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 33040
    },
    {
      "epoch": 7.728803648696059,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 33041
    },
    {
      "epoch": 7.729037539469068,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6975,
      "step": 33042
    },
    {
      "epoch": 7.729271430242077,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7728,
      "step": 33043
    },
    {
      "epoch": 7.729505321015086,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 33044
    },
    {
      "epoch": 7.729739211788095,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 33045
    },
    {
      "epoch": 7.729973102561104,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7398,
      "step": 33046
    },
    {
      "epoch": 7.730206993334113,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6736,
      "step": 33047
    },
    {
      "epoch": 7.7304408841071215,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 33048
    },
    {
      "epoch": 7.730674774880131,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 33049
    },
    {
      "epoch": 7.73090866565314,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4458,
      "step": 33050
    },
    {
      "epoch": 7.731142556426149,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 33051
    },
    {
      "epoch": 7.731376447199158,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 33052
    },
    {
      "epoch": 7.7316103379721675,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 33053
    },
    {
      "epoch": 7.731844228745176,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5964,
      "step": 33054
    },
    {
      "epoch": 7.732078119518185,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6593,
      "step": 33055
    },
    {
      "epoch": 7.732312010291194,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 33056
    },
    {
      "epoch": 7.732545901064203,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 33057
    },
    {
      "epoch": 7.732779791837212,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 33058
    },
    {
      "epoch": 7.733013682610221,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 33059
    },
    {
      "epoch": 7.73324757338323,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.3777,
      "step": 33060
    },
    {
      "epoch": 7.733481464156239,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 33061
    },
    {
      "epoch": 7.733715354929248,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6366,
      "step": 33062
    },
    {
      "epoch": 7.733949245702257,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 33063
    },
    {
      "epoch": 7.7341831364752665,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8945,
      "step": 33064
    },
    {
      "epoch": 7.734417027248275,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 33065
    },
    {
      "epoch": 7.734650918021284,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 33066
    },
    {
      "epoch": 7.734884808794293,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 33067
    },
    {
      "epoch": 7.735118699567302,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6396,
      "step": 33068
    },
    {
      "epoch": 7.735352590340311,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 33069
    },
    {
      "epoch": 7.73558648111332,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.4854,
      "step": 33070
    },
    {
      "epoch": 7.735820371886329,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5975,
      "step": 33071
    },
    {
      "epoch": 7.736054262659338,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5731,
      "step": 33072
    },
    {
      "epoch": 7.736288153432347,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 33073
    },
    {
      "epoch": 7.736522044205356,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4208,
      "step": 33074
    },
    {
      "epoch": 7.7367559349783654,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 33075
    },
    {
      "epoch": 7.736989825751374,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2614,
      "step": 33076
    },
    {
      "epoch": 7.737223716524383,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 33077
    },
    {
      "epoch": 7.737457607297392,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 33078
    },
    {
      "epoch": 7.737691498070401,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5103,
      "step": 33079
    },
    {
      "epoch": 7.73792538884341,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8508,
      "step": 33080
    },
    {
      "epoch": 7.738159279616419,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7104,
      "step": 33081
    },
    {
      "epoch": 7.738393170389428,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 33082
    },
    {
      "epoch": 7.738627061162437,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 33083
    },
    {
      "epoch": 7.738860951935446,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7113,
      "step": 33084
    },
    {
      "epoch": 7.739094842708456,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 33085
    },
    {
      "epoch": 7.739328733481464,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 33086
    },
    {
      "epoch": 7.739562624254473,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 33087
    },
    {
      "epoch": 7.739796515027482,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 33088
    },
    {
      "epoch": 7.740030405800491,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.3982,
      "step": 33089
    },
    {
      "epoch": 7.7402642965735,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6853,
      "step": 33090
    },
    {
      "epoch": 7.7404981873465095,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 33091
    },
    {
      "epoch": 7.740732078119518,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 33092
    },
    {
      "epoch": 7.740965968892527,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 33093
    },
    {
      "epoch": 7.741199859665536,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6606,
      "step": 33094
    },
    {
      "epoch": 7.741433750438545,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 33095
    },
    {
      "epoch": 7.741667641211555,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 33096
    },
    {
      "epoch": 7.741901531984563,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 33097
    },
    {
      "epoch": 7.742135422757572,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 33098
    },
    {
      "epoch": 7.742369313530581,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 33099
    },
    {
      "epoch": 7.74260320430359,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7443,
      "step": 33100
    },
    {
      "epoch": 7.74260320430359,
      "eval_runtime": 4.616,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 33100
    },
    {
      "epoch": 7.742837095076599,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 33101
    },
    {
      "epoch": 7.7430709858496085,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 33102
    },
    {
      "epoch": 7.743304876622617,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 33103
    },
    {
      "epoch": 7.743538767395626,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3304,
      "step": 33104
    },
    {
      "epoch": 7.743772658168635,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4187,
      "step": 33105
    },
    {
      "epoch": 7.744006548941644,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5895,
      "step": 33106
    },
    {
      "epoch": 7.744240439714654,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7017,
      "step": 33107
    },
    {
      "epoch": 7.744474330487662,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 33108
    },
    {
      "epoch": 7.744708221260671,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 33109
    },
    {
      "epoch": 7.74494211203368,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7147,
      "step": 33110
    },
    {
      "epoch": 7.745176002806689,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 33111
    },
    {
      "epoch": 7.745409893579699,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 33112
    },
    {
      "epoch": 7.7456437843527075,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.598,
      "step": 33113
    },
    {
      "epoch": 7.745877675125716,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.3888,
      "step": 33114
    },
    {
      "epoch": 7.746111565898725,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 33115
    },
    {
      "epoch": 7.746345456671734,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6106,
      "step": 33116
    },
    {
      "epoch": 7.746579347444744,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5995,
      "step": 33117
    },
    {
      "epoch": 7.746813238217753,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5934,
      "step": 33118
    },
    {
      "epoch": 7.747047128990761,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 33119
    },
    {
      "epoch": 7.74728101976377,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 33120
    },
    {
      "epoch": 7.747514910536779,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5444,
      "step": 33121
    },
    {
      "epoch": 7.747748801309788,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 33122
    },
    {
      "epoch": 7.747982692082798,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 33123
    },
    {
      "epoch": 7.7482165828558065,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 33124
    },
    {
      "epoch": 7.748450473628815,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 33125
    },
    {
      "epoch": 7.748684364401824,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5607,
      "step": 33126
    },
    {
      "epoch": 7.748918255174833,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 33127
    },
    {
      "epoch": 7.749152145947843,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 33128
    },
    {
      "epoch": 7.749386036720852,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7282,
      "step": 33129
    },
    {
      "epoch": 7.74961992749386,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 33130
    },
    {
      "epoch": 7.749853818266869,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 33131
    },
    {
      "epoch": 7.750087709039878,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5027,
      "step": 33132
    },
    {
      "epoch": 7.750321599812887,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 33133
    },
    {
      "epoch": 7.750555490585897,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 33134
    },
    {
      "epoch": 7.7507893813589055,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 33135
    },
    {
      "epoch": 7.751023272131914,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6467,
      "step": 33136
    },
    {
      "epoch": 7.751257162904923,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 33137
    },
    {
      "epoch": 7.751491053677933,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 33138
    },
    {
      "epoch": 7.751724944450942,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 33139
    },
    {
      "epoch": 7.751958835223951,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 33140
    },
    {
      "epoch": 7.752192725996959,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6259,
      "step": 33141
    },
    {
      "epoch": 7.752426616769968,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 33142
    },
    {
      "epoch": 7.752660507542977,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 33143
    },
    {
      "epoch": 7.752894398315987,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 33144
    },
    {
      "epoch": 7.753128289088996,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6483,
      "step": 33145
    },
    {
      "epoch": 7.7533621798620045,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 33146
    },
    {
      "epoch": 7.753596070635013,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5666,
      "step": 33147
    },
    {
      "epoch": 7.753829961408022,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5579,
      "step": 33148
    },
    {
      "epoch": 7.754063852181032,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7996,
      "step": 33149
    },
    {
      "epoch": 7.754297742954041,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 33150
    },
    {
      "epoch": 7.75453163372705,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 33151
    },
    {
      "epoch": 7.754765524500058,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 33152
    },
    {
      "epoch": 7.754999415273067,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8693,
      "step": 33153
    },
    {
      "epoch": 7.755233306046076,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 33154
    },
    {
      "epoch": 7.755467196819086,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.4068,
      "step": 33155
    },
    {
      "epoch": 7.755701087592095,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 33156
    },
    {
      "epoch": 7.7559349783651035,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.4567,
      "step": 33157
    },
    {
      "epoch": 7.756168869138112,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 33158
    },
    {
      "epoch": 7.756402759911121,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 33159
    },
    {
      "epoch": 7.756636650684131,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 33160
    },
    {
      "epoch": 7.75687054145714,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 33161
    },
    {
      "epoch": 7.757104432230149,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.6124,
      "step": 33162
    },
    {
      "epoch": 7.757338323003157,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 33163
    },
    {
      "epoch": 7.757572213776166,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 33164
    },
    {
      "epoch": 7.757806104549175,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7579,
      "step": 33165
    },
    {
      "epoch": 7.758039995322185,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7367,
      "step": 33166
    },
    {
      "epoch": 7.758273886095194,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 33167
    },
    {
      "epoch": 7.7585077768682025,
      "grad_norm": 7.21875,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 33168
    },
    {
      "epoch": 7.758741667641211,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5635,
      "step": 33169
    },
    {
      "epoch": 7.758975558414221,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 33170
    },
    {
      "epoch": 7.75920944918723,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6653,
      "step": 33171
    },
    {
      "epoch": 7.759443339960239,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 33172
    },
    {
      "epoch": 7.7596772307332476,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5794,
      "step": 33173
    },
    {
      "epoch": 7.759911121506256,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 33174
    },
    {
      "epoch": 7.760145012279265,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 33175
    },
    {
      "epoch": 7.760378903052275,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 33176
    },
    {
      "epoch": 7.760612793825284,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 33177
    },
    {
      "epoch": 7.760846684598293,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6662,
      "step": 33178
    },
    {
      "epoch": 7.7610805753713015,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5453,
      "step": 33179
    },
    {
      "epoch": 7.76131446614431,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 33180
    },
    {
      "epoch": 7.76154835691732,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6904,
      "step": 33181
    },
    {
      "epoch": 7.761782247690329,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 33182
    },
    {
      "epoch": 7.762016138463338,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5675,
      "step": 33183
    },
    {
      "epoch": 7.7622500292363465,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6851,
      "step": 33184
    },
    {
      "epoch": 7.762483920009355,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6775,
      "step": 33185
    },
    {
      "epoch": 7.762717810782364,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.697,
      "step": 33186
    },
    {
      "epoch": 7.762951701555374,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5992,
      "step": 33187
    },
    {
      "epoch": 7.763185592328383,
      "grad_norm": 8.3125,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 33188
    },
    {
      "epoch": 7.763419483101392,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 33189
    },
    {
      "epoch": 7.7636533738744005,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 33190
    },
    {
      "epoch": 7.763887264647409,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 33191
    },
    {
      "epoch": 7.764121155420419,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5534,
      "step": 33192
    },
    {
      "epoch": 7.764355046193428,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6854,
      "step": 33193
    },
    {
      "epoch": 7.764588936966437,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 33194
    },
    {
      "epoch": 7.7648228277394455,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7215,
      "step": 33195
    },
    {
      "epoch": 7.765056718512454,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 33196
    },
    {
      "epoch": 7.765290609285463,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 33197
    },
    {
      "epoch": 7.765524500058473,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 33198
    },
    {
      "epoch": 7.765758390831482,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 33199
    },
    {
      "epoch": 7.765992281604491,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 33200
    },
    {
      "epoch": 7.765992281604491,
      "eval_runtime": 4.5994,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 33200
    },
    {
      "epoch": 7.7662261723774995,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 33201
    },
    {
      "epoch": 7.766460063150509,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 33202
    },
    {
      "epoch": 7.766693953923518,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 33203
    },
    {
      "epoch": 7.766927844696527,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7308,
      "step": 33204
    },
    {
      "epoch": 7.767161735469536,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 33205
    },
    {
      "epoch": 7.7673956262425445,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 33206
    },
    {
      "epoch": 7.767629517015553,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 33207
    },
    {
      "epoch": 7.767863407788563,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 33208
    },
    {
      "epoch": 7.768097298561572,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6409,
      "step": 33209
    },
    {
      "epoch": 7.768331189334581,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 33210
    },
    {
      "epoch": 7.76856508010759,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 33211
    },
    {
      "epoch": 7.7687989708805985,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 33212
    },
    {
      "epoch": 7.769032861653608,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 33213
    },
    {
      "epoch": 7.769266752426617,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 33214
    },
    {
      "epoch": 7.769500643199626,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 33215
    },
    {
      "epoch": 7.769734533972635,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 33216
    },
    {
      "epoch": 7.7699684247456435,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.5552,
      "step": 33217
    },
    {
      "epoch": 7.770202315518652,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6195,
      "step": 33218
    },
    {
      "epoch": 7.770436206291662,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 33219
    },
    {
      "epoch": 7.770670097064671,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 33220
    },
    {
      "epoch": 7.77090398783768,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 33221
    },
    {
      "epoch": 7.771137878610689,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 33222
    },
    {
      "epoch": 7.7713717693836974,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7262,
      "step": 33223
    },
    {
      "epoch": 7.771605660156707,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 33224
    },
    {
      "epoch": 7.771839550929716,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5038,
      "step": 33225
    },
    {
      "epoch": 7.772073441702725,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 33226
    },
    {
      "epoch": 7.772307332475734,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 33227
    },
    {
      "epoch": 7.7725412232487425,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 33228
    },
    {
      "epoch": 7.772775114021751,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.646,
      "step": 33229
    },
    {
      "epoch": 7.773009004794761,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 33230
    },
    {
      "epoch": 7.77324289556777,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 33231
    },
    {
      "epoch": 7.773476786340779,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 33232
    },
    {
      "epoch": 7.773710677113788,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5541,
      "step": 33233
    },
    {
      "epoch": 7.773944567886797,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 33234
    },
    {
      "epoch": 7.774178458659806,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 33235
    },
    {
      "epoch": 7.774412349432815,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 33236
    },
    {
      "epoch": 7.774646240205824,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 33237
    },
    {
      "epoch": 7.774880130978833,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 33238
    },
    {
      "epoch": 7.7751140217518415,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.4586,
      "step": 33239
    },
    {
      "epoch": 7.775347912524851,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7281,
      "step": 33240
    },
    {
      "epoch": 7.77558180329786,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6834,
      "step": 33241
    },
    {
      "epoch": 7.775815694070869,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 33242
    },
    {
      "epoch": 7.776049584843878,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 33243
    },
    {
      "epoch": 7.776283475616887,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6146,
      "step": 33244
    },
    {
      "epoch": 7.776517366389896,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5536,
      "step": 33245
    },
    {
      "epoch": 7.776751257162905,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 33246
    },
    {
      "epoch": 7.776985147935914,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 33247
    },
    {
      "epoch": 7.777219038708923,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 33248
    },
    {
      "epoch": 7.777452929481932,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.776,
      "step": 33249
    },
    {
      "epoch": 7.7776868202549405,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5044,
      "step": 33250
    },
    {
      "epoch": 7.77792071102795,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 33251
    },
    {
      "epoch": 7.778154601800959,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 33252
    },
    {
      "epoch": 7.778388492573968,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 33253
    },
    {
      "epoch": 7.778622383346977,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 33254
    },
    {
      "epoch": 7.778856274119986,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 33255
    },
    {
      "epoch": 7.779090164892995,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8409,
      "step": 33256
    },
    {
      "epoch": 7.779324055666004,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6434,
      "step": 33257
    },
    {
      "epoch": 7.779557946439013,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 33258
    },
    {
      "epoch": 7.779791837212022,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.3638,
      "step": 33259
    },
    {
      "epoch": 7.780025727985031,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6089,
      "step": 33260
    },
    {
      "epoch": 7.7802596187580395,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 33261
    },
    {
      "epoch": 7.780493509531049,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 33262
    },
    {
      "epoch": 7.780727400304058,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9758,
      "step": 33263
    },
    {
      "epoch": 7.780961291077067,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7414,
      "step": 33264
    },
    {
      "epoch": 7.781195181850076,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2022,
      "step": 33265
    },
    {
      "epoch": 7.7814290726230855,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7099,
      "step": 33266
    },
    {
      "epoch": 7.781662963396094,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 33267
    },
    {
      "epoch": 7.781896854169103,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 33268
    },
    {
      "epoch": 7.782130744942112,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 33269
    },
    {
      "epoch": 7.782364635715121,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.667,
      "step": 33270
    },
    {
      "epoch": 7.78259852648813,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6562,
      "step": 33271
    },
    {
      "epoch": 7.782832417261139,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.4164,
      "step": 33272
    },
    {
      "epoch": 7.783066308034148,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 33273
    },
    {
      "epoch": 7.783300198807157,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7356,
      "step": 33274
    },
    {
      "epoch": 7.783534089580166,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.688,
      "step": 33275
    },
    {
      "epoch": 7.783767980353175,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5992,
      "step": 33276
    },
    {
      "epoch": 7.7840018711261845,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5267,
      "step": 33277
    },
    {
      "epoch": 7.784235761899193,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5465,
      "step": 33278
    },
    {
      "epoch": 7.784469652672202,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 33279
    },
    {
      "epoch": 7.784703543445211,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6805,
      "step": 33280
    },
    {
      "epoch": 7.78493743421822,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 33281
    },
    {
      "epoch": 7.785171324991229,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 33282
    },
    {
      "epoch": 7.785405215764238,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8708,
      "step": 33283
    },
    {
      "epoch": 7.785639106537247,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 33284
    },
    {
      "epoch": 7.785872997310256,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 33285
    },
    {
      "epoch": 7.786106888083265,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7219,
      "step": 33286
    },
    {
      "epoch": 7.786340778856274,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 33287
    },
    {
      "epoch": 7.7865746696292835,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.706,
      "step": 33288
    },
    {
      "epoch": 7.786808560402292,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6199,
      "step": 33289
    },
    {
      "epoch": 7.787042451175301,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.7656,
      "step": 33290
    },
    {
      "epoch": 7.78727634194831,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 33291
    },
    {
      "epoch": 7.787510232721319,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 33292
    },
    {
      "epoch": 7.787744123494328,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 33293
    },
    {
      "epoch": 7.787978014267337,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 33294
    },
    {
      "epoch": 7.788211905040346,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.6137,
      "step": 33295
    },
    {
      "epoch": 7.788445795813355,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 33296
    },
    {
      "epoch": 7.788679686586364,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 33297
    },
    {
      "epoch": 7.788913577359374,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 33298
    },
    {
      "epoch": 7.7891474681323825,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 33299
    },
    {
      "epoch": 7.789381358905391,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 33300
    },
    {
      "epoch": 7.789381358905391,
      "eval_runtime": 4.6759,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 33300
    },
    {
      "epoch": 7.7896152496784,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6734,
      "step": 33301
    },
    {
      "epoch": 7.789849140451409,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6459,
      "step": 33302
    },
    {
      "epoch": 7.790083031224418,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.114,
      "step": 33303
    },
    {
      "epoch": 7.7903169219974275,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 33304
    },
    {
      "epoch": 7.790550812770436,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 33305
    },
    {
      "epoch": 7.790784703543445,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 33306
    },
    {
      "epoch": 7.791018594316454,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 33307
    },
    {
      "epoch": 7.791252485089463,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.3012,
      "step": 33308
    },
    {
      "epoch": 7.791486375862473,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 33309
    },
    {
      "epoch": 7.7917202666354815,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 33310
    },
    {
      "epoch": 7.79195415740849,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 33311
    },
    {
      "epoch": 7.792188048181499,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6772,
      "step": 33312
    },
    {
      "epoch": 7.792421938954508,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6436,
      "step": 33313
    },
    {
      "epoch": 7.792655829727517,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6898,
      "step": 33314
    },
    {
      "epoch": 7.7928897205005265,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 33315
    },
    {
      "epoch": 7.793123611273535,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 33316
    },
    {
      "epoch": 7.793357502046544,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 33317
    },
    {
      "epoch": 7.793591392819553,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.585,
      "step": 33318
    },
    {
      "epoch": 7.793825283592563,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8466,
      "step": 33319
    },
    {
      "epoch": 7.794059174365572,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 33320
    },
    {
      "epoch": 7.7942930651385804,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6823,
      "step": 33321
    },
    {
      "epoch": 7.794526955911589,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 33322
    },
    {
      "epoch": 7.794760846684598,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 33323
    },
    {
      "epoch": 7.794994737457607,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 33324
    },
    {
      "epoch": 7.795228628230617,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6882,
      "step": 33325
    },
    {
      "epoch": 7.7954625190036255,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6407,
      "step": 33326
    },
    {
      "epoch": 7.795696409776634,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 33327
    },
    {
      "epoch": 7.795930300549643,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 33328
    },
    {
      "epoch": 7.796164191322652,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 33329
    },
    {
      "epoch": 7.796398082095662,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7667,
      "step": 33330
    },
    {
      "epoch": 7.796631972868671,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 33331
    },
    {
      "epoch": 7.796865863641679,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 33332
    },
    {
      "epoch": 7.797099754414688,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5803,
      "step": 33333
    },
    {
      "epoch": 7.797333645187697,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 33334
    },
    {
      "epoch": 7.797567535960706,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6599,
      "step": 33335
    },
    {
      "epoch": 7.797801426733716,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 33336
    },
    {
      "epoch": 7.7980353175067245,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6544,
      "step": 33337
    },
    {
      "epoch": 7.798269208279733,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 33338
    },
    {
      "epoch": 7.798503099052742,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 33339
    },
    {
      "epoch": 7.798736989825751,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 33340
    },
    {
      "epoch": 7.798970880598761,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 33341
    },
    {
      "epoch": 7.79920477137177,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 33342
    },
    {
      "epoch": 7.799438662144778,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6384,
      "step": 33343
    },
    {
      "epoch": 7.799672552917787,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 33344
    },
    {
      "epoch": 7.799906443690796,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6902,
      "step": 33345
    },
    {
      "epoch": 7.800140334463805,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 33346
    },
    {
      "epoch": 7.800374225236815,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7061,
      "step": 33347
    },
    {
      "epoch": 7.8006081160098235,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.675,
      "step": 33348
    },
    {
      "epoch": 7.800842006782832,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 33349
    },
    {
      "epoch": 7.801075897555841,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 33350
    },
    {
      "epoch": 7.801309788328851,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 33351
    },
    {
      "epoch": 7.80154367910186,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 33352
    },
    {
      "epoch": 7.801777569874869,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 33353
    },
    {
      "epoch": 7.802011460647877,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 33354
    },
    {
      "epoch": 7.802245351420886,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6485,
      "step": 33355
    },
    {
      "epoch": 7.802479242193895,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 33356
    },
    {
      "epoch": 7.802713132966905,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 33357
    },
    {
      "epoch": 7.802947023739914,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6742,
      "step": 33358
    },
    {
      "epoch": 7.8031809145129225,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 33359
    },
    {
      "epoch": 7.803414805285931,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5272,
      "step": 33360
    },
    {
      "epoch": 7.80364869605894,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 33361
    },
    {
      "epoch": 7.80388258683195,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.523,
      "step": 33362
    },
    {
      "epoch": 7.804116477604959,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 33363
    },
    {
      "epoch": 7.804350368377968,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5717,
      "step": 33364
    },
    {
      "epoch": 7.804584259150976,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 33365
    },
    {
      "epoch": 7.804818149923985,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 33366
    },
    {
      "epoch": 7.805052040696994,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 33367
    },
    {
      "epoch": 7.805285931470004,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 33368
    },
    {
      "epoch": 7.805519822243013,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 33369
    },
    {
      "epoch": 7.8057537130160215,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5783,
      "step": 33370
    },
    {
      "epoch": 7.80598760378903,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7184,
      "step": 33371
    },
    {
      "epoch": 7.806221494562039,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 33372
    },
    {
      "epoch": 7.806455385335049,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 33373
    },
    {
      "epoch": 7.806689276108058,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 33374
    },
    {
      "epoch": 7.806923166881067,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6647,
      "step": 33375
    },
    {
      "epoch": 7.807157057654075,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5478,
      "step": 33376
    },
    {
      "epoch": 7.807390948427084,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 33377
    },
    {
      "epoch": 7.807624839200093,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 33378
    },
    {
      "epoch": 7.807858729973103,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 33379
    },
    {
      "epoch": 7.808092620746112,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 33380
    },
    {
      "epoch": 7.8083265115191205,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 33381
    },
    {
      "epoch": 7.808560402292129,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5256,
      "step": 33382
    },
    {
      "epoch": 7.808794293065139,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 33383
    },
    {
      "epoch": 7.809028183838148,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 33384
    },
    {
      "epoch": 7.809262074611157,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 33385
    },
    {
      "epoch": 7.809495965384166,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6788,
      "step": 33386
    },
    {
      "epoch": 7.809729856157174,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 33387
    },
    {
      "epoch": 7.809963746930183,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 33388
    },
    {
      "epoch": 7.810197637703193,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7335,
      "step": 33389
    },
    {
      "epoch": 7.810431528476202,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 33390
    },
    {
      "epoch": 7.810665419249211,
      "grad_norm": 7.25,
      "learning_rate": 3e-05,
      "loss": 2.4606,
      "step": 33391
    },
    {
      "epoch": 7.8108993100222195,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 33392
    },
    {
      "epoch": 7.811133200795228,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 33393
    },
    {
      "epoch": 7.811367091568238,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6793,
      "step": 33394
    },
    {
      "epoch": 7.811600982341247,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 33395
    },
    {
      "epoch": 7.811834873114256,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 33396
    },
    {
      "epoch": 7.812068763887265,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.5618,
      "step": 33397
    },
    {
      "epoch": 7.812302654660273,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6856,
      "step": 33398
    },
    {
      "epoch": 7.812536545433282,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5693,
      "step": 33399
    },
    {
      "epoch": 7.812770436206292,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 33400
    },
    {
      "epoch": 7.812770436206292,
      "eval_runtime": 4.6191,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 33400
    },
    {
      "epoch": 7.813004326979301,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7422,
      "step": 33401
    },
    {
      "epoch": 7.81323821775231,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 33402
    },
    {
      "epoch": 7.8134721085253185,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 33403
    },
    {
      "epoch": 7.813705999298327,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 33404
    },
    {
      "epoch": 7.813939890071337,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 33405
    },
    {
      "epoch": 7.814173780844346,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.5541,
      "step": 33406
    },
    {
      "epoch": 7.814407671617355,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 33407
    },
    {
      "epoch": 7.814641562390364,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 33408
    },
    {
      "epoch": 7.814875453163372,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 33409
    },
    {
      "epoch": 7.815109343936381,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 33410
    },
    {
      "epoch": 7.815343234709391,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 33411
    },
    {
      "epoch": 7.8155771254824,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 33412
    },
    {
      "epoch": 7.815811016255409,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 33413
    },
    {
      "epoch": 7.8160449070284175,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 33414
    },
    {
      "epoch": 7.816278797801427,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 33415
    },
    {
      "epoch": 7.816512688574436,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 33416
    },
    {
      "epoch": 7.816746579347445,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 33417
    },
    {
      "epoch": 7.816980470120454,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 33418
    },
    {
      "epoch": 7.8172143608934626,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 33419
    },
    {
      "epoch": 7.817448251666471,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 33420
    },
    {
      "epoch": 7.817682142439481,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 33421
    },
    {
      "epoch": 7.81791603321249,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 33422
    },
    {
      "epoch": 7.818149923985499,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9442,
      "step": 33423
    },
    {
      "epoch": 7.818383814758508,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6348,
      "step": 33424
    },
    {
      "epoch": 7.8186177055315165,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5943,
      "step": 33425
    },
    {
      "epoch": 7.818851596304526,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 33426
    },
    {
      "epoch": 7.819085487077535,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 33427
    },
    {
      "epoch": 7.819319377850544,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 33428
    },
    {
      "epoch": 7.819553268623553,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5843,
      "step": 33429
    },
    {
      "epoch": 7.8197871593965615,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3588,
      "step": 33430
    },
    {
      "epoch": 7.82002105016957,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 33431
    },
    {
      "epoch": 7.82025494094258,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 33432
    },
    {
      "epoch": 7.820488831715589,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 33433
    },
    {
      "epoch": 7.820722722488598,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 33434
    },
    {
      "epoch": 7.820956613261607,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7281,
      "step": 33435
    },
    {
      "epoch": 7.8211905040346155,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6297,
      "step": 33436
    },
    {
      "epoch": 7.821424394807625,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 33437
    },
    {
      "epoch": 7.821658285580634,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 33438
    },
    {
      "epoch": 7.821892176353643,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 33439
    },
    {
      "epoch": 7.822126067126652,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 33440
    },
    {
      "epoch": 7.8223599578996605,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.5786,
      "step": 33441
    },
    {
      "epoch": 7.822593848672669,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 33442
    },
    {
      "epoch": 7.822827739445679,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9255,
      "step": 33443
    },
    {
      "epoch": 7.823061630218688,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7617,
      "step": 33444
    },
    {
      "epoch": 7.823295520991697,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 33445
    },
    {
      "epoch": 7.823529411764706,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 33446
    },
    {
      "epoch": 7.823763302537715,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7568,
      "step": 33447
    },
    {
      "epoch": 7.823997193310724,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6226,
      "step": 33448
    },
    {
      "epoch": 7.824231084083733,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 33449
    },
    {
      "epoch": 7.824464974856742,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 33450
    },
    {
      "epoch": 7.824698865629751,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6902,
      "step": 33451
    },
    {
      "epoch": 7.8249327564027595,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7708,
      "step": 33452
    },
    {
      "epoch": 7.825166647175769,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6691,
      "step": 33453
    },
    {
      "epoch": 7.825400537948778,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 33454
    },
    {
      "epoch": 7.825634428721787,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 33455
    },
    {
      "epoch": 7.825868319494796,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7017,
      "step": 33456
    },
    {
      "epoch": 7.826102210267805,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 33457
    },
    {
      "epoch": 7.826336101040814,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 33458
    },
    {
      "epoch": 7.826569991813823,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 33459
    },
    {
      "epoch": 7.826803882586832,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 33460
    },
    {
      "epoch": 7.827037773359841,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6153,
      "step": 33461
    },
    {
      "epoch": 7.82727166413285,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6831,
      "step": 33462
    },
    {
      "epoch": 7.8275055549058585,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7239,
      "step": 33463
    },
    {
      "epoch": 7.827739445678868,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7093,
      "step": 33464
    },
    {
      "epoch": 7.827973336451877,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6837,
      "step": 33465
    },
    {
      "epoch": 7.828207227224886,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 33466
    },
    {
      "epoch": 7.828441117997895,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7332,
      "step": 33467
    },
    {
      "epoch": 7.828675008770904,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 33468
    },
    {
      "epoch": 7.828908899543913,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 33469
    },
    {
      "epoch": 7.829142790316922,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.566,
      "step": 33470
    },
    {
      "epoch": 7.829376681089931,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6446,
      "step": 33471
    },
    {
      "epoch": 7.82961057186294,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 33472
    },
    {
      "epoch": 7.829844462635949,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 33473
    },
    {
      "epoch": 7.8300783534089575,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5456,
      "step": 33474
    },
    {
      "epoch": 7.830312244181967,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 33475
    },
    {
      "epoch": 7.830546134954976,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 33476
    },
    {
      "epoch": 7.830780025727985,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 33477
    },
    {
      "epoch": 7.831013916500994,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7059,
      "step": 33478
    },
    {
      "epoch": 7.8312478072740035,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.746,
      "step": 33479
    },
    {
      "epoch": 7.831481698047012,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 33480
    },
    {
      "epoch": 7.831715588820021,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 33481
    },
    {
      "epoch": 7.83194947959303,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 33482
    },
    {
      "epoch": 7.832183370366039,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 33483
    },
    {
      "epoch": 7.832417261139048,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 33484
    },
    {
      "epoch": 7.832651151912057,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 33485
    },
    {
      "epoch": 7.832885042685066,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6478,
      "step": 33486
    },
    {
      "epoch": 7.833118933458075,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5815,
      "step": 33487
    },
    {
      "epoch": 7.833352824231084,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 33488
    },
    {
      "epoch": 7.833586715004093,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 33489
    },
    {
      "epoch": 7.8338206057771025,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 33490
    },
    {
      "epoch": 7.834054496550111,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 33491
    },
    {
      "epoch": 7.83428838732312,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 33492
    },
    {
      "epoch": 7.834522278096129,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 33493
    },
    {
      "epoch": 7.834756168869138,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 33494
    },
    {
      "epoch": 7.834990059642147,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 33495
    },
    {
      "epoch": 7.835223950415156,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 33496
    },
    {
      "epoch": 7.835457841188165,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 33497
    },
    {
      "epoch": 7.835691731961174,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 33498
    },
    {
      "epoch": 7.835925622734183,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6145,
      "step": 33499
    },
    {
      "epoch": 7.836159513507192,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7277,
      "step": 33500
    },
    {
      "epoch": 7.836159513507192,
      "eval_runtime": 4.5802,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 33500
    },
    {
      "epoch": 7.8363934042802015,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 33501
    },
    {
      "epoch": 7.83662729505321,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5658,
      "step": 33502
    },
    {
      "epoch": 7.836861185826219,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6622,
      "step": 33503
    },
    {
      "epoch": 7.837095076599228,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.4795,
      "step": 33504
    },
    {
      "epoch": 7.837328967372237,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 33505
    },
    {
      "epoch": 7.837562858145247,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 33506
    },
    {
      "epoch": 7.837796748918255,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4909,
      "step": 33507
    },
    {
      "epoch": 7.838030639691264,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5366,
      "step": 33508
    },
    {
      "epoch": 7.838264530464273,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 33509
    },
    {
      "epoch": 7.838498421237282,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 33510
    },
    {
      "epoch": 7.838732312010292,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 33511
    },
    {
      "epoch": 7.8389662027833005,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 33512
    },
    {
      "epoch": 7.839200093556309,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7413,
      "step": 33513
    },
    {
      "epoch": 7.839433984329318,
      "grad_norm": 10.25,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 33514
    },
    {
      "epoch": 7.839667875102327,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8176,
      "step": 33515
    },
    {
      "epoch": 7.839901765875336,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 33516
    },
    {
      "epoch": 7.8401356566483456,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.513,
      "step": 33517
    },
    {
      "epoch": 7.840369547421354,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 33518
    },
    {
      "epoch": 7.840603438194363,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 33519
    },
    {
      "epoch": 7.840837328967372,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4535,
      "step": 33520
    },
    {
      "epoch": 7.841071219740381,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 33521
    },
    {
      "epoch": 7.841305110513391,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 33522
    },
    {
      "epoch": 7.8415390012863995,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.742,
      "step": 33523
    },
    {
      "epoch": 7.841772892059408,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 33524
    },
    {
      "epoch": 7.842006782832417,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 33525
    },
    {
      "epoch": 7.842240673605426,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 33526
    },
    {
      "epoch": 7.842474564378435,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 33527
    },
    {
      "epoch": 7.8427084551514445,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6346,
      "step": 33528
    },
    {
      "epoch": 7.842942345924453,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7191,
      "step": 33529
    },
    {
      "epoch": 7.843176236697462,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 33530
    },
    {
      "epoch": 7.843410127470471,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 33531
    },
    {
      "epoch": 7.843644018243481,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 33532
    },
    {
      "epoch": 7.84387790901649,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 33533
    },
    {
      "epoch": 7.8441117997894985,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5717,
      "step": 33534
    },
    {
      "epoch": 7.844345690562507,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 33535
    },
    {
      "epoch": 7.844579581335516,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.3713,
      "step": 33536
    },
    {
      "epoch": 7.844813472108525,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 33537
    },
    {
      "epoch": 7.845047362881535,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 33538
    },
    {
      "epoch": 7.8452812536545435,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2128,
      "step": 33539
    },
    {
      "epoch": 7.845515144427552,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6869,
      "step": 33540
    },
    {
      "epoch": 7.845749035200561,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 33541
    },
    {
      "epoch": 7.84598292597357,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 33542
    },
    {
      "epoch": 7.84621681674658,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5898,
      "step": 33543
    },
    {
      "epoch": 7.846450707519589,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6998,
      "step": 33544
    },
    {
      "epoch": 7.8466845982925975,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.3685,
      "step": 33545
    },
    {
      "epoch": 7.846918489065606,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5091,
      "step": 33546
    },
    {
      "epoch": 7.847152379838615,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 33547
    },
    {
      "epoch": 7.847386270611624,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 33548
    },
    {
      "epoch": 7.847620161384634,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6128,
      "step": 33549
    },
    {
      "epoch": 7.8478540521576425,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5292,
      "step": 33550
    },
    {
      "epoch": 7.848087942930651,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 33551
    },
    {
      "epoch": 7.84832183370366,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6399,
      "step": 33552
    },
    {
      "epoch": 7.848555724476669,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8677,
      "step": 33553
    },
    {
      "epoch": 7.848789615249679,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 33554
    },
    {
      "epoch": 7.849023506022688,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.441,
      "step": 33555
    },
    {
      "epoch": 7.8492573967956965,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 33556
    },
    {
      "epoch": 7.849491287568705,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 33557
    },
    {
      "epoch": 7.849725178341714,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 33558
    },
    {
      "epoch": 7.849959069114723,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6148,
      "step": 33559
    },
    {
      "epoch": 7.850192959887733,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 33560
    },
    {
      "epoch": 7.8504268506607415,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7073,
      "step": 33561
    },
    {
      "epoch": 7.85066074143375,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 33562
    },
    {
      "epoch": 7.850894632206759,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 33563
    },
    {
      "epoch": 7.851128522979769,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 33564
    },
    {
      "epoch": 7.851362413752778,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7511,
      "step": 33565
    },
    {
      "epoch": 7.851596304525787,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 33566
    },
    {
      "epoch": 7.8518301952987954,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 33567
    },
    {
      "epoch": 7.852064086071804,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6186,
      "step": 33568
    },
    {
      "epoch": 7.852297976844813,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 33569
    },
    {
      "epoch": 7.852531867617823,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3763,
      "step": 33570
    },
    {
      "epoch": 7.852765758390832,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.3316,
      "step": 33571
    },
    {
      "epoch": 7.8529996491638405,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 33572
    },
    {
      "epoch": 7.853233539936849,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 33573
    },
    {
      "epoch": 7.853467430709858,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 33574
    },
    {
      "epoch": 7.853701321482868,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 33575
    },
    {
      "epoch": 7.853935212255877,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6918,
      "step": 33576
    },
    {
      "epoch": 7.854169103028886,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 33577
    },
    {
      "epoch": 7.854402993801894,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 33578
    },
    {
      "epoch": 7.854636884574903,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 33579
    },
    {
      "epoch": 7.854870775347912,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.618,
      "step": 33580
    },
    {
      "epoch": 7.855104666120922,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 33581
    },
    {
      "epoch": 7.855338556893931,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 33582
    },
    {
      "epoch": 7.8555724476669395,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 33583
    },
    {
      "epoch": 7.855806338439948,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.6574,
      "step": 33584
    },
    {
      "epoch": 7.856040229212957,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7086,
      "step": 33585
    },
    {
      "epoch": 7.856274119985967,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.611,
      "step": 33586
    },
    {
      "epoch": 7.856508010758976,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 33587
    },
    {
      "epoch": 7.856741901531985,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 33588
    },
    {
      "epoch": 7.856975792304993,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 33589
    },
    {
      "epoch": 7.857209683078002,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 33590
    },
    {
      "epoch": 7.857443573851011,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 33591
    },
    {
      "epoch": 7.857677464624021,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5931,
      "step": 33592
    },
    {
      "epoch": 7.85791135539703,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5275,
      "step": 33593
    },
    {
      "epoch": 7.8581452461700385,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5324,
      "step": 33594
    },
    {
      "epoch": 7.858379136943047,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 33595
    },
    {
      "epoch": 7.858613027716057,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7292,
      "step": 33596
    },
    {
      "epoch": 7.858846918489066,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 33597
    },
    {
      "epoch": 7.859080809262075,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 33598
    },
    {
      "epoch": 7.859314700035084,
      "grad_norm": 13.0625,
      "learning_rate": 3e-05,
      "loss": 1.6921,
      "step": 33599
    },
    {
      "epoch": 7.859548590808092,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 33600
    },
    {
      "epoch": 7.859548590808092,
      "eval_runtime": 4.6682,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 33600
    },
    {
      "epoch": 7.859782481581101,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 33601
    },
    {
      "epoch": 7.860016372354111,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6359,
      "step": 33602
    },
    {
      "epoch": 7.86025026312712,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8897,
      "step": 33603
    },
    {
      "epoch": 7.860484153900129,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 33604
    },
    {
      "epoch": 7.8607180446731375,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 33605
    },
    {
      "epoch": 7.860951935446146,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5723,
      "step": 33606
    },
    {
      "epoch": 7.861185826219156,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 33607
    },
    {
      "epoch": 7.861419716992165,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7123,
      "step": 33608
    },
    {
      "epoch": 7.861653607765174,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 33609
    },
    {
      "epoch": 7.861887498538183,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 33610
    },
    {
      "epoch": 7.862121389311191,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 33611
    },
    {
      "epoch": 7.8623552800842,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.4887,
      "step": 33612
    },
    {
      "epoch": 7.86258917085721,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8776,
      "step": 33613
    },
    {
      "epoch": 7.862823061630219,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5705,
      "step": 33614
    },
    {
      "epoch": 7.863056952403228,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6232,
      "step": 33615
    },
    {
      "epoch": 7.8632908431762365,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5387,
      "step": 33616
    },
    {
      "epoch": 7.863524733949245,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5628,
      "step": 33617
    },
    {
      "epoch": 7.863758624722255,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 33618
    },
    {
      "epoch": 7.863992515495264,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7082,
      "step": 33619
    },
    {
      "epoch": 7.864226406268273,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6356,
      "step": 33620
    },
    {
      "epoch": 7.864460297041282,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 33621
    },
    {
      "epoch": 7.86469418781429,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 33622
    },
    {
      "epoch": 7.864928078587299,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5409,
      "step": 33623
    },
    {
      "epoch": 7.865161969360309,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.4812,
      "step": 33624
    },
    {
      "epoch": 7.865395860133318,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 33625
    },
    {
      "epoch": 7.865629750906327,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 33626
    },
    {
      "epoch": 7.8658636416793355,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 33627
    },
    {
      "epoch": 7.866097532452345,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6586,
      "step": 33628
    },
    {
      "epoch": 7.866331423225354,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.719,
      "step": 33629
    },
    {
      "epoch": 7.866565313998363,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 33630
    },
    {
      "epoch": 7.866799204771372,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 33631
    },
    {
      "epoch": 7.867033095544381,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 33632
    },
    {
      "epoch": 7.867266986317389,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 33633
    },
    {
      "epoch": 7.867500877090399,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 33634
    },
    {
      "epoch": 7.867734767863408,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 33635
    },
    {
      "epoch": 7.867968658636417,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.652,
      "step": 33636
    },
    {
      "epoch": 7.868202549409426,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 33637
    },
    {
      "epoch": 7.8684364401824345,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 33638
    },
    {
      "epoch": 7.868670330955444,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 33639
    },
    {
      "epoch": 7.868904221728453,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 33640
    },
    {
      "epoch": 7.869138112501462,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 33641
    },
    {
      "epoch": 7.869372003274471,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.427,
      "step": 33642
    },
    {
      "epoch": 7.86960589404748,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6805,
      "step": 33643
    },
    {
      "epoch": 7.869839784820488,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 33644
    },
    {
      "epoch": 7.870073675593498,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 33645
    },
    {
      "epoch": 7.870307566366507,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 33646
    },
    {
      "epoch": 7.870541457139516,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 33647
    },
    {
      "epoch": 7.870775347912525,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6171,
      "step": 33648
    },
    {
      "epoch": 7.8710092386855335,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 33649
    },
    {
      "epoch": 7.871243129458543,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 33650
    },
    {
      "epoch": 7.871477020231552,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7253,
      "step": 33651
    },
    {
      "epoch": 7.871710911004561,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 33652
    },
    {
      "epoch": 7.87194480177757,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5896,
      "step": 33653
    },
    {
      "epoch": 7.872178692550579,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 33654
    },
    {
      "epoch": 7.872412583323587,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6274,
      "step": 33655
    },
    {
      "epoch": 7.872646474096597,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.588,
      "step": 33656
    },
    {
      "epoch": 7.872880364869606,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 33657
    },
    {
      "epoch": 7.873114255642615,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 33658
    },
    {
      "epoch": 7.873348146415624,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6568,
      "step": 33659
    },
    {
      "epoch": 7.873582037188633,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7908,
      "step": 33660
    },
    {
      "epoch": 7.873815927961642,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6574,
      "step": 33661
    },
    {
      "epoch": 7.874049818734651,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7253,
      "step": 33662
    },
    {
      "epoch": 7.87428370950766,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6512,
      "step": 33663
    },
    {
      "epoch": 7.874517600280669,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 33664
    },
    {
      "epoch": 7.8747514910536776,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 33665
    },
    {
      "epoch": 7.874985381826687,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 33666
    },
    {
      "epoch": 7.875219272599696,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5571,
      "step": 33667
    },
    {
      "epoch": 7.875453163372705,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6425,
      "step": 33668
    },
    {
      "epoch": 7.875687054145714,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 33669
    },
    {
      "epoch": 7.875920944918723,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 33670
    },
    {
      "epoch": 7.876154835691732,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 33671
    },
    {
      "epoch": 7.876388726464741,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7007,
      "step": 33672
    },
    {
      "epoch": 7.87662261723775,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 33673
    },
    {
      "epoch": 7.876856508010759,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 33674
    },
    {
      "epoch": 7.877090398783768,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7358,
      "step": 33675
    },
    {
      "epoch": 7.8773242895567765,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.5748,
      "step": 33676
    },
    {
      "epoch": 7.877558180329786,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.651,
      "step": 33677
    },
    {
      "epoch": 7.877792071102795,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 33678
    },
    {
      "epoch": 7.878025961875804,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 33679
    },
    {
      "epoch": 7.878259852648813,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 33680
    },
    {
      "epoch": 7.878493743421822,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 33681
    },
    {
      "epoch": 7.878727634194831,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 33682
    },
    {
      "epoch": 7.87896152496784,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6913,
      "step": 33683
    },
    {
      "epoch": 7.879195415740849,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 33684
    },
    {
      "epoch": 7.879429306513858,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 33685
    },
    {
      "epoch": 7.879663197286867,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5875,
      "step": 33686
    },
    {
      "epoch": 7.8798970880598755,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 33687
    },
    {
      "epoch": 7.880130978832885,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 33688
    },
    {
      "epoch": 7.880364869605894,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 33689
    },
    {
      "epoch": 7.880598760378903,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 33690
    },
    {
      "epoch": 7.880832651151912,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 33691
    },
    {
      "epoch": 7.8810665419249215,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 33692
    },
    {
      "epoch": 7.88130043269793,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.538,
      "step": 33693
    },
    {
      "epoch": 7.881534323470939,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 33694
    },
    {
      "epoch": 7.881768214243948,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.474,
      "step": 33695
    },
    {
      "epoch": 7.882002105016957,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 33696
    },
    {
      "epoch": 7.882235995789966,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 33697
    },
    {
      "epoch": 7.882469886562975,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 33698
    },
    {
      "epoch": 7.882703777335984,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4506,
      "step": 33699
    },
    {
      "epoch": 7.882937668108993,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 33700
    },
    {
      "epoch": 7.882937668108993,
      "eval_runtime": 4.651,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 33700
    },
    {
      "epoch": 7.883171558882002,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.2101,
      "step": 33701
    },
    {
      "epoch": 7.883405449655011,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6484,
      "step": 33702
    },
    {
      "epoch": 7.8836393404280205,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 33703
    },
    {
      "epoch": 7.883873231201029,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 33704
    },
    {
      "epoch": 7.884107121974038,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8132,
      "step": 33705
    },
    {
      "epoch": 7.884341012747047,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.23,
      "step": 33706
    },
    {
      "epoch": 7.884574903520056,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.4184,
      "step": 33707
    },
    {
      "epoch": 7.884808794293065,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 33708
    },
    {
      "epoch": 7.885042685066074,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 33709
    },
    {
      "epoch": 7.885276575839083,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7127,
      "step": 33710
    },
    {
      "epoch": 7.885510466612092,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 33711
    },
    {
      "epoch": 7.885744357385101,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.748,
      "step": 33712
    },
    {
      "epoch": 7.885978248158111,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6635,
      "step": 33713
    },
    {
      "epoch": 7.8862121389311195,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 33714
    },
    {
      "epoch": 7.886446029704128,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 33715
    },
    {
      "epoch": 7.886679920477137,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 33716
    },
    {
      "epoch": 7.886913811250146,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 33717
    },
    {
      "epoch": 7.887147702023155,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6894,
      "step": 33718
    },
    {
      "epoch": 7.887381592796165,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 33719
    },
    {
      "epoch": 7.887615483569173,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 33720
    },
    {
      "epoch": 7.887849374342182,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 33721
    },
    {
      "epoch": 7.888083265115191,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 33722
    },
    {
      "epoch": 7.8883171558882,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7607,
      "step": 33723
    },
    {
      "epoch": 7.88855104666121,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 33724
    },
    {
      "epoch": 7.8887849374342185,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7893,
      "step": 33725
    },
    {
      "epoch": 7.889018828207227,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 33726
    },
    {
      "epoch": 7.889252718980236,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5248,
      "step": 33727
    },
    {
      "epoch": 7.889486609753245,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 33728
    },
    {
      "epoch": 7.889720500526254,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6986,
      "step": 33729
    },
    {
      "epoch": 7.889954391299264,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7128,
      "step": 33730
    },
    {
      "epoch": 7.890188282072272,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 33731
    },
    {
      "epoch": 7.890422172845281,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.414,
      "step": 33732
    },
    {
      "epoch": 7.89065606361829,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5699,
      "step": 33733
    },
    {
      "epoch": 7.890889954391299,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 33734
    },
    {
      "epoch": 7.891123845164309,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 33735
    },
    {
      "epoch": 7.8913577359373175,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 33736
    },
    {
      "epoch": 7.891591626710326,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 33737
    },
    {
      "epoch": 7.891825517483335,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 33738
    },
    {
      "epoch": 7.892059408256344,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 33739
    },
    {
      "epoch": 7.892293299029353,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 33740
    },
    {
      "epoch": 7.892527189802363,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.599,
      "step": 33741
    },
    {
      "epoch": 7.892761080575371,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 33742
    },
    {
      "epoch": 7.89299497134838,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8409,
      "step": 33743
    },
    {
      "epoch": 7.893228862121389,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 33744
    },
    {
      "epoch": 7.893462752894399,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.714,
      "step": 33745
    },
    {
      "epoch": 7.893696643667408,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 33746
    },
    {
      "epoch": 7.8939305344404165,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 33747
    },
    {
      "epoch": 7.894164425213425,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 33748
    },
    {
      "epoch": 7.894398315986434,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 33749
    },
    {
      "epoch": 7.894632206759443,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5543,
      "step": 33750
    },
    {
      "epoch": 7.894866097532453,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8634,
      "step": 33751
    },
    {
      "epoch": 7.895099988305462,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 33752
    },
    {
      "epoch": 7.89533387907847,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7367,
      "step": 33753
    },
    {
      "epoch": 7.895567769851479,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 33754
    },
    {
      "epoch": 7.895801660624488,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7043,
      "step": 33755
    },
    {
      "epoch": 7.896035551397498,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6867,
      "step": 33756
    },
    {
      "epoch": 7.896269442170507,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.4878,
      "step": 33757
    },
    {
      "epoch": 7.8965033329435155,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3028,
      "step": 33758
    },
    {
      "epoch": 7.896737223716524,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7548,
      "step": 33759
    },
    {
      "epoch": 7.896971114489533,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9461,
      "step": 33760
    },
    {
      "epoch": 7.897205005262542,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6326,
      "step": 33761
    },
    {
      "epoch": 7.897438896035552,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6284,
      "step": 33762
    },
    {
      "epoch": 7.8976727868085606,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5582,
      "step": 33763
    },
    {
      "epoch": 7.897906677581569,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 33764
    },
    {
      "epoch": 7.898140568354578,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6121,
      "step": 33765
    },
    {
      "epoch": 7.898374459127587,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7449,
      "step": 33766
    },
    {
      "epoch": 7.898608349900597,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 33767
    },
    {
      "epoch": 7.898842240673606,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7321,
      "step": 33768
    },
    {
      "epoch": 7.8990761314466145,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.4021,
      "step": 33769
    },
    {
      "epoch": 7.899310022219623,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 33770
    },
    {
      "epoch": 7.899543912992632,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 33771
    },
    {
      "epoch": 7.899777803765641,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 33772
    },
    {
      "epoch": 7.900011694538651,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6331,
      "step": 33773
    },
    {
      "epoch": 7.9002455853116595,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 33774
    },
    {
      "epoch": 7.900479476084668,
      "grad_norm": 2.515625,
      "learning_rate": 3e-05,
      "loss": 1.4949,
      "step": 33775
    },
    {
      "epoch": 7.900713366857677,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6164,
      "step": 33776
    },
    {
      "epoch": 7.900947257630687,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 33777
    },
    {
      "epoch": 7.901181148403696,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5497,
      "step": 33778
    },
    {
      "epoch": 7.901415039176705,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 33779
    },
    {
      "epoch": 7.9016489299497135,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0299,
      "step": 33780
    },
    {
      "epoch": 7.901882820722722,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7794,
      "step": 33781
    },
    {
      "epoch": 7.902116711495731,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 33782
    },
    {
      "epoch": 7.902350602268741,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1904,
      "step": 33783
    },
    {
      "epoch": 7.90258449304175,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 33784
    },
    {
      "epoch": 7.9028183838147585,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 33785
    },
    {
      "epoch": 7.903052274587767,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6203,
      "step": 33786
    },
    {
      "epoch": 7.903286165360776,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 33787
    },
    {
      "epoch": 7.903520056133786,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 33788
    },
    {
      "epoch": 7.903753946906795,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 33789
    },
    {
      "epoch": 7.903987837679804,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 33790
    },
    {
      "epoch": 7.9042217284528125,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 33791
    },
    {
      "epoch": 7.904455619225821,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 33792
    },
    {
      "epoch": 7.90468950999883,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7375,
      "step": 33793
    },
    {
      "epoch": 7.90492340077184,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6439,
      "step": 33794
    },
    {
      "epoch": 7.905157291544849,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0835,
      "step": 33795
    },
    {
      "epoch": 7.9053911823178575,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6156,
      "step": 33796
    },
    {
      "epoch": 7.905625073090866,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 33797
    },
    {
      "epoch": 7.905858963863875,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 33798
    },
    {
      "epoch": 7.906092854636885,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7708,
      "step": 33799
    },
    {
      "epoch": 7.906326745409894,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7819,
      "step": 33800
    },
    {
      "epoch": 7.906326745409894,
      "eval_runtime": 4.706,
      "eval_samples_per_second": 0.212,
      "eval_steps_per_second": 0.212,
      "step": 33800
    },
    {
      "epoch": 7.906560636182903,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6247,
      "step": 33801
    },
    {
      "epoch": 7.9067945269559115,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6222,
      "step": 33802
    },
    {
      "epoch": 7.90702841772892,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 33803
    },
    {
      "epoch": 7.907262308501929,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 33804
    },
    {
      "epoch": 7.907496199274939,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.69,
      "step": 33805
    },
    {
      "epoch": 7.907730090047948,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 33806
    },
    {
      "epoch": 7.9079639808209565,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 33807
    },
    {
      "epoch": 7.908197871593965,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.515,
      "step": 33808
    },
    {
      "epoch": 7.908431762366975,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5441,
      "step": 33809
    },
    {
      "epoch": 7.908665653139984,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 33810
    },
    {
      "epoch": 7.908899543912993,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 33811
    },
    {
      "epoch": 7.909133434686002,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.645,
      "step": 33812
    },
    {
      "epoch": 7.9093673254590104,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6538,
      "step": 33813
    },
    {
      "epoch": 7.909601216232019,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5467,
      "step": 33814
    },
    {
      "epoch": 7.909835107005029,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7819,
      "step": 33815
    },
    {
      "epoch": 7.910068997778038,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 33816
    },
    {
      "epoch": 7.910302888551047,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7112,
      "step": 33817
    },
    {
      "epoch": 7.9105367793240555,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 33818
    },
    {
      "epoch": 7.910770670097064,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 33819
    },
    {
      "epoch": 7.911004560870074,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 33820
    },
    {
      "epoch": 7.911238451643083,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5948,
      "step": 33821
    },
    {
      "epoch": 7.911472342416092,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4989,
      "step": 33822
    },
    {
      "epoch": 7.911706233189101,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.2383,
      "step": 33823
    },
    {
      "epoch": 7.911940123962109,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 33824
    },
    {
      "epoch": 7.912174014735118,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 33825
    },
    {
      "epoch": 7.912407905508128,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.3974,
      "step": 33826
    },
    {
      "epoch": 7.912641796281137,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.7328,
      "step": 33827
    },
    {
      "epoch": 7.912875687054146,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 33828
    },
    {
      "epoch": 7.9131095778271545,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 33829
    },
    {
      "epoch": 7.913343468600163,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 33830
    },
    {
      "epoch": 7.913577359373173,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6673,
      "step": 33831
    },
    {
      "epoch": 7.913811250146182,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 33832
    },
    {
      "epoch": 7.914045140919191,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 33833
    },
    {
      "epoch": 7.9142790316922,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6963,
      "step": 33834
    },
    {
      "epoch": 7.914512922465208,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 33835
    },
    {
      "epoch": 7.914746813238217,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 33836
    },
    {
      "epoch": 7.914980704011227,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5169,
      "step": 33837
    },
    {
      "epoch": 7.915214594784236,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 33838
    },
    {
      "epoch": 7.915448485557245,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8891,
      "step": 33839
    },
    {
      "epoch": 7.9156823763302535,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 33840
    },
    {
      "epoch": 7.915916267103263,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4903,
      "step": 33841
    },
    {
      "epoch": 7.916150157876272,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7089,
      "step": 33842
    },
    {
      "epoch": 7.916384048649281,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.62,
      "step": 33843
    },
    {
      "epoch": 7.91661793942229,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 33844
    },
    {
      "epoch": 7.916851830195299,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5425,
      "step": 33845
    },
    {
      "epoch": 7.917085720968307,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5958,
      "step": 33846
    },
    {
      "epoch": 7.917319611741317,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5614,
      "step": 33847
    },
    {
      "epoch": 7.917553502514326,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 33848
    },
    {
      "epoch": 7.917787393287335,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 33849
    },
    {
      "epoch": 7.918021284060344,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 33850
    },
    {
      "epoch": 7.9182551748333525,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 33851
    },
    {
      "epoch": 7.918489065606362,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6106,
      "step": 33852
    },
    {
      "epoch": 7.918722956379371,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 33853
    },
    {
      "epoch": 7.91895684715238,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 33854
    },
    {
      "epoch": 7.919190737925389,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 33855
    },
    {
      "epoch": 7.919424628698398,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.4005,
      "step": 33856
    },
    {
      "epoch": 7.919658519471406,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 33857
    },
    {
      "epoch": 7.919892410244416,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 33858
    },
    {
      "epoch": 7.920126301017425,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 33859
    },
    {
      "epoch": 7.920360191790434,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 33860
    },
    {
      "epoch": 7.920594082563443,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 33861
    },
    {
      "epoch": 7.9208279733364515,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 33862
    },
    {
      "epoch": 7.921061864109461,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 33863
    },
    {
      "epoch": 7.92129575488247,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 33864
    },
    {
      "epoch": 7.921529645655479,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6252,
      "step": 33865
    },
    {
      "epoch": 7.921763536428488,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 33866
    },
    {
      "epoch": 7.921997427201497,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.3773,
      "step": 33867
    },
    {
      "epoch": 7.922231317974505,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 33868
    },
    {
      "epoch": 7.922465208747515,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6929,
      "step": 33869
    },
    {
      "epoch": 7.922699099520524,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 33870
    },
    {
      "epoch": 7.922932990293533,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 33871
    },
    {
      "epoch": 7.923166881066542,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 33872
    },
    {
      "epoch": 7.923400771839551,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9388,
      "step": 33873
    },
    {
      "epoch": 7.92363466261256,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 33874
    },
    {
      "epoch": 7.923868553385569,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8027,
      "step": 33875
    },
    {
      "epoch": 7.924102444158578,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7332,
      "step": 33876
    },
    {
      "epoch": 7.924336334931587,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6091,
      "step": 33877
    },
    {
      "epoch": 7.924570225704596,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5861,
      "step": 33878
    },
    {
      "epoch": 7.924804116477605,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 33879
    },
    {
      "epoch": 7.925038007250614,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 33880
    },
    {
      "epoch": 7.925271898023623,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 33881
    },
    {
      "epoch": 7.925505788796632,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 33882
    },
    {
      "epoch": 7.925739679569641,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 33883
    },
    {
      "epoch": 7.92597357034265,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 33884
    },
    {
      "epoch": 7.926207461115659,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6084,
      "step": 33885
    },
    {
      "epoch": 7.926441351888668,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3876,
      "step": 33886
    },
    {
      "epoch": 7.926675242661677,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 33887
    },
    {
      "epoch": 7.926909133434686,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5725,
      "step": 33888
    },
    {
      "epoch": 7.927143024207695,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.5408,
      "step": 33889
    },
    {
      "epoch": 7.927376914980704,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 33890
    },
    {
      "epoch": 7.927610805753713,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.6248,
      "step": 33891
    },
    {
      "epoch": 7.927844696526722,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 33892
    },
    {
      "epoch": 7.928078587299731,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6372,
      "step": 33893
    },
    {
      "epoch": 7.92831247807274,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7251,
      "step": 33894
    },
    {
      "epoch": 7.928546368845749,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6778,
      "step": 33895
    },
    {
      "epoch": 7.928780259618758,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5572,
      "step": 33896
    },
    {
      "epoch": 7.929014150391767,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 33897
    },
    {
      "epoch": 7.929248041164776,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 33898
    },
    {
      "epoch": 7.929481931937785,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 33899
    },
    {
      "epoch": 7.929715822710794,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 33900
    },
    {
      "epoch": 7.929715822710794,
      "eval_runtime": 4.5703,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 33900
    },
    {
      "epoch": 7.929949713483803,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 33901
    },
    {
      "epoch": 7.930183604256812,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 33902
    },
    {
      "epoch": 7.930417495029821,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7718,
      "step": 33903
    },
    {
      "epoch": 7.93065138580283,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6541,
      "step": 33904
    },
    {
      "epoch": 7.9308852765758395,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 33905
    },
    {
      "epoch": 7.931119167348848,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 33906
    },
    {
      "epoch": 7.931353058121857,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6442,
      "step": 33907
    },
    {
      "epoch": 7.931586948894866,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 33908
    },
    {
      "epoch": 7.931820839667875,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 33909
    },
    {
      "epoch": 7.932054730440884,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7083,
      "step": 33910
    },
    {
      "epoch": 7.932288621213893,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7135,
      "step": 33911
    },
    {
      "epoch": 7.932522511986902,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.4693,
      "step": 33912
    },
    {
      "epoch": 7.932756402759911,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 33913
    },
    {
      "epoch": 7.93299029353292,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 33914
    },
    {
      "epoch": 7.933224184305929,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 33915
    },
    {
      "epoch": 7.9334580750789385,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 33916
    },
    {
      "epoch": 7.933691965851947,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 33917
    },
    {
      "epoch": 7.933925856624956,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 33918
    },
    {
      "epoch": 7.934159747397965,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.697,
      "step": 33919
    },
    {
      "epoch": 7.934393638170974,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6116,
      "step": 33920
    },
    {
      "epoch": 7.934627528943983,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 33921
    },
    {
      "epoch": 7.934861419716992,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3797,
      "step": 33922
    },
    {
      "epoch": 7.935095310490001,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4783,
      "step": 33923
    },
    {
      "epoch": 7.93532920126301,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 33924
    },
    {
      "epoch": 7.935563092036019,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 33925
    },
    {
      "epoch": 7.935796982809029,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 33926
    },
    {
      "epoch": 7.9360308735820375,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 33927
    },
    {
      "epoch": 7.936264764355046,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 33928
    },
    {
      "epoch": 7.936498655128055,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5359,
      "step": 33929
    },
    {
      "epoch": 7.936732545901064,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 33930
    },
    {
      "epoch": 7.936966436674073,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.8289,
      "step": 33931
    },
    {
      "epoch": 7.937200327447083,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 33932
    },
    {
      "epoch": 7.937434218220091,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 33933
    },
    {
      "epoch": 7.9376681089931,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 33934
    },
    {
      "epoch": 7.937901999766109,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6335,
      "step": 33935
    },
    {
      "epoch": 7.938135890539118,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 33936
    },
    {
      "epoch": 7.938369781312128,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6915,
      "step": 33937
    },
    {
      "epoch": 7.9386036720851365,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 33938
    },
    {
      "epoch": 7.938837562858145,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 33939
    },
    {
      "epoch": 7.939071453631154,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5326,
      "step": 33940
    },
    {
      "epoch": 7.939305344404163,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 33941
    },
    {
      "epoch": 7.939539235177172,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.619,
      "step": 33942
    },
    {
      "epoch": 7.939773125950182,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.3282,
      "step": 33943
    },
    {
      "epoch": 7.94000701672319,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5057,
      "step": 33944
    },
    {
      "epoch": 7.940240907496199,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.4295,
      "step": 33945
    },
    {
      "epoch": 7.940474798269208,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 33946
    },
    {
      "epoch": 7.940708689042217,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 33947
    },
    {
      "epoch": 7.940942579815227,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 33948
    },
    {
      "epoch": 7.9411764705882355,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 33949
    },
    {
      "epoch": 7.941410361361244,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 33950
    },
    {
      "epoch": 7.941644252134253,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7586,
      "step": 33951
    },
    {
      "epoch": 7.941878142907262,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 33952
    },
    {
      "epoch": 7.942112033680271,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8219,
      "step": 33953
    },
    {
      "epoch": 7.942345924453281,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 33954
    },
    {
      "epoch": 7.942579815226289,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 33955
    },
    {
      "epoch": 7.942813705999298,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 33956
    },
    {
      "epoch": 7.943047596772307,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 33957
    },
    {
      "epoch": 7.943281487545317,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 33958
    },
    {
      "epoch": 7.943515378318326,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.373,
      "step": 33959
    },
    {
      "epoch": 7.9437492690913345,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7066,
      "step": 33960
    },
    {
      "epoch": 7.943983159864343,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 33961
    },
    {
      "epoch": 7.944217050637352,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 33962
    },
    {
      "epoch": 7.944450941410361,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 33963
    },
    {
      "epoch": 7.944684832183371,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7405,
      "step": 33964
    },
    {
      "epoch": 7.94491872295638,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 33965
    },
    {
      "epoch": 7.945152613729388,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 33966
    },
    {
      "epoch": 7.945386504502397,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 33967
    },
    {
      "epoch": 7.945620395275406,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 33968
    },
    {
      "epoch": 7.945854286048416,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 33969
    },
    {
      "epoch": 7.946088176821425,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7622,
      "step": 33970
    },
    {
      "epoch": 7.9463220675944335,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 33971
    },
    {
      "epoch": 7.946555958367442,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 33972
    },
    {
      "epoch": 7.946789849140451,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 33973
    },
    {
      "epoch": 7.94702373991346,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 33974
    },
    {
      "epoch": 7.94725763068647,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 33975
    },
    {
      "epoch": 7.947491521459479,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 33976
    },
    {
      "epoch": 7.947725412232487,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 33977
    },
    {
      "epoch": 7.947959303005496,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 33978
    },
    {
      "epoch": 7.948193193778505,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8061,
      "step": 33979
    },
    {
      "epoch": 7.948427084551515,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 33980
    },
    {
      "epoch": 7.948660975324524,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 33981
    },
    {
      "epoch": 7.9488948660975325,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 33982
    },
    {
      "epoch": 7.949128756870541,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 33983
    },
    {
      "epoch": 7.94936264764355,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 33984
    },
    {
      "epoch": 7.949596538416559,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 33985
    },
    {
      "epoch": 7.949830429189569,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 33986
    },
    {
      "epoch": 7.950064319962578,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 33987
    },
    {
      "epoch": 7.950298210735586,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 33988
    },
    {
      "epoch": 7.950532101508595,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 33989
    },
    {
      "epoch": 7.950765992281605,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 33990
    },
    {
      "epoch": 7.950999883054614,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1534,
      "step": 33991
    },
    {
      "epoch": 7.951233773827623,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.72,
      "step": 33992
    },
    {
      "epoch": 7.9514676646006315,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 33993
    },
    {
      "epoch": 7.95170155537364,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 33994
    },
    {
      "epoch": 7.951935446146649,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 33995
    },
    {
      "epoch": 7.952169336919659,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.4843,
      "step": 33996
    },
    {
      "epoch": 7.952403227692668,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.667,
      "step": 33997
    },
    {
      "epoch": 7.952637118465677,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7971,
      "step": 33998
    },
    {
      "epoch": 7.952871009238685,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.614,
      "step": 33999
    },
    {
      "epoch": 7.953104900011694,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 34000
    },
    {
      "epoch": 7.953104900011694,
      "eval_runtime": 4.6106,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 34000
    },
    {
      "epoch": 7.953338790784704,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 34001
    },
    {
      "epoch": 7.953572681557713,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 34002
    },
    {
      "epoch": 7.953806572330722,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.478,
      "step": 34003
    },
    {
      "epoch": 7.9540404631037305,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 34004
    },
    {
      "epoch": 7.954274353876739,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 34005
    },
    {
      "epoch": 7.954508244649748,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7343,
      "step": 34006
    },
    {
      "epoch": 7.954742135422758,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 34007
    },
    {
      "epoch": 7.954976026195767,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 34008
    },
    {
      "epoch": 7.9552099169687756,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8838,
      "step": 34009
    },
    {
      "epoch": 7.955443807741784,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7828,
      "step": 34010
    },
    {
      "epoch": 7.955677698514793,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 34011
    },
    {
      "epoch": 7.955911589287803,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 34012
    },
    {
      "epoch": 7.956145480060812,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 34013
    },
    {
      "epoch": 7.956379370833821,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6884,
      "step": 34014
    },
    {
      "epoch": 7.9566132616068295,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 34015
    },
    {
      "epoch": 7.956847152379838,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2388,
      "step": 34016
    },
    {
      "epoch": 7.957081043152847,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 34017
    },
    {
      "epoch": 7.957314933925857,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 34018
    },
    {
      "epoch": 7.957548824698866,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 34019
    },
    {
      "epoch": 7.9577827154718745,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 34020
    },
    {
      "epoch": 7.958016606244883,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7105,
      "step": 34021
    },
    {
      "epoch": 7.958250497017893,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 34022
    },
    {
      "epoch": 7.958484387790902,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 34023
    },
    {
      "epoch": 7.958718278563911,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6064,
      "step": 34024
    },
    {
      "epoch": 7.95895216933692,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 34025
    },
    {
      "epoch": 7.9591860601099285,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 34026
    },
    {
      "epoch": 7.959419950882937,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 34027
    },
    {
      "epoch": 7.959653841655947,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 34028
    },
    {
      "epoch": 7.959887732428956,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 34029
    },
    {
      "epoch": 7.960121623201965,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 34030
    },
    {
      "epoch": 7.9603555139749735,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 34031
    },
    {
      "epoch": 7.960589404747982,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 34032
    },
    {
      "epoch": 7.960823295520992,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6317,
      "step": 34033
    },
    {
      "epoch": 7.961057186294001,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6133,
      "step": 34034
    },
    {
      "epoch": 7.96129107706701,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 34035
    },
    {
      "epoch": 7.961524967840019,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 34036
    },
    {
      "epoch": 7.9617588586130275,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.7549,
      "step": 34037
    },
    {
      "epoch": 7.961992749386036,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 34038
    },
    {
      "epoch": 7.962226640159046,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 34039
    },
    {
      "epoch": 7.962460530932055,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 34040
    },
    {
      "epoch": 7.962694421705064,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5665,
      "step": 34041
    },
    {
      "epoch": 7.9629283124780725,
      "grad_norm": 12.6875,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 34042
    },
    {
      "epoch": 7.963162203251081,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 34043
    },
    {
      "epoch": 7.963396094024091,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 34044
    },
    {
      "epoch": 7.9636299847971,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 34045
    },
    {
      "epoch": 7.963863875570109,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4747,
      "step": 34046
    },
    {
      "epoch": 7.964097766343118,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 34047
    },
    {
      "epoch": 7.9643316571161265,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 34048
    },
    {
      "epoch": 7.964565547889135,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 34049
    },
    {
      "epoch": 7.964799438662145,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8038,
      "step": 34050
    },
    {
      "epoch": 7.965033329435154,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 34051
    },
    {
      "epoch": 7.965267220208163,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6639,
      "step": 34052
    },
    {
      "epoch": 7.9655011109811715,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 34053
    },
    {
      "epoch": 7.965735001754181,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 34054
    },
    {
      "epoch": 7.96596889252719,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7352,
      "step": 34055
    },
    {
      "epoch": 7.966202783300199,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 34056
    },
    {
      "epoch": 7.966436674073208,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 34057
    },
    {
      "epoch": 7.966670564846217,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5261,
      "step": 34058
    },
    {
      "epoch": 7.9669044556192254,
      "grad_norm": 7.65625,
      "learning_rate": 3e-05,
      "loss": 1.5747,
      "step": 34059
    },
    {
      "epoch": 7.967138346392235,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7317,
      "step": 34060
    },
    {
      "epoch": 7.967372237165244,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8577,
      "step": 34061
    },
    {
      "epoch": 7.967606127938253,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 34062
    },
    {
      "epoch": 7.967840018711262,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 34063
    },
    {
      "epoch": 7.9680739094842705,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 34064
    },
    {
      "epoch": 7.96830780025728,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 34065
    },
    {
      "epoch": 7.968541691030289,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 34066
    },
    {
      "epoch": 7.968775581803298,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6399,
      "step": 34067
    },
    {
      "epoch": 7.969009472576307,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.4826,
      "step": 34068
    },
    {
      "epoch": 7.969243363349316,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 34069
    },
    {
      "epoch": 7.969477254122324,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 34070
    },
    {
      "epoch": 7.969711144895334,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8083,
      "step": 34071
    },
    {
      "epoch": 7.969945035668343,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 34072
    },
    {
      "epoch": 7.970178926441352,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 34073
    },
    {
      "epoch": 7.970412817214361,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 34074
    },
    {
      "epoch": 7.9706467079873695,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6155,
      "step": 34075
    },
    {
      "epoch": 7.970880598760379,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 34076
    },
    {
      "epoch": 7.971114489533388,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 34077
    },
    {
      "epoch": 7.971348380306397,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 34078
    },
    {
      "epoch": 7.971582271079406,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8721,
      "step": 34079
    },
    {
      "epoch": 7.971816161852415,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 34080
    },
    {
      "epoch": 7.972050052625423,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.541,
      "step": 34081
    },
    {
      "epoch": 7.972283943398433,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.509,
      "step": 34082
    },
    {
      "epoch": 7.972517834171442,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7188,
      "step": 34083
    },
    {
      "epoch": 7.972751724944451,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 34084
    },
    {
      "epoch": 7.97298561571746,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 34085
    },
    {
      "epoch": 7.973219506490469,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5169,
      "step": 34086
    },
    {
      "epoch": 7.973453397263478,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 34087
    },
    {
      "epoch": 7.973687288036487,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 34088
    },
    {
      "epoch": 7.973921178809496,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 34089
    },
    {
      "epoch": 7.974155069582505,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6292,
      "step": 34090
    },
    {
      "epoch": 7.974388960355514,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 34091
    },
    {
      "epoch": 7.974622851128523,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5423,
      "step": 34092
    },
    {
      "epoch": 7.974856741901532,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 34093
    },
    {
      "epoch": 7.975090632674541,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 34094
    },
    {
      "epoch": 7.97532452344755,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 34095
    },
    {
      "epoch": 7.975558414220559,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6861,
      "step": 34096
    },
    {
      "epoch": 7.975792304993568,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5069,
      "step": 34097
    },
    {
      "epoch": 7.976026195766577,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6166,
      "step": 34098
    },
    {
      "epoch": 7.976260086539586,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 34099
    },
    {
      "epoch": 7.976493977312595,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 34100
    },
    {
      "epoch": 7.976493977312595,
      "eval_runtime": 4.6516,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 34100
    },
    {
      "epoch": 7.976727868085604,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 34101
    },
    {
      "epoch": 7.976961758858613,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.5461,
      "step": 34102
    },
    {
      "epoch": 7.977195649631622,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 34103
    },
    {
      "epoch": 7.977429540404631,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.627,
      "step": 34104
    },
    {
      "epoch": 7.97766343117764,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 34105
    },
    {
      "epoch": 7.977897321950649,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5005,
      "step": 34106
    },
    {
      "epoch": 7.9781312127236585,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5868,
      "step": 34107
    },
    {
      "epoch": 7.978365103496667,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 34108
    },
    {
      "epoch": 7.978598994269676,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 34109
    },
    {
      "epoch": 7.978832885042685,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 34110
    },
    {
      "epoch": 7.979066775815694,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 34111
    },
    {
      "epoch": 7.979300666588703,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6393,
      "step": 34112
    },
    {
      "epoch": 7.9795345573617125,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6231,
      "step": 34113
    },
    {
      "epoch": 7.979768448134721,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 34114
    },
    {
      "epoch": 7.98000233890773,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 34115
    },
    {
      "epoch": 7.980236229680739,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 34116
    },
    {
      "epoch": 7.980470120453748,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 34117
    },
    {
      "epoch": 7.9807040112267575,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 34118
    },
    {
      "epoch": 7.980937901999766,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 34119
    },
    {
      "epoch": 7.981171792772775,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 34120
    },
    {
      "epoch": 7.981405683545784,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 34121
    },
    {
      "epoch": 7.981639574318793,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 34122
    },
    {
      "epoch": 7.981873465091802,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 34123
    },
    {
      "epoch": 7.9821073558648115,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 34124
    },
    {
      "epoch": 7.98234124663782,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 34125
    },
    {
      "epoch": 7.982575137410829,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 34126
    },
    {
      "epoch": 7.982809028183838,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 34127
    },
    {
      "epoch": 7.983042918956847,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 34128
    },
    {
      "epoch": 7.9832768097298565,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6422,
      "step": 34129
    },
    {
      "epoch": 7.983510700502865,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.51,
      "step": 34130
    },
    {
      "epoch": 7.983744591275874,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 34131
    },
    {
      "epoch": 7.983978482048883,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.4548,
      "step": 34132
    },
    {
      "epoch": 7.984212372821892,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.582,
      "step": 34133
    },
    {
      "epoch": 7.984446263594901,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 34134
    },
    {
      "epoch": 7.9846801543679105,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7019,
      "step": 34135
    },
    {
      "epoch": 7.984914045140919,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 34136
    },
    {
      "epoch": 7.985147935913928,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 34137
    },
    {
      "epoch": 7.985381826686937,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 34138
    },
    {
      "epoch": 7.985615717459947,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 34139
    },
    {
      "epoch": 7.9858496082329555,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6501,
      "step": 34140
    },
    {
      "epoch": 7.986083499005964,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 34141
    },
    {
      "epoch": 7.986317389778973,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.6776,
      "step": 34142
    },
    {
      "epoch": 7.986551280551982,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6716,
      "step": 34143
    },
    {
      "epoch": 7.986785171324991,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 34144
    },
    {
      "epoch": 7.987019062098001,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 34145
    },
    {
      "epoch": 7.9872529528710094,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 34146
    },
    {
      "epoch": 7.987486843644018,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 34147
    },
    {
      "epoch": 7.987720734417027,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6988,
      "step": 34148
    },
    {
      "epoch": 7.987954625190036,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5695,
      "step": 34149
    },
    {
      "epoch": 7.988188515963046,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 34150
    },
    {
      "epoch": 7.9884224067360545,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 34151
    },
    {
      "epoch": 7.988656297509063,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 34152
    },
    {
      "epoch": 7.988890188282072,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.4424,
      "step": 34153
    },
    {
      "epoch": 7.989124079055081,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 34154
    },
    {
      "epoch": 7.98935796982809,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 34155
    },
    {
      "epoch": 7.9895918606011,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 34156
    },
    {
      "epoch": 7.989825751374108,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 34157
    },
    {
      "epoch": 7.990059642147117,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 34158
    },
    {
      "epoch": 7.990293532920126,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5997,
      "step": 34159
    },
    {
      "epoch": 7.990527423693135,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5823,
      "step": 34160
    },
    {
      "epoch": 7.990761314466145,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 34161
    },
    {
      "epoch": 7.9909952052391535,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 34162
    },
    {
      "epoch": 7.991229096012162,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 34163
    },
    {
      "epoch": 7.991462986785171,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 34164
    },
    {
      "epoch": 7.99169687755818,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 34165
    },
    {
      "epoch": 7.991930768331189,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 34166
    },
    {
      "epoch": 7.992164659104199,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6484,
      "step": 34167
    },
    {
      "epoch": 7.992398549877207,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9946,
      "step": 34168
    },
    {
      "epoch": 7.992632440650216,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 34169
    },
    {
      "epoch": 7.992866331423225,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7162,
      "step": 34170
    },
    {
      "epoch": 7.993100222196235,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 34171
    },
    {
      "epoch": 7.993334112969244,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 34172
    },
    {
      "epoch": 7.9935680037422525,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 34173
    },
    {
      "epoch": 7.993801894515261,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.709,
      "step": 34174
    },
    {
      "epoch": 7.99403578528827,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 34175
    },
    {
      "epoch": 7.994269676061279,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5772,
      "step": 34176
    },
    {
      "epoch": 7.994503566834289,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 34177
    },
    {
      "epoch": 7.994737457607298,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 34178
    },
    {
      "epoch": 7.994971348380306,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 34179
    },
    {
      "epoch": 7.995205239153315,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 34180
    },
    {
      "epoch": 7.995439129926324,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 34181
    },
    {
      "epoch": 7.995673020699334,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7198,
      "step": 34182
    },
    {
      "epoch": 7.995906911472343,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 34183
    },
    {
      "epoch": 7.9961408022453515,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.827,
      "step": 34184
    },
    {
      "epoch": 7.99637469301836,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8257,
      "step": 34185
    },
    {
      "epoch": 7.996608583791369,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 34186
    },
    {
      "epoch": 7.996842474564378,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.3532,
      "step": 34187
    },
    {
      "epoch": 7.997076365337388,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 34188
    },
    {
      "epoch": 7.997310256110397,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 34189
    },
    {
      "epoch": 7.997544146883405,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 34190
    },
    {
      "epoch": 7.997778037656414,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 34191
    },
    {
      "epoch": 7.998011928429423,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4808,
      "step": 34192
    },
    {
      "epoch": 7.998245819202433,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 34193
    },
    {
      "epoch": 7.998479709975442,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7488,
      "step": 34194
    },
    {
      "epoch": 7.9987136007484505,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5369,
      "step": 34195
    },
    {
      "epoch": 7.998947491521459,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 34196
    },
    {
      "epoch": 7.999181382294468,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5551,
      "step": 34197
    },
    {
      "epoch": 7.999415273067477,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8002,
      "step": 34198
    },
    {
      "epoch": 7.999649163840487,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 34199
    },
    {
      "epoch": 7.999883054613496,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.754,
      "step": 34200
    },
    {
      "epoch": 7.999883054613496,
      "eval_runtime": 4.6273,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 34200
    },
    {
      "epoch": 8.000233890773009,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 3.4717,
      "step": 34201
    },
    {
      "epoch": 8.000467781546018,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7997,
      "step": 34202
    },
    {
      "epoch": 8.000701672319027,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 34203
    },
    {
      "epoch": 8.000935563092035,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8278,
      "step": 34204
    },
    {
      "epoch": 8.001169453865044,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 34205
    },
    {
      "epoch": 8.001403344638055,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 34206
    },
    {
      "epoch": 8.001637235411064,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7792,
      "step": 34207
    },
    {
      "epoch": 8.001871126184072,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5316,
      "step": 34208
    },
    {
      "epoch": 8.002105016957081,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7608,
      "step": 34209
    },
    {
      "epoch": 8.00233890773009,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 34210
    },
    {
      "epoch": 8.002572798503099,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 34211
    },
    {
      "epoch": 8.002806689276108,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 34212
    },
    {
      "epoch": 8.003040580049117,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 34213
    },
    {
      "epoch": 8.003274470822126,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7284,
      "step": 34214
    },
    {
      "epoch": 8.003508361595134,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6749,
      "step": 34215
    },
    {
      "epoch": 8.003742252368145,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.916,
      "step": 34216
    },
    {
      "epoch": 8.003976143141154,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5982,
      "step": 34217
    },
    {
      "epoch": 8.004210033914163,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7687,
      "step": 34218
    },
    {
      "epoch": 8.004443924687171,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4173,
      "step": 34219
    },
    {
      "epoch": 8.00467781546018,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 34220
    },
    {
      "epoch": 8.00491170623319,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 34221
    },
    {
      "epoch": 8.005145597006198,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 34222
    },
    {
      "epoch": 8.005379487779207,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 34223
    },
    {
      "epoch": 8.005613378552216,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6027,
      "step": 34224
    },
    {
      "epoch": 8.005847269325225,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 34225
    },
    {
      "epoch": 8.006081160098233,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6814,
      "step": 34226
    },
    {
      "epoch": 8.006315050871244,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4566,
      "step": 34227
    },
    {
      "epoch": 8.006548941644253,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 34228
    },
    {
      "epoch": 8.006782832417262,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8363,
      "step": 34229
    },
    {
      "epoch": 8.00701672319027,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 34230
    },
    {
      "epoch": 8.00725061396328,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 34231
    },
    {
      "epoch": 8.007484504736288,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 34232
    },
    {
      "epoch": 8.007718395509297,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 34233
    },
    {
      "epoch": 8.007952286282306,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.3874,
      "step": 34234
    },
    {
      "epoch": 8.008186177055315,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 34235
    },
    {
      "epoch": 8.008420067828323,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.641,
      "step": 34236
    },
    {
      "epoch": 8.008653958601332,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 34237
    },
    {
      "epoch": 8.008887849374343,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 34238
    },
    {
      "epoch": 8.009121740147352,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7054,
      "step": 34239
    },
    {
      "epoch": 8.00935563092036,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 34240
    },
    {
      "epoch": 8.00958952169337,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 34241
    },
    {
      "epoch": 8.009823412466378,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 34242
    },
    {
      "epoch": 8.010057303239387,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7033,
      "step": 34243
    },
    {
      "epoch": 8.010291194012396,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 34244
    },
    {
      "epoch": 8.010525084785405,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5612,
      "step": 34245
    },
    {
      "epoch": 8.010758975558414,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7838,
      "step": 34246
    },
    {
      "epoch": 8.010992866331422,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5924,
      "step": 34247
    },
    {
      "epoch": 8.011226757104433,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6372,
      "step": 34248
    },
    {
      "epoch": 8.011460647877442,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 34249
    },
    {
      "epoch": 8.01169453865045,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4903,
      "step": 34250
    },
    {
      "epoch": 8.01192842942346,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 34251
    },
    {
      "epoch": 8.012162320196468,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 34252
    },
    {
      "epoch": 8.012396210969477,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7377,
      "step": 34253
    },
    {
      "epoch": 8.012630101742486,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 34254
    },
    {
      "epoch": 8.012863992515495,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5554,
      "step": 34255
    },
    {
      "epoch": 8.013097883288504,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.358,
      "step": 34256
    },
    {
      "epoch": 8.013331774061513,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6868,
      "step": 34257
    },
    {
      "epoch": 8.013565664834521,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 34258
    },
    {
      "epoch": 8.013799555607532,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 34259
    },
    {
      "epoch": 8.014033446380541,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 34260
    },
    {
      "epoch": 8.01426733715355,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4149,
      "step": 34261
    },
    {
      "epoch": 8.014501227926559,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6445,
      "step": 34262
    },
    {
      "epoch": 8.014735118699567,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 34263
    },
    {
      "epoch": 8.014969009472576,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 34264
    },
    {
      "epoch": 8.015202900245585,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 34265
    },
    {
      "epoch": 8.015436791018594,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 34266
    },
    {
      "epoch": 8.015670681791603,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6428,
      "step": 34267
    },
    {
      "epoch": 8.015904572564612,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 34268
    },
    {
      "epoch": 8.01613846333762,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 1.6857,
      "step": 34269
    },
    {
      "epoch": 8.016372354110631,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 34270
    },
    {
      "epoch": 8.01660624488364,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 34271
    },
    {
      "epoch": 8.016840135656649,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 34272
    },
    {
      "epoch": 8.017074026429658,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5584,
      "step": 34273
    },
    {
      "epoch": 8.017307917202666,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 34274
    },
    {
      "epoch": 8.017541807975675,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 34275
    },
    {
      "epoch": 8.017775698748684,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 34276
    },
    {
      "epoch": 8.018009589521693,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 34277
    },
    {
      "epoch": 8.018243480294702,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.7391,
      "step": 34278
    },
    {
      "epoch": 8.01847737106771,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 34279
    },
    {
      "epoch": 8.018711261840721,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 34280
    },
    {
      "epoch": 8.01894515261373,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 34281
    },
    {
      "epoch": 8.019179043386739,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6078,
      "step": 34282
    },
    {
      "epoch": 8.019412934159748,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6801,
      "step": 34283
    },
    {
      "epoch": 8.019646824932757,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6731,
      "step": 34284
    },
    {
      "epoch": 8.019880715705765,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 34285
    },
    {
      "epoch": 8.020114606478774,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6399,
      "step": 34286
    },
    {
      "epoch": 8.020348497251783,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 34287
    },
    {
      "epoch": 8.020582388024792,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8176,
      "step": 34288
    },
    {
      "epoch": 8.0208162787978,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 34289
    },
    {
      "epoch": 8.02105016957081,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5929,
      "step": 34290
    },
    {
      "epoch": 8.02128406034382,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6935,
      "step": 34291
    },
    {
      "epoch": 8.021517951116829,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 34292
    },
    {
      "epoch": 8.021751841889838,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 34293
    },
    {
      "epoch": 8.021985732662847,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5971,
      "step": 34294
    },
    {
      "epoch": 8.022219623435856,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7366,
      "step": 34295
    },
    {
      "epoch": 8.022453514208864,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.556,
      "step": 34296
    },
    {
      "epoch": 8.022687404981873,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 34297
    },
    {
      "epoch": 8.022921295754882,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 34298
    },
    {
      "epoch": 8.023155186527891,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.1022,
      "step": 34299
    },
    {
      "epoch": 8.0233890773009,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 34300
    },
    {
      "epoch": 8.0233890773009,
      "eval_runtime": 4.6308,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 34300
    },
    {
      "epoch": 8.023622968073909,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 34301
    },
    {
      "epoch": 8.02385685884692,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8607,
      "step": 34302
    },
    {
      "epoch": 8.024090749619928,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 34303
    },
    {
      "epoch": 8.024324640392937,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 34304
    },
    {
      "epoch": 8.024558531165946,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 34305
    },
    {
      "epoch": 8.024792421938955,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 34306
    },
    {
      "epoch": 8.025026312711963,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7079,
      "step": 34307
    },
    {
      "epoch": 8.025260203484972,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5297,
      "step": 34308
    },
    {
      "epoch": 8.025494094257981,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6391,
      "step": 34309
    },
    {
      "epoch": 8.02572798503099,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6422,
      "step": 34310
    },
    {
      "epoch": 8.025961875803999,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 34311
    },
    {
      "epoch": 8.02619576657701,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5307,
      "step": 34312
    },
    {
      "epoch": 8.026429657350018,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.5954,
      "step": 34313
    },
    {
      "epoch": 8.026663548123027,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 34314
    },
    {
      "epoch": 8.026897438896036,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 34315
    },
    {
      "epoch": 8.027131329669045,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5892,
      "step": 34316
    },
    {
      "epoch": 8.027365220442054,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 34317
    },
    {
      "epoch": 8.027599111215062,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7653,
      "step": 34318
    },
    {
      "epoch": 8.027833001988071,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 34319
    },
    {
      "epoch": 8.02806689276108,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 34320
    },
    {
      "epoch": 8.028300783534089,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 34321
    },
    {
      "epoch": 8.028534674307098,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 34322
    },
    {
      "epoch": 8.028768565080108,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 34323
    },
    {
      "epoch": 8.029002455853117,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6516,
      "step": 34324
    },
    {
      "epoch": 8.029236346626126,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5975,
      "step": 34325
    },
    {
      "epoch": 8.029470237399135,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.4072,
      "step": 34326
    },
    {
      "epoch": 8.029704128172144,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6394,
      "step": 34327
    },
    {
      "epoch": 8.029938018945153,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 34328
    },
    {
      "epoch": 8.030171909718161,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 34329
    },
    {
      "epoch": 8.03040580049117,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 34330
    },
    {
      "epoch": 8.030639691264179,
      "grad_norm": 7.375,
      "learning_rate": 3e-05,
      "loss": 1.5414,
      "step": 34331
    },
    {
      "epoch": 8.030873582037188,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 34332
    },
    {
      "epoch": 8.031107472810197,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7586,
      "step": 34333
    },
    {
      "epoch": 8.031341363583207,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 34334
    },
    {
      "epoch": 8.031575254356216,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 34335
    },
    {
      "epoch": 8.031809145129225,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6486,
      "step": 34336
    },
    {
      "epoch": 8.032043035902234,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6151,
      "step": 34337
    },
    {
      "epoch": 8.032276926675243,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 34338
    },
    {
      "epoch": 8.032510817448252,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6322,
      "step": 34339
    },
    {
      "epoch": 8.03274470822126,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 1.9033,
      "step": 34340
    },
    {
      "epoch": 8.03297859899427,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4901,
      "step": 34341
    },
    {
      "epoch": 8.033212489767278,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 34342
    },
    {
      "epoch": 8.033446380540287,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 34343
    },
    {
      "epoch": 8.033680271313298,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 34344
    },
    {
      "epoch": 8.033914162086306,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 34345
    },
    {
      "epoch": 8.034148052859315,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 34346
    },
    {
      "epoch": 8.034381943632324,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 34347
    },
    {
      "epoch": 8.034615834405333,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6768,
      "step": 34348
    },
    {
      "epoch": 8.034849725178342,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 34349
    },
    {
      "epoch": 8.03508361595135,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 34350
    },
    {
      "epoch": 8.03531750672436,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 34351
    },
    {
      "epoch": 8.035551397497368,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 34352
    },
    {
      "epoch": 8.035785288270377,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7034,
      "step": 34353
    },
    {
      "epoch": 8.036019179043386,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5787,
      "step": 34354
    },
    {
      "epoch": 8.036253069816397,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 34355
    },
    {
      "epoch": 8.036486960589405,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 34356
    },
    {
      "epoch": 8.036720851362414,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 34357
    },
    {
      "epoch": 8.036954742135423,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 34358
    },
    {
      "epoch": 8.037188632908432,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 34359
    },
    {
      "epoch": 8.03742252368144,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.3671,
      "step": 34360
    },
    {
      "epoch": 8.03765641445445,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 34361
    },
    {
      "epoch": 8.037890305227458,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 34362
    },
    {
      "epoch": 8.038124196000467,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6547,
      "step": 34363
    },
    {
      "epoch": 8.038358086773476,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 34364
    },
    {
      "epoch": 8.038591977546485,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 34365
    },
    {
      "epoch": 8.038825868319496,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4367,
      "step": 34366
    },
    {
      "epoch": 8.039059759092504,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 34367
    },
    {
      "epoch": 8.039293649865513,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 34368
    },
    {
      "epoch": 8.039527540638522,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 34369
    },
    {
      "epoch": 8.03976143141153,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 34370
    },
    {
      "epoch": 8.03999532218454,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 34371
    },
    {
      "epoch": 8.040229212957549,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 34372
    },
    {
      "epoch": 8.040463103730557,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 34373
    },
    {
      "epoch": 8.040696994503566,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5569,
      "step": 34374
    },
    {
      "epoch": 8.040930885276575,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.529,
      "step": 34375
    },
    {
      "epoch": 8.041164776049586,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5077,
      "step": 34376
    },
    {
      "epoch": 8.041398666822595,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 34377
    },
    {
      "epoch": 8.041632557595603,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 34378
    },
    {
      "epoch": 8.041866448368612,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5605,
      "step": 34379
    },
    {
      "epoch": 8.042100339141621,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 34380
    },
    {
      "epoch": 8.04233422991463,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 34381
    },
    {
      "epoch": 8.042568120687639,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 34382
    },
    {
      "epoch": 8.042802011460648,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6695,
      "step": 34383
    },
    {
      "epoch": 8.043035902233656,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.5829,
      "step": 34384
    },
    {
      "epoch": 8.043269793006665,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6547,
      "step": 34385
    },
    {
      "epoch": 8.043503683779674,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 34386
    },
    {
      "epoch": 8.043737574552685,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5186,
      "step": 34387
    },
    {
      "epoch": 8.043971465325694,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.4588,
      "step": 34388
    },
    {
      "epoch": 8.044205356098702,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6382,
      "step": 34389
    },
    {
      "epoch": 8.044439246871711,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 34390
    },
    {
      "epoch": 8.04467313764472,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 34391
    },
    {
      "epoch": 8.044907028417729,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.653,
      "step": 34392
    },
    {
      "epoch": 8.045140919190738,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 34393
    },
    {
      "epoch": 8.045374809963747,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6571,
      "step": 34394
    },
    {
      "epoch": 8.045608700736755,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5502,
      "step": 34395
    },
    {
      "epoch": 8.045842591509764,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7419,
      "step": 34396
    },
    {
      "epoch": 8.046076482282775,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 34397
    },
    {
      "epoch": 8.046310373055784,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7413,
      "step": 34398
    },
    {
      "epoch": 8.046544263828793,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 34399
    },
    {
      "epoch": 8.046778154601801,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 34400
    },
    {
      "epoch": 8.046778154601801,
      "eval_runtime": 4.5928,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 34400
    },
    {
      "epoch": 8.04701204537481,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 34401
    },
    {
      "epoch": 8.047245936147819,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 34402
    },
    {
      "epoch": 8.047479826920828,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7906,
      "step": 34403
    },
    {
      "epoch": 8.047713717693837,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 34404
    },
    {
      "epoch": 8.047947608466846,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 34405
    },
    {
      "epoch": 8.048181499239854,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6432,
      "step": 34406
    },
    {
      "epoch": 8.048415390012863,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 34407
    },
    {
      "epoch": 8.048649280785874,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.6386,
      "step": 34408
    },
    {
      "epoch": 8.048883171558883,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 34409
    },
    {
      "epoch": 8.049117062331892,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6551,
      "step": 34410
    },
    {
      "epoch": 8.0493509531049,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6878,
      "step": 34411
    },
    {
      "epoch": 8.04958484387791,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 34412
    },
    {
      "epoch": 8.049818734650918,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5688,
      "step": 34413
    },
    {
      "epoch": 8.050052625423927,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.4652,
      "step": 34414
    },
    {
      "epoch": 8.050286516196936,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 34415
    },
    {
      "epoch": 8.050520406969945,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 34416
    },
    {
      "epoch": 8.050754297742953,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6972,
      "step": 34417
    },
    {
      "epoch": 8.050988188515962,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.4621,
      "step": 34418
    },
    {
      "epoch": 8.051222079288973,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 34419
    },
    {
      "epoch": 8.051455970061982,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 34420
    },
    {
      "epoch": 8.05168986083499,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 34421
    },
    {
      "epoch": 8.051923751608,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4557,
      "step": 34422
    },
    {
      "epoch": 8.052157642381008,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 34423
    },
    {
      "epoch": 8.052391533154017,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 34424
    },
    {
      "epoch": 8.052625423927026,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 34425
    },
    {
      "epoch": 8.052859314700035,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.679,
      "step": 34426
    },
    {
      "epoch": 8.053093205473044,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.723,
      "step": 34427
    },
    {
      "epoch": 8.053327096246052,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.6643,
      "step": 34428
    },
    {
      "epoch": 8.053560987019063,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 34429
    },
    {
      "epoch": 8.053794877792072,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 34430
    },
    {
      "epoch": 8.05402876856508,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6362,
      "step": 34431
    },
    {
      "epoch": 8.05426265933809,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 34432
    },
    {
      "epoch": 8.054496550111098,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5865,
      "step": 34433
    },
    {
      "epoch": 8.054730440884107,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5975,
      "step": 34434
    },
    {
      "epoch": 8.054964331657116,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5879,
      "step": 34435
    },
    {
      "epoch": 8.055198222430125,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 34436
    },
    {
      "epoch": 8.055432113203134,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6699,
      "step": 34437
    },
    {
      "epoch": 8.055666003976143,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 34438
    },
    {
      "epoch": 8.055899894749151,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7355,
      "step": 34439
    },
    {
      "epoch": 8.056133785522162,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5825,
      "step": 34440
    },
    {
      "epoch": 8.05636767629517,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4843,
      "step": 34441
    },
    {
      "epoch": 8.05660156706818,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5908,
      "step": 34442
    },
    {
      "epoch": 8.056835457841188,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 34443
    },
    {
      "epoch": 8.057069348614197,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 34444
    },
    {
      "epoch": 8.057303239387206,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 34445
    },
    {
      "epoch": 8.057537130160215,
      "grad_norm": 2.484375,
      "learning_rate": 3e-05,
      "loss": 1.5302,
      "step": 34446
    },
    {
      "epoch": 8.057771020933224,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 34447
    },
    {
      "epoch": 8.058004911706233,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6116,
      "step": 34448
    },
    {
      "epoch": 8.058238802479242,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 34449
    },
    {
      "epoch": 8.05847269325225,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8118,
      "step": 34450
    },
    {
      "epoch": 8.058706584025261,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 34451
    },
    {
      "epoch": 8.05894047479827,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 34452
    },
    {
      "epoch": 8.059174365571279,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6016,
      "step": 34453
    },
    {
      "epoch": 8.059408256344287,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7318,
      "step": 34454
    },
    {
      "epoch": 8.059642147117296,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 34455
    },
    {
      "epoch": 8.059876037890305,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 34456
    },
    {
      "epoch": 8.060109928663314,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6467,
      "step": 34457
    },
    {
      "epoch": 8.060343819436323,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 34458
    },
    {
      "epoch": 8.060577710209332,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 34459
    },
    {
      "epoch": 8.06081160098234,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 34460
    },
    {
      "epoch": 8.061045491755351,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 34461
    },
    {
      "epoch": 8.06127938252836,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5412,
      "step": 34462
    },
    {
      "epoch": 8.061513273301369,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6862,
      "step": 34463
    },
    {
      "epoch": 8.061747164074378,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7037,
      "step": 34464
    },
    {
      "epoch": 8.061981054847386,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4713,
      "step": 34465
    },
    {
      "epoch": 8.062214945620395,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 34466
    },
    {
      "epoch": 8.062448836393404,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6249,
      "step": 34467
    },
    {
      "epoch": 8.062682727166413,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 34468
    },
    {
      "epoch": 8.062916617939422,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.3222,
      "step": 34469
    },
    {
      "epoch": 8.06315050871243,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 34470
    },
    {
      "epoch": 8.06338439948544,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 34471
    },
    {
      "epoch": 8.06361829025845,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 34472
    },
    {
      "epoch": 8.063852181031459,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2336,
      "step": 34473
    },
    {
      "epoch": 8.064086071804468,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 34474
    },
    {
      "epoch": 8.064319962577477,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 34475
    },
    {
      "epoch": 8.064553853350485,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 34476
    },
    {
      "epoch": 8.064787744123494,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5994,
      "step": 34477
    },
    {
      "epoch": 8.065021634896503,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8337,
      "step": 34478
    },
    {
      "epoch": 8.065255525669512,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5603,
      "step": 34479
    },
    {
      "epoch": 8.06548941644252,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6564,
      "step": 34480
    },
    {
      "epoch": 8.06572330721553,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.618,
      "step": 34481
    },
    {
      "epoch": 8.065957197988538,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7146,
      "step": 34482
    },
    {
      "epoch": 8.066191088761549,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 34483
    },
    {
      "epoch": 8.066424979534558,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.4818,
      "step": 34484
    },
    {
      "epoch": 8.066658870307567,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2206,
      "step": 34485
    },
    {
      "epoch": 8.066892761080576,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7152,
      "step": 34486
    },
    {
      "epoch": 8.067126651853584,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 34487
    },
    {
      "epoch": 8.067360542626593,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 34488
    },
    {
      "epoch": 8.067594433399602,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.6132,
      "step": 34489
    },
    {
      "epoch": 8.067828324172611,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 34490
    },
    {
      "epoch": 8.06806221494562,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 34491
    },
    {
      "epoch": 8.068296105718629,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 34492
    },
    {
      "epoch": 8.06852999649164,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6871,
      "step": 34493
    },
    {
      "epoch": 8.068763887264648,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 34494
    },
    {
      "epoch": 8.068997778037657,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 34495
    },
    {
      "epoch": 8.069231668810666,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7049,
      "step": 34496
    },
    {
      "epoch": 8.069465559583675,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5539,
      "step": 34497
    },
    {
      "epoch": 8.069699450356683,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 34498
    },
    {
      "epoch": 8.069933341129692,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5939,
      "step": 34499
    },
    {
      "epoch": 8.070167231902701,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 34500
    },
    {
      "epoch": 8.070167231902701,
      "eval_runtime": 4.6365,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 34500
    },
    {
      "epoch": 8.07040112267571,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 34501
    },
    {
      "epoch": 8.070635013448719,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5991,
      "step": 34502
    },
    {
      "epoch": 8.070868904221728,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 34503
    },
    {
      "epoch": 8.071102794994738,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 34504
    },
    {
      "epoch": 8.071336685767747,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 34505
    },
    {
      "epoch": 8.071570576540756,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.439,
      "step": 34506
    },
    {
      "epoch": 8.071804467313765,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 34507
    },
    {
      "epoch": 8.072038358086774,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7244,
      "step": 34508
    },
    {
      "epoch": 8.072272248859782,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 34509
    },
    {
      "epoch": 8.072506139632791,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 34510
    },
    {
      "epoch": 8.0727400304058,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 34511
    },
    {
      "epoch": 8.072973921178809,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4875,
      "step": 34512
    },
    {
      "epoch": 8.073207811951818,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6579,
      "step": 34513
    },
    {
      "epoch": 8.073441702724827,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5131,
      "step": 34514
    },
    {
      "epoch": 8.073675593497837,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 34515
    },
    {
      "epoch": 8.073909484270846,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5929,
      "step": 34516
    },
    {
      "epoch": 8.074143375043855,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 34517
    },
    {
      "epoch": 8.074377265816864,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 34518
    },
    {
      "epoch": 8.074611156589873,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 34519
    },
    {
      "epoch": 8.074845047362881,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5532,
      "step": 34520
    },
    {
      "epoch": 8.07507893813589,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 34521
    },
    {
      "epoch": 8.0753128289089,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 34522
    },
    {
      "epoch": 8.075546719681908,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 34523
    },
    {
      "epoch": 8.075780610454917,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 34524
    },
    {
      "epoch": 8.076014501227927,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4547,
      "step": 34525
    },
    {
      "epoch": 8.076248392000936,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 34526
    },
    {
      "epoch": 8.076482282773945,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 34527
    },
    {
      "epoch": 8.076716173546954,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 34528
    },
    {
      "epoch": 8.076950064319963,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7858,
      "step": 34529
    },
    {
      "epoch": 8.077183955092972,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 34530
    },
    {
      "epoch": 8.07741784586598,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 34531
    },
    {
      "epoch": 8.07765173663899,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5912,
      "step": 34532
    },
    {
      "epoch": 8.077885627411998,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 34533
    },
    {
      "epoch": 8.078119518185007,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4217,
      "step": 34534
    },
    {
      "epoch": 8.078353408958016,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5948,
      "step": 34535
    },
    {
      "epoch": 8.078587299731026,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 34536
    },
    {
      "epoch": 8.078821190504035,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5247,
      "step": 34537
    },
    {
      "epoch": 8.079055081277044,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 34538
    },
    {
      "epoch": 8.079288972050053,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 34539
    },
    {
      "epoch": 8.079522862823062,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7769,
      "step": 34540
    },
    {
      "epoch": 8.07975675359607,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 34541
    },
    {
      "epoch": 8.07999064436908,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.584,
      "step": 34542
    },
    {
      "epoch": 8.080224535142088,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 34543
    },
    {
      "epoch": 8.080458425915097,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 34544
    },
    {
      "epoch": 8.080692316688106,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 34545
    },
    {
      "epoch": 8.080926207461115,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7278,
      "step": 34546
    },
    {
      "epoch": 8.081160098234125,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 34547
    },
    {
      "epoch": 8.081393989007134,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7392,
      "step": 34548
    },
    {
      "epoch": 8.081627879780143,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 34549
    },
    {
      "epoch": 8.081861770553152,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6951,
      "step": 34550
    },
    {
      "epoch": 8.08209566132616,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.718,
      "step": 34551
    },
    {
      "epoch": 8.08232955209917,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.752,
      "step": 34552
    },
    {
      "epoch": 8.082563442872178,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7999,
      "step": 34553
    },
    {
      "epoch": 8.082797333645187,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5797,
      "step": 34554
    },
    {
      "epoch": 8.083031224418196,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 34555
    },
    {
      "epoch": 8.083265115191205,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 34556
    },
    {
      "epoch": 8.083499005964216,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6121,
      "step": 34557
    },
    {
      "epoch": 8.083732896737224,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6724,
      "step": 34558
    },
    {
      "epoch": 8.083966787510233,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.569,
      "step": 34559
    },
    {
      "epoch": 8.084200678283242,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 34560
    },
    {
      "epoch": 8.084434569056251,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 34561
    },
    {
      "epoch": 8.08466845982926,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 34562
    },
    {
      "epoch": 8.084902350602269,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6004,
      "step": 34563
    },
    {
      "epoch": 8.085136241375277,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6084,
      "step": 34564
    },
    {
      "epoch": 8.085370132148286,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 34565
    },
    {
      "epoch": 8.085604022921295,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 34566
    },
    {
      "epoch": 8.085837913694304,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 34567
    },
    {
      "epoch": 8.086071804467315,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4014,
      "step": 34568
    },
    {
      "epoch": 8.086305695240323,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 34569
    },
    {
      "epoch": 8.086539586013332,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 34570
    },
    {
      "epoch": 8.086773476786341,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 34571
    },
    {
      "epoch": 8.08700736755935,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6983,
      "step": 34572
    },
    {
      "epoch": 8.087241258332359,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 34573
    },
    {
      "epoch": 8.087475149105368,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 34574
    },
    {
      "epoch": 8.087709039878376,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4231,
      "step": 34575
    },
    {
      "epoch": 8.087942930651385,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6411,
      "step": 34576
    },
    {
      "epoch": 8.088176821424394,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6464,
      "step": 34577
    },
    {
      "epoch": 8.088410712197405,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.53,
      "step": 34578
    },
    {
      "epoch": 8.088644602970414,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 34579
    },
    {
      "epoch": 8.088878493743422,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5294,
      "step": 34580
    },
    {
      "epoch": 8.089112384516431,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 34581
    },
    {
      "epoch": 8.08934627528944,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 34582
    },
    {
      "epoch": 8.089580166062449,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6597,
      "step": 34583
    },
    {
      "epoch": 8.089814056835458,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5162,
      "step": 34584
    },
    {
      "epoch": 8.090047947608467,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 34585
    },
    {
      "epoch": 8.090281838381475,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6851,
      "step": 34586
    },
    {
      "epoch": 8.090515729154484,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8026,
      "step": 34587
    },
    {
      "epoch": 8.090749619927493,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6229,
      "step": 34588
    },
    {
      "epoch": 8.090983510700504,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 34589
    },
    {
      "epoch": 8.091217401473513,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5921,
      "step": 34590
    },
    {
      "epoch": 8.091451292246521,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 34591
    },
    {
      "epoch": 8.09168518301953,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 34592
    },
    {
      "epoch": 8.091919073792539,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6473,
      "step": 34593
    },
    {
      "epoch": 8.092152964565548,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.741,
      "step": 34594
    },
    {
      "epoch": 8.092386855338557,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 34595
    },
    {
      "epoch": 8.092620746111566,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 34596
    },
    {
      "epoch": 8.092854636884574,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 34597
    },
    {
      "epoch": 8.093088527657583,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 34598
    },
    {
      "epoch": 8.093322418430592,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 34599
    },
    {
      "epoch": 8.093556309203603,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 34600
    },
    {
      "epoch": 8.093556309203603,
      "eval_runtime": 4.6,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 34600
    },
    {
      "epoch": 8.093790199976612,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6502,
      "step": 34601
    },
    {
      "epoch": 8.09402409074962,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 34602
    },
    {
      "epoch": 8.09425798152263,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 34603
    },
    {
      "epoch": 8.094491872295638,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 34604
    },
    {
      "epoch": 8.094725763068647,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7351,
      "step": 34605
    },
    {
      "epoch": 8.094959653841656,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5786,
      "step": 34606
    },
    {
      "epoch": 8.095193544614665,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 34607
    },
    {
      "epoch": 8.095427435387673,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 34608
    },
    {
      "epoch": 8.095661326160682,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.4819,
      "step": 34609
    },
    {
      "epoch": 8.095895216933691,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 34610
    },
    {
      "epoch": 8.096129107706702,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4096,
      "step": 34611
    },
    {
      "epoch": 8.09636299847971,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5825,
      "step": 34612
    },
    {
      "epoch": 8.09659688925272,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 34613
    },
    {
      "epoch": 8.096830780025728,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.581,
      "step": 34614
    },
    {
      "epoch": 8.097064670798737,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 34615
    },
    {
      "epoch": 8.097298561571746,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 34616
    },
    {
      "epoch": 8.097532452344755,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 34617
    },
    {
      "epoch": 8.097766343117764,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.5895,
      "step": 34618
    },
    {
      "epoch": 8.098000233890772,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 34619
    },
    {
      "epoch": 8.098234124663781,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6337,
      "step": 34620
    },
    {
      "epoch": 8.098468015436792,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 34621
    },
    {
      "epoch": 8.0987019062098,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.0265,
      "step": 34622
    },
    {
      "epoch": 8.09893579698281,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5739,
      "step": 34623
    },
    {
      "epoch": 8.099169687755818,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 34624
    },
    {
      "epoch": 8.099403578528827,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6239,
      "step": 34625
    },
    {
      "epoch": 8.099637469301836,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 34626
    },
    {
      "epoch": 8.099871360074845,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5952,
      "step": 34627
    },
    {
      "epoch": 8.100105250847854,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 34628
    },
    {
      "epoch": 8.100339141620863,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.4179,
      "step": 34629
    },
    {
      "epoch": 8.100573032393871,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.4941,
      "step": 34630
    },
    {
      "epoch": 8.10080692316688,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 34631
    },
    {
      "epoch": 8.10104081393989,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 34632
    },
    {
      "epoch": 8.1012747047129,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 34633
    },
    {
      "epoch": 8.101508595485909,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 34634
    },
    {
      "epoch": 8.101742486258917,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 34635
    },
    {
      "epoch": 8.101976377031926,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 34636
    },
    {
      "epoch": 8.102210267804935,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 34637
    },
    {
      "epoch": 8.102444158577944,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5212,
      "step": 34638
    },
    {
      "epoch": 8.102678049350953,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6828,
      "step": 34639
    },
    {
      "epoch": 8.102911940123962,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.7827,
      "step": 34640
    },
    {
      "epoch": 8.10314583089697,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6783,
      "step": 34641
    },
    {
      "epoch": 8.103379721669981,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 34642
    },
    {
      "epoch": 8.10361361244299,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8151,
      "step": 34643
    },
    {
      "epoch": 8.103847503215999,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 34644
    },
    {
      "epoch": 8.104081393989008,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5621,
      "step": 34645
    },
    {
      "epoch": 8.104315284762016,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 34646
    },
    {
      "epoch": 8.104549175535025,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 34647
    },
    {
      "epoch": 8.104783066308034,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6577,
      "step": 34648
    },
    {
      "epoch": 8.105016957081043,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 34649
    },
    {
      "epoch": 8.105250847854052,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5051,
      "step": 34650
    },
    {
      "epoch": 8.10548473862706,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 34651
    },
    {
      "epoch": 8.10571862940007,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 34652
    },
    {
      "epoch": 8.10595252017308,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6986,
      "step": 34653
    },
    {
      "epoch": 8.106186410946089,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7365,
      "step": 34654
    },
    {
      "epoch": 8.106420301719098,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 34655
    },
    {
      "epoch": 8.106654192492107,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 34656
    },
    {
      "epoch": 8.106888083265115,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 34657
    },
    {
      "epoch": 8.107121974038124,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 34658
    },
    {
      "epoch": 8.107355864811133,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7468,
      "step": 34659
    },
    {
      "epoch": 8.107589755584142,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5555,
      "step": 34660
    },
    {
      "epoch": 8.10782364635715,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 34661
    },
    {
      "epoch": 8.10805753713016,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 34662
    },
    {
      "epoch": 8.108291427903168,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 34663
    },
    {
      "epoch": 8.108525318676179,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 34664
    },
    {
      "epoch": 8.108759209449188,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 34665
    },
    {
      "epoch": 8.108993100222197,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 34666
    },
    {
      "epoch": 8.109226990995205,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 34667
    },
    {
      "epoch": 8.109460881768214,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 34668
    },
    {
      "epoch": 8.109694772541223,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 34669
    },
    {
      "epoch": 8.109928663314232,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 34670
    },
    {
      "epoch": 8.11016255408724,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.608,
      "step": 34671
    },
    {
      "epoch": 8.11039644486025,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 34672
    },
    {
      "epoch": 8.110630335633259,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 34673
    },
    {
      "epoch": 8.11086422640627,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 34674
    },
    {
      "epoch": 8.111098117179278,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 34675
    },
    {
      "epoch": 8.111332007952287,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.4578,
      "step": 34676
    },
    {
      "epoch": 8.111565898725296,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6858,
      "step": 34677
    },
    {
      "epoch": 8.111799789498304,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6197,
      "step": 34678
    },
    {
      "epoch": 8.112033680271313,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.819,
      "step": 34679
    },
    {
      "epoch": 8.112267571044322,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5219,
      "step": 34680
    },
    {
      "epoch": 8.112501461817331,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.4836,
      "step": 34681
    },
    {
      "epoch": 8.11273535259034,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 34682
    },
    {
      "epoch": 8.112969243363349,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 34683
    },
    {
      "epoch": 8.113203134136358,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 34684
    },
    {
      "epoch": 8.113437024909368,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 34685
    },
    {
      "epoch": 8.113670915682377,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7135,
      "step": 34686
    },
    {
      "epoch": 8.113904806455386,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7271,
      "step": 34687
    },
    {
      "epoch": 8.114138697228395,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6873,
      "step": 34688
    },
    {
      "epoch": 8.114372588001403,
      "grad_norm": 8.75,
      "learning_rate": 3e-05,
      "loss": 2.3682,
      "step": 34689
    },
    {
      "epoch": 8.114606478774412,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 34690
    },
    {
      "epoch": 8.114840369547421,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.571,
      "step": 34691
    },
    {
      "epoch": 8.11507426032043,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 34692
    },
    {
      "epoch": 8.115308151093439,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 34693
    },
    {
      "epoch": 8.115542041866448,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.3394,
      "step": 34694
    },
    {
      "epoch": 8.115775932639457,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 34695
    },
    {
      "epoch": 8.116009823412467,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 34696
    },
    {
      "epoch": 8.116243714185476,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 34697
    },
    {
      "epoch": 8.116477604958485,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 34698
    },
    {
      "epoch": 8.116711495731494,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6031,
      "step": 34699
    },
    {
      "epoch": 8.116945386504502,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6367,
      "step": 34700
    },
    {
      "epoch": 8.116945386504502,
      "eval_runtime": 4.584,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 34700
    },
    {
      "epoch": 8.117179277277511,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 34701
    },
    {
      "epoch": 8.11741316805052,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 34702
    },
    {
      "epoch": 8.117647058823529,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 34703
    },
    {
      "epoch": 8.117880949596538,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 34704
    },
    {
      "epoch": 8.118114840369547,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.659,
      "step": 34705
    },
    {
      "epoch": 8.118348731142557,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 34706
    },
    {
      "epoch": 8.118582621915566,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5326,
      "step": 34707
    },
    {
      "epoch": 8.118816512688575,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 34708
    },
    {
      "epoch": 8.119050403461584,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 34709
    },
    {
      "epoch": 8.119284294234593,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 34710
    },
    {
      "epoch": 8.119518185007601,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.3219,
      "step": 34711
    },
    {
      "epoch": 8.11975207578061,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 34712
    },
    {
      "epoch": 8.11998596655362,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 34713
    },
    {
      "epoch": 8.120219857326628,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 34714
    },
    {
      "epoch": 8.120453748099637,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6742,
      "step": 34715
    },
    {
      "epoch": 8.120687638872646,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 34716
    },
    {
      "epoch": 8.120921529645656,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6478,
      "step": 34717
    },
    {
      "epoch": 8.121155420418665,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.632,
      "step": 34718
    },
    {
      "epoch": 8.121389311191674,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 34719
    },
    {
      "epoch": 8.121623201964683,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 34720
    },
    {
      "epoch": 8.121857092737692,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6272,
      "step": 34721
    },
    {
      "epoch": 8.1220909835107,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4089,
      "step": 34722
    },
    {
      "epoch": 8.12232487428371,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 34723
    },
    {
      "epoch": 8.122558765056718,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6272,
      "step": 34724
    },
    {
      "epoch": 8.122792655829727,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6358,
      "step": 34725
    },
    {
      "epoch": 8.123026546602736,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 34726
    },
    {
      "epoch": 8.123260437375745,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5026,
      "step": 34727
    },
    {
      "epoch": 8.123494328148755,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 34728
    },
    {
      "epoch": 8.123728218921764,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5767,
      "step": 34729
    },
    {
      "epoch": 8.123962109694773,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 34730
    },
    {
      "epoch": 8.124196000467782,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 34731
    },
    {
      "epoch": 8.12442989124079,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 34732
    },
    {
      "epoch": 8.1246637820138,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 34733
    },
    {
      "epoch": 8.124897672786808,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 34734
    },
    {
      "epoch": 8.125131563559817,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6726,
      "step": 34735
    },
    {
      "epoch": 8.125365454332826,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 34736
    },
    {
      "epoch": 8.125599345105835,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 34737
    },
    {
      "epoch": 8.125833235878845,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 34738
    },
    {
      "epoch": 8.126067126651854,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.426,
      "step": 34739
    },
    {
      "epoch": 8.126301017424863,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 34740
    },
    {
      "epoch": 8.126534908197872,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 34741
    },
    {
      "epoch": 8.12676879897088,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5351,
      "step": 34742
    },
    {
      "epoch": 8.12700268974389,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.3592,
      "step": 34743
    },
    {
      "epoch": 8.127236580516898,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5925,
      "step": 34744
    },
    {
      "epoch": 8.127470471289907,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6176,
      "step": 34745
    },
    {
      "epoch": 8.127704362062916,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 34746
    },
    {
      "epoch": 8.127938252835925,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 34747
    },
    {
      "epoch": 8.128172143608934,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5853,
      "step": 34748
    },
    {
      "epoch": 8.128406034381944,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 34749
    },
    {
      "epoch": 8.128639925154953,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5254,
      "step": 34750
    },
    {
      "epoch": 8.128873815927962,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 34751
    },
    {
      "epoch": 8.129107706700971,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7165,
      "step": 34752
    },
    {
      "epoch": 8.12934159747398,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 34753
    },
    {
      "epoch": 8.129575488246989,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6065,
      "step": 34754
    },
    {
      "epoch": 8.129809379019997,
      "grad_norm": 9.5625,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 34755
    },
    {
      "epoch": 8.130043269793006,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8135,
      "step": 34756
    },
    {
      "epoch": 8.130277160566015,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 34757
    },
    {
      "epoch": 8.130511051339024,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8801,
      "step": 34758
    },
    {
      "epoch": 8.130744942112035,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 34759
    },
    {
      "epoch": 8.130978832885043,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6077,
      "step": 34760
    },
    {
      "epoch": 8.131212723658052,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7737,
      "step": 34761
    },
    {
      "epoch": 8.131446614431061,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 34762
    },
    {
      "epoch": 8.13168050520407,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4846,
      "step": 34763
    },
    {
      "epoch": 8.131914395977079,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6186,
      "step": 34764
    },
    {
      "epoch": 8.132148286750088,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 34765
    },
    {
      "epoch": 8.132382177523096,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 34766
    },
    {
      "epoch": 8.132616068296105,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7342,
      "step": 34767
    },
    {
      "epoch": 8.132849959069114,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 34768
    },
    {
      "epoch": 8.133083849842123,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 34769
    },
    {
      "epoch": 8.133317740615134,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 34770
    },
    {
      "epoch": 8.133551631388142,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7862,
      "step": 34771
    },
    {
      "epoch": 8.133785522161151,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.3588,
      "step": 34772
    },
    {
      "epoch": 8.13401941293416,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.4069,
      "step": 34773
    },
    {
      "epoch": 8.134253303707169,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 34774
    },
    {
      "epoch": 8.134487194480178,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6539,
      "step": 34775
    },
    {
      "epoch": 8.134721085253187,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 34776
    },
    {
      "epoch": 8.134954976026195,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6365,
      "step": 34777
    },
    {
      "epoch": 8.135188866799204,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 34778
    },
    {
      "epoch": 8.135422757572213,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 34779
    },
    {
      "epoch": 8.135656648345222,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 34780
    },
    {
      "epoch": 8.135890539118233,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 34781
    },
    {
      "epoch": 8.136124429891241,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6201,
      "step": 34782
    },
    {
      "epoch": 8.13635832066425,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.3521,
      "step": 34783
    },
    {
      "epoch": 8.136592211437259,
      "grad_norm": 2.515625,
      "learning_rate": 3e-05,
      "loss": 1.5696,
      "step": 34784
    },
    {
      "epoch": 8.136826102210268,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 34785
    },
    {
      "epoch": 8.137059992983277,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.5833,
      "step": 34786
    },
    {
      "epoch": 8.137293883756286,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.3796,
      "step": 34787
    },
    {
      "epoch": 8.137527774529294,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.702,
      "step": 34788
    },
    {
      "epoch": 8.137761665302303,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6958,
      "step": 34789
    },
    {
      "epoch": 8.137995556075312,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 34790
    },
    {
      "epoch": 8.138229446848321,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6502,
      "step": 34791
    },
    {
      "epoch": 8.138463337621332,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7163,
      "step": 34792
    },
    {
      "epoch": 8.13869722839434,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 34793
    },
    {
      "epoch": 8.13893111916735,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 34794
    },
    {
      "epoch": 8.139165009940358,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4899,
      "step": 34795
    },
    {
      "epoch": 8.139398900713367,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 34796
    },
    {
      "epoch": 8.139632791486376,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 34797
    },
    {
      "epoch": 8.139866682259385,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.775,
      "step": 34798
    },
    {
      "epoch": 8.140100573032393,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 34799
    },
    {
      "epoch": 8.140334463805402,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 34800
    },
    {
      "epoch": 8.140334463805402,
      "eval_runtime": 4.6327,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 34800
    },
    {
      "epoch": 8.140568354578411,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 34801
    },
    {
      "epoch": 8.140802245351422,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7058,
      "step": 34802
    },
    {
      "epoch": 8.14103613612443,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.4944,
      "step": 34803
    },
    {
      "epoch": 8.14127002689744,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 34804
    },
    {
      "epoch": 8.141503917670448,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 34805
    },
    {
      "epoch": 8.141737808443457,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 34806
    },
    {
      "epoch": 8.141971699216466,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5908,
      "step": 34807
    },
    {
      "epoch": 8.142205589989475,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6136,
      "step": 34808
    },
    {
      "epoch": 8.142439480762484,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6935,
      "step": 34809
    },
    {
      "epoch": 8.142673371535492,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4953,
      "step": 34810
    },
    {
      "epoch": 8.142907262308501,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 34811
    },
    {
      "epoch": 8.14314115308151,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 34812
    },
    {
      "epoch": 8.14337504385452,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 34813
    },
    {
      "epoch": 8.14360893462753,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 34814
    },
    {
      "epoch": 8.143842825400538,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.3692,
      "step": 34815
    },
    {
      "epoch": 8.144076716173547,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6068,
      "step": 34816
    },
    {
      "epoch": 8.144310606946556,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5383,
      "step": 34817
    },
    {
      "epoch": 8.144544497719565,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 34818
    },
    {
      "epoch": 8.144778388492574,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4704,
      "step": 34819
    },
    {
      "epoch": 8.145012279265583,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7005,
      "step": 34820
    },
    {
      "epoch": 8.145246170038591,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6612,
      "step": 34821
    },
    {
      "epoch": 8.1454800608116,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 34822
    },
    {
      "epoch": 8.14571395158461,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 34823
    },
    {
      "epoch": 8.14594784235762,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5741,
      "step": 34824
    },
    {
      "epoch": 8.146181733130629,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 34825
    },
    {
      "epoch": 8.146415623903637,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 34826
    },
    {
      "epoch": 8.146649514676646,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 34827
    },
    {
      "epoch": 8.146883405449655,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 34828
    },
    {
      "epoch": 8.147117296222664,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 34829
    },
    {
      "epoch": 8.147351186995673,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 34830
    },
    {
      "epoch": 8.147585077768682,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6841,
      "step": 34831
    },
    {
      "epoch": 8.14781896854169,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6386,
      "step": 34832
    },
    {
      "epoch": 8.1480528593147,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6279,
      "step": 34833
    },
    {
      "epoch": 8.14828675008771,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 34834
    },
    {
      "epoch": 8.148520640860719,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.6572,
      "step": 34835
    },
    {
      "epoch": 8.148754531633728,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 34836
    },
    {
      "epoch": 8.148988422406736,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 34837
    },
    {
      "epoch": 8.149222313179745,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 34838
    },
    {
      "epoch": 8.149456203952754,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7598,
      "step": 34839
    },
    {
      "epoch": 8.149690094725763,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 34840
    },
    {
      "epoch": 8.149923985498772,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 34841
    },
    {
      "epoch": 8.15015787627178,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5364,
      "step": 34842
    },
    {
      "epoch": 8.15039176704479,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 34843
    },
    {
      "epoch": 8.150625657817798,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 34844
    },
    {
      "epoch": 8.150859548590809,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5913,
      "step": 34845
    },
    {
      "epoch": 8.151093439363818,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 34846
    },
    {
      "epoch": 8.151327330136827,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7254,
      "step": 34847
    },
    {
      "epoch": 8.151561220909835,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 34848
    },
    {
      "epoch": 8.151795111682844,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.5145,
      "step": 34849
    },
    {
      "epoch": 8.152029002455853,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6449,
      "step": 34850
    },
    {
      "epoch": 8.152262893228862,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5205,
      "step": 34851
    },
    {
      "epoch": 8.15249678400187,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 34852
    },
    {
      "epoch": 8.15273067477488,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5273,
      "step": 34853
    },
    {
      "epoch": 8.152964565547888,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6945,
      "step": 34854
    },
    {
      "epoch": 8.153198456320899,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 34855
    },
    {
      "epoch": 8.153432347093908,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6756,
      "step": 34856
    },
    {
      "epoch": 8.153666237866917,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7552,
      "step": 34857
    },
    {
      "epoch": 8.153900128639926,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6893,
      "step": 34858
    },
    {
      "epoch": 8.154134019412934,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 34859
    },
    {
      "epoch": 8.154367910185943,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.641,
      "step": 34860
    },
    {
      "epoch": 8.154601800958952,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 34861
    },
    {
      "epoch": 8.15483569173196,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 34862
    },
    {
      "epoch": 8.15506958250497,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 34863
    },
    {
      "epoch": 8.155303473277979,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4668,
      "step": 34864
    },
    {
      "epoch": 8.155537364050987,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 34865
    },
    {
      "epoch": 8.155771254823998,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8361,
      "step": 34866
    },
    {
      "epoch": 8.156005145597007,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 34867
    },
    {
      "epoch": 8.156239036370016,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5849,
      "step": 34868
    },
    {
      "epoch": 8.156472927143025,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 34869
    },
    {
      "epoch": 8.156706817916033,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 34870
    },
    {
      "epoch": 8.156940708689042,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 34871
    },
    {
      "epoch": 8.157174599462051,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.498,
      "step": 34872
    },
    {
      "epoch": 8.15740849023506,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 34873
    },
    {
      "epoch": 8.157642381008069,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5871,
      "step": 34874
    },
    {
      "epoch": 8.157876271781078,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 34875
    },
    {
      "epoch": 8.158110162554086,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7226,
      "step": 34876
    },
    {
      "epoch": 8.158344053327097,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6941,
      "step": 34877
    },
    {
      "epoch": 8.158577944100106,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5644,
      "step": 34878
    },
    {
      "epoch": 8.158811834873115,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 34879
    },
    {
      "epoch": 8.159045725646124,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7729,
      "step": 34880
    },
    {
      "epoch": 8.159279616419132,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.731,
      "step": 34881
    },
    {
      "epoch": 8.159513507192141,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 34882
    },
    {
      "epoch": 8.15974739796515,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 34883
    },
    {
      "epoch": 8.159981288738159,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 34884
    },
    {
      "epoch": 8.160215179511168,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5261,
      "step": 34885
    },
    {
      "epoch": 8.160449070284177,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6136,
      "step": 34886
    },
    {
      "epoch": 8.160682961057187,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 34887
    },
    {
      "epoch": 8.160916851830196,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 34888
    },
    {
      "epoch": 8.161150742603205,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6321,
      "step": 34889
    },
    {
      "epoch": 8.161384633376214,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6039,
      "step": 34890
    },
    {
      "epoch": 8.161618524149223,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6143,
      "step": 34891
    },
    {
      "epoch": 8.161852414922231,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5461,
      "step": 34892
    },
    {
      "epoch": 8.16208630569524,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7231,
      "step": 34893
    },
    {
      "epoch": 8.162320196468249,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 34894
    },
    {
      "epoch": 8.162554087241258,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6584,
      "step": 34895
    },
    {
      "epoch": 8.162787978014267,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 34896
    },
    {
      "epoch": 8.163021868787276,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6201,
      "step": 34897
    },
    {
      "epoch": 8.163255759560286,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 34898
    },
    {
      "epoch": 8.163489650333295,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 34899
    },
    {
      "epoch": 8.163723541106304,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8424,
      "step": 34900
    },
    {
      "epoch": 8.163723541106304,
      "eval_runtime": 4.6076,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 34900
    },
    {
      "epoch": 8.163957431879313,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 34901
    },
    {
      "epoch": 8.164191322652322,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 34902
    },
    {
      "epoch": 8.16442521342533,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6858,
      "step": 34903
    },
    {
      "epoch": 8.16465910419834,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.726,
      "step": 34904
    },
    {
      "epoch": 8.164892994971348,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7444,
      "step": 34905
    },
    {
      "epoch": 8.165126885744357,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 34906
    },
    {
      "epoch": 8.165360776517366,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 34907
    },
    {
      "epoch": 8.165594667290375,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 34908
    },
    {
      "epoch": 8.165828558063385,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.526,
      "step": 34909
    },
    {
      "epoch": 8.166062448836394,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6524,
      "step": 34910
    },
    {
      "epoch": 8.166296339609403,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 34911
    },
    {
      "epoch": 8.166530230382412,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 34912
    },
    {
      "epoch": 8.16676412115542,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 34913
    },
    {
      "epoch": 8.16699801192843,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 34914
    },
    {
      "epoch": 8.167231902701438,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 34915
    },
    {
      "epoch": 8.167465793474447,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 34916
    },
    {
      "epoch": 8.167699684247456,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6113,
      "step": 34917
    },
    {
      "epoch": 8.167933575020465,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 34918
    },
    {
      "epoch": 8.168167465793475,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5358,
      "step": 34919
    },
    {
      "epoch": 8.168401356566484,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 34920
    },
    {
      "epoch": 8.168635247339493,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6898,
      "step": 34921
    },
    {
      "epoch": 8.168869138112502,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5611,
      "step": 34922
    },
    {
      "epoch": 8.16910302888551,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6309,
      "step": 34923
    },
    {
      "epoch": 8.16933691965852,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.3932,
      "step": 34924
    },
    {
      "epoch": 8.169570810431528,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.3791,
      "step": 34925
    },
    {
      "epoch": 8.169804701204537,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.493,
      "step": 34926
    },
    {
      "epoch": 8.170038591977546,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6118,
      "step": 34927
    },
    {
      "epoch": 8.170272482750555,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 34928
    },
    {
      "epoch": 8.170506373523564,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 34929
    },
    {
      "epoch": 8.170740264296574,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.4262,
      "step": 34930
    },
    {
      "epoch": 8.170974155069583,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 34931
    },
    {
      "epoch": 8.171208045842592,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5639,
      "step": 34932
    },
    {
      "epoch": 8.1714419366156,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 34933
    },
    {
      "epoch": 8.17167582738861,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 34934
    },
    {
      "epoch": 8.171909718161618,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5738,
      "step": 34935
    },
    {
      "epoch": 8.172143608934627,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 34936
    },
    {
      "epoch": 8.172377499707636,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6486,
      "step": 34937
    },
    {
      "epoch": 8.172611390480645,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 34938
    },
    {
      "epoch": 8.172845281253654,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 34939
    },
    {
      "epoch": 8.173079172026663,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 34940
    },
    {
      "epoch": 8.173313062799673,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.5489,
      "step": 34941
    },
    {
      "epoch": 8.173546953572682,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6572,
      "step": 34942
    },
    {
      "epoch": 8.173780844345691,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 34943
    },
    {
      "epoch": 8.1740147351187,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5114,
      "step": 34944
    },
    {
      "epoch": 8.174248625891709,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7333,
      "step": 34945
    },
    {
      "epoch": 8.174482516664717,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 34946
    },
    {
      "epoch": 8.174716407437726,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 34947
    },
    {
      "epoch": 8.174950298210735,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 34948
    },
    {
      "epoch": 8.175184188983744,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 34949
    },
    {
      "epoch": 8.175418079756753,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 34950
    },
    {
      "epoch": 8.175651970529763,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 34951
    },
    {
      "epoch": 8.175885861302772,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 34952
    },
    {
      "epoch": 8.176119752075781,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6644,
      "step": 34953
    },
    {
      "epoch": 8.17635364284879,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5283,
      "step": 34954
    },
    {
      "epoch": 8.176587533621799,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 34955
    },
    {
      "epoch": 8.176821424394808,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6781,
      "step": 34956
    },
    {
      "epoch": 8.177055315167816,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 34957
    },
    {
      "epoch": 8.177289205940825,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5712,
      "step": 34958
    },
    {
      "epoch": 8.177523096713834,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6948,
      "step": 34959
    },
    {
      "epoch": 8.177756987486843,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 34960
    },
    {
      "epoch": 8.177990878259852,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 34961
    },
    {
      "epoch": 8.178224769032862,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 34962
    },
    {
      "epoch": 8.178458659805871,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 34963
    },
    {
      "epoch": 8.17869255057888,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 34964
    },
    {
      "epoch": 8.178926441351889,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4765,
      "step": 34965
    },
    {
      "epoch": 8.179160332124898,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.951,
      "step": 34966
    },
    {
      "epoch": 8.179394222897907,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6022,
      "step": 34967
    },
    {
      "epoch": 8.179628113670915,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6643,
      "step": 34968
    },
    {
      "epoch": 8.179862004443924,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 34969
    },
    {
      "epoch": 8.180095895216933,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6316,
      "step": 34970
    },
    {
      "epoch": 8.180329785989942,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.5752,
      "step": 34971
    },
    {
      "epoch": 8.18056367676295,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 34972
    },
    {
      "epoch": 8.180797567535961,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 34973
    },
    {
      "epoch": 8.18103145830897,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.3762,
      "step": 34974
    },
    {
      "epoch": 8.181265349081979,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 34975
    },
    {
      "epoch": 8.181499239854988,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.5435,
      "step": 34976
    },
    {
      "epoch": 8.181733130627997,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 34977
    },
    {
      "epoch": 8.181967021401006,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.76,
      "step": 34978
    },
    {
      "epoch": 8.182200912174014,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.3958,
      "step": 34979
    },
    {
      "epoch": 8.182434802947023,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 34980
    },
    {
      "epoch": 8.182668693720032,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 34981
    },
    {
      "epoch": 8.182902584493041,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 34982
    },
    {
      "epoch": 8.183136475266052,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5086,
      "step": 34983
    },
    {
      "epoch": 8.18337036603906,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.421,
      "step": 34984
    },
    {
      "epoch": 8.18360425681207,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 34985
    },
    {
      "epoch": 8.183838147585078,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.655,
      "step": 34986
    },
    {
      "epoch": 8.184072038358087,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 34987
    },
    {
      "epoch": 8.184305929131096,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.596,
      "step": 34988
    },
    {
      "epoch": 8.184539819904105,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5542,
      "step": 34989
    },
    {
      "epoch": 8.184773710677113,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6501,
      "step": 34990
    },
    {
      "epoch": 8.185007601450122,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7624,
      "step": 34991
    },
    {
      "epoch": 8.185241492223131,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5068,
      "step": 34992
    },
    {
      "epoch": 8.18547538299614,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 34993
    },
    {
      "epoch": 8.18570927376915,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 34994
    },
    {
      "epoch": 8.18594316454216,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5426,
      "step": 34995
    },
    {
      "epoch": 8.186177055315168,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6382,
      "step": 34996
    },
    {
      "epoch": 8.186410946088177,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.155,
      "step": 34997
    },
    {
      "epoch": 8.186644836861186,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6871,
      "step": 34998
    },
    {
      "epoch": 8.186878727634195,
      "grad_norm": 10.875,
      "learning_rate": 3e-05,
      "loss": 1.7978,
      "step": 34999
    },
    {
      "epoch": 8.187112618407204,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 35000
    },
    {
      "epoch": 8.187112618407204,
      "eval_runtime": 4.6293,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 35000
    },
    {
      "epoch": 8.187346509180212,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 35001
    },
    {
      "epoch": 8.187580399953221,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 35002
    },
    {
      "epoch": 8.18781429072623,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 35003
    },
    {
      "epoch": 8.18804818149924,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 35004
    },
    {
      "epoch": 8.18828207227225,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 35005
    },
    {
      "epoch": 8.188515963045258,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7184,
      "step": 35006
    },
    {
      "epoch": 8.188749853818267,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 35007
    },
    {
      "epoch": 8.188983744591276,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 35008
    },
    {
      "epoch": 8.189217635364285,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 35009
    },
    {
      "epoch": 8.189451526137294,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 35010
    },
    {
      "epoch": 8.189685416910303,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 35011
    },
    {
      "epoch": 8.189919307683311,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6797,
      "step": 35012
    },
    {
      "epoch": 8.19015319845632,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6921,
      "step": 35013
    },
    {
      "epoch": 8.19038708922933,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 35014
    },
    {
      "epoch": 8.19062098000234,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 35015
    },
    {
      "epoch": 8.190854870775349,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 35016
    },
    {
      "epoch": 8.191088761548357,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 35017
    },
    {
      "epoch": 8.191322652321366,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 35018
    },
    {
      "epoch": 8.191556543094375,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6344,
      "step": 35019
    },
    {
      "epoch": 8.191790433867384,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 35020
    },
    {
      "epoch": 8.192024324640393,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.68,
      "step": 35021
    },
    {
      "epoch": 8.192258215413402,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 35022
    },
    {
      "epoch": 8.19249210618641,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.5438,
      "step": 35023
    },
    {
      "epoch": 8.19272599695942,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5244,
      "step": 35024
    },
    {
      "epoch": 8.192959887732428,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 35025
    },
    {
      "epoch": 8.193193778505439,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 35026
    },
    {
      "epoch": 8.193427669278448,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 35027
    },
    {
      "epoch": 8.193661560051456,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 35028
    },
    {
      "epoch": 8.193895450824465,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5503,
      "step": 35029
    },
    {
      "epoch": 8.194129341597474,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 35030
    },
    {
      "epoch": 8.194363232370483,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 35031
    },
    {
      "epoch": 8.194597123143492,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.567,
      "step": 35032
    },
    {
      "epoch": 8.1948310139165,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6719,
      "step": 35033
    },
    {
      "epoch": 8.19506490468951,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6803,
      "step": 35034
    },
    {
      "epoch": 8.195298795462518,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5693,
      "step": 35035
    },
    {
      "epoch": 8.195532686235527,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 35036
    },
    {
      "epoch": 8.195766577008538,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 35037
    },
    {
      "epoch": 8.196000467781547,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 35038
    },
    {
      "epoch": 8.196234358554555,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.782,
      "step": 35039
    },
    {
      "epoch": 8.196468249327564,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6412,
      "step": 35040
    },
    {
      "epoch": 8.196702140100573,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.669,
      "step": 35041
    },
    {
      "epoch": 8.196936030873582,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 35042
    },
    {
      "epoch": 8.19716992164659,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.5551,
      "step": 35043
    },
    {
      "epoch": 8.1974038124196,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 35044
    },
    {
      "epoch": 8.197637703192608,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 35045
    },
    {
      "epoch": 8.197871593965617,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 35046
    },
    {
      "epoch": 8.198105484738628,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5782,
      "step": 35047
    },
    {
      "epoch": 8.198339375511637,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5884,
      "step": 35048
    },
    {
      "epoch": 8.198573266284646,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 35049
    },
    {
      "epoch": 8.198807157057654,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 35050
    },
    {
      "epoch": 8.199041047830663,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 35051
    },
    {
      "epoch": 8.199274938603672,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6882,
      "step": 35052
    },
    {
      "epoch": 8.199508829376681,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5921,
      "step": 35053
    },
    {
      "epoch": 8.19974272014969,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 35054
    },
    {
      "epoch": 8.199976610922699,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6912,
      "step": 35055
    },
    {
      "epoch": 8.200210501695707,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7439,
      "step": 35056
    },
    {
      "epoch": 8.200444392468716,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 35057
    },
    {
      "epoch": 8.200678283241727,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 35058
    },
    {
      "epoch": 8.200912174014736,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6591,
      "step": 35059
    },
    {
      "epoch": 8.201146064787745,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6468,
      "step": 35060
    },
    {
      "epoch": 8.201379955560753,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6206,
      "step": 35061
    },
    {
      "epoch": 8.201613846333762,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 35062
    },
    {
      "epoch": 8.201847737106771,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 35063
    },
    {
      "epoch": 8.20208162787978,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4138,
      "step": 35064
    },
    {
      "epoch": 8.202315518652789,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4817,
      "step": 35065
    },
    {
      "epoch": 8.202549409425798,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 35066
    },
    {
      "epoch": 8.202783300198806,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 35067
    },
    {
      "epoch": 8.203017190971817,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 35068
    },
    {
      "epoch": 8.203251081744826,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 35069
    },
    {
      "epoch": 8.203484972517835,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6519,
      "step": 35070
    },
    {
      "epoch": 8.203718863290844,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 35071
    },
    {
      "epoch": 8.203952754063852,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 35072
    },
    {
      "epoch": 8.204186644836861,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 35073
    },
    {
      "epoch": 8.20442053560987,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 35074
    },
    {
      "epoch": 8.204654426382879,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 35075
    },
    {
      "epoch": 8.204888317155888,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 35076
    },
    {
      "epoch": 8.205122207928897,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 35077
    },
    {
      "epoch": 8.205356098701905,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 35078
    },
    {
      "epoch": 8.205589989474916,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 35079
    },
    {
      "epoch": 8.205823880247925,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 35080
    },
    {
      "epoch": 8.206057771020934,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6189,
      "step": 35081
    },
    {
      "epoch": 8.206291661793943,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6253,
      "step": 35082
    },
    {
      "epoch": 8.206525552566951,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 35083
    },
    {
      "epoch": 8.20675944333996,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5201,
      "step": 35084
    },
    {
      "epoch": 8.206993334112969,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7048,
      "step": 35085
    },
    {
      "epoch": 8.207227224885978,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 35086
    },
    {
      "epoch": 8.207461115658987,
      "grad_norm": 9.1875,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 35087
    },
    {
      "epoch": 8.207695006431996,
      "grad_norm": 8.1875,
      "learning_rate": 3e-05,
      "loss": 2.344,
      "step": 35088
    },
    {
      "epoch": 8.207928897205004,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.545,
      "step": 35089
    },
    {
      "epoch": 8.208162787978015,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6559,
      "step": 35090
    },
    {
      "epoch": 8.208396678751024,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6501,
      "step": 35091
    },
    {
      "epoch": 8.208630569524033,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8434,
      "step": 35092
    },
    {
      "epoch": 8.208864460297042,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 35093
    },
    {
      "epoch": 8.20909835107005,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6436,
      "step": 35094
    },
    {
      "epoch": 8.20933224184306,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6585,
      "step": 35095
    },
    {
      "epoch": 8.209566132616068,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 35096
    },
    {
      "epoch": 8.209800023389077,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 35097
    },
    {
      "epoch": 8.210033914162086,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 35098
    },
    {
      "epoch": 8.210267804935095,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6603,
      "step": 35099
    },
    {
      "epoch": 8.210501695708105,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.514,
      "step": 35100
    },
    {
      "epoch": 8.210501695708105,
      "eval_runtime": 4.8093,
      "eval_samples_per_second": 0.208,
      "eval_steps_per_second": 0.208,
      "step": 35100
    },
    {
      "epoch": 8.210735586481114,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 35101
    },
    {
      "epoch": 8.210969477254123,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 35102
    },
    {
      "epoch": 8.211203368027132,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 35103
    },
    {
      "epoch": 8.21143725880014,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 35104
    },
    {
      "epoch": 8.21167114957315,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 35105
    },
    {
      "epoch": 8.211905040346158,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7419,
      "step": 35106
    },
    {
      "epoch": 8.212138931119167,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4361,
      "step": 35107
    },
    {
      "epoch": 8.212372821892176,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 35108
    },
    {
      "epoch": 8.212606712665185,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7532,
      "step": 35109
    },
    {
      "epoch": 8.212840603438194,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 35110
    },
    {
      "epoch": 8.213074494211204,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 35111
    },
    {
      "epoch": 8.213308384984213,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5989,
      "step": 35112
    },
    {
      "epoch": 8.213542275757222,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 35113
    },
    {
      "epoch": 8.21377616653023,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5646,
      "step": 35114
    },
    {
      "epoch": 8.21401005730324,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.5809,
      "step": 35115
    },
    {
      "epoch": 8.214243948076248,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7219,
      "step": 35116
    },
    {
      "epoch": 8.214477838849257,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.814,
      "step": 35117
    },
    {
      "epoch": 8.214711729622266,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6831,
      "step": 35118
    },
    {
      "epoch": 8.214945620395275,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 35119
    },
    {
      "epoch": 8.215179511168284,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 35120
    },
    {
      "epoch": 8.215413401941293,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6285,
      "step": 35121
    },
    {
      "epoch": 8.215647292714303,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 35122
    },
    {
      "epoch": 8.215881183487312,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 35123
    },
    {
      "epoch": 8.21611507426032,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5789,
      "step": 35124
    },
    {
      "epoch": 8.21634896503333,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5939,
      "step": 35125
    },
    {
      "epoch": 8.216582855806339,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 35126
    },
    {
      "epoch": 8.216816746579347,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 35127
    },
    {
      "epoch": 8.217050637352356,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.488,
      "step": 35128
    },
    {
      "epoch": 8.217284528125365,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6242,
      "step": 35129
    },
    {
      "epoch": 8.217518418898374,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 35130
    },
    {
      "epoch": 8.217752309671383,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 35131
    },
    {
      "epoch": 8.217986200444393,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6764,
      "step": 35132
    },
    {
      "epoch": 8.218220091217402,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5847,
      "step": 35133
    },
    {
      "epoch": 8.218453981990411,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5784,
      "step": 35134
    },
    {
      "epoch": 8.21868787276342,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5504,
      "step": 35135
    },
    {
      "epoch": 8.218921763536429,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 35136
    },
    {
      "epoch": 8.219155654309438,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 35137
    },
    {
      "epoch": 8.219389545082446,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5243,
      "step": 35138
    },
    {
      "epoch": 8.219623435855455,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 35139
    },
    {
      "epoch": 8.219857326628464,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 35140
    },
    {
      "epoch": 8.220091217401473,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7592,
      "step": 35141
    },
    {
      "epoch": 8.220325108174482,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6925,
      "step": 35142
    },
    {
      "epoch": 8.220558998947492,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.6355,
      "step": 35143
    },
    {
      "epoch": 8.220792889720501,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7,
      "step": 35144
    },
    {
      "epoch": 8.22102678049351,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7642,
      "step": 35145
    },
    {
      "epoch": 8.221260671266519,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4778,
      "step": 35146
    },
    {
      "epoch": 8.221494562039528,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 35147
    },
    {
      "epoch": 8.221728452812537,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 35148
    },
    {
      "epoch": 8.221962343585545,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6165,
      "step": 35149
    },
    {
      "epoch": 8.222196234358554,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 35150
    },
    {
      "epoch": 8.222430125131563,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 35151
    },
    {
      "epoch": 8.222664015904572,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4853,
      "step": 35152
    },
    {
      "epoch": 8.22289790667758,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 35153
    },
    {
      "epoch": 8.223131797450591,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7038,
      "step": 35154
    },
    {
      "epoch": 8.2233656882236,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 35155
    },
    {
      "epoch": 8.223599578996609,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5813,
      "step": 35156
    },
    {
      "epoch": 8.223833469769618,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 35157
    },
    {
      "epoch": 8.224067360542627,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6678,
      "step": 35158
    },
    {
      "epoch": 8.224301251315635,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 35159
    },
    {
      "epoch": 8.224535142088644,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 35160
    },
    {
      "epoch": 8.224769032861653,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 35161
    },
    {
      "epoch": 8.225002923634662,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8838,
      "step": 35162
    },
    {
      "epoch": 8.22523681440767,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 35163
    },
    {
      "epoch": 8.225470705180681,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 35164
    },
    {
      "epoch": 8.22570459595369,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5975,
      "step": 35165
    },
    {
      "epoch": 8.2259384867267,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 35166
    },
    {
      "epoch": 8.226172377499708,
      "grad_norm": 8.5,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 35167
    },
    {
      "epoch": 8.226406268272717,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6835,
      "step": 35168
    },
    {
      "epoch": 8.226640159045726,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 35169
    },
    {
      "epoch": 8.226874049818734,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4984,
      "step": 35170
    },
    {
      "epoch": 8.227107940591743,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5807,
      "step": 35171
    },
    {
      "epoch": 8.227341831364752,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 35172
    },
    {
      "epoch": 8.227575722137761,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 35173
    },
    {
      "epoch": 8.22780961291077,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 35174
    },
    {
      "epoch": 8.22804350368378,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 35175
    },
    {
      "epoch": 8.22827739445679,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 35176
    },
    {
      "epoch": 8.228511285229798,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 35177
    },
    {
      "epoch": 8.228745176002807,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 35178
    },
    {
      "epoch": 8.228979066775816,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5373,
      "step": 35179
    },
    {
      "epoch": 8.229212957548825,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 35180
    },
    {
      "epoch": 8.229446848321833,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 35181
    },
    {
      "epoch": 8.229680739094842,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5595,
      "step": 35182
    },
    {
      "epoch": 8.229914629867851,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 35183
    },
    {
      "epoch": 8.23014852064086,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 35184
    },
    {
      "epoch": 8.23038241141387,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 35185
    },
    {
      "epoch": 8.23061630218688,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5601,
      "step": 35186
    },
    {
      "epoch": 8.230850192959888,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 35187
    },
    {
      "epoch": 8.231084083732897,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 35188
    },
    {
      "epoch": 8.231317974505906,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8988,
      "step": 35189
    },
    {
      "epoch": 8.231551865278915,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6361,
      "step": 35190
    },
    {
      "epoch": 8.231785756051924,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 35191
    },
    {
      "epoch": 8.232019646824932,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 35192
    },
    {
      "epoch": 8.232253537597941,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 35193
    },
    {
      "epoch": 8.23248742837095,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4728,
      "step": 35194
    },
    {
      "epoch": 8.232721319143959,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 35195
    },
    {
      "epoch": 8.23295520991697,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 35196
    },
    {
      "epoch": 8.233189100689978,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6786,
      "step": 35197
    },
    {
      "epoch": 8.233422991462987,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 35198
    },
    {
      "epoch": 8.233656882235996,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8001,
      "step": 35199
    },
    {
      "epoch": 8.233890773009005,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 35200
    },
    {
      "epoch": 8.233890773009005,
      "eval_runtime": 4.6021,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 35200
    },
    {
      "epoch": 8.234124663782014,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6889,
      "step": 35201
    },
    {
      "epoch": 8.234358554555023,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 35202
    },
    {
      "epoch": 8.234592445328031,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 35203
    },
    {
      "epoch": 8.23482633610104,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 35204
    },
    {
      "epoch": 8.23506022687405,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 35205
    },
    {
      "epoch": 8.235294117647058,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6243,
      "step": 35206
    },
    {
      "epoch": 8.235528008420069,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 35207
    },
    {
      "epoch": 8.235761899193077,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 35208
    },
    {
      "epoch": 8.235995789966086,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5924,
      "step": 35209
    },
    {
      "epoch": 8.236229680739095,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4762,
      "step": 35210
    },
    {
      "epoch": 8.236463571512104,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7295,
      "step": 35211
    },
    {
      "epoch": 8.236697462285113,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8242,
      "step": 35212
    },
    {
      "epoch": 8.236931353058122,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 35213
    },
    {
      "epoch": 8.23716524383113,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 35214
    },
    {
      "epoch": 8.23739913460414,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 35215
    },
    {
      "epoch": 8.237633025377148,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 35216
    },
    {
      "epoch": 8.237866916150157,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 35217
    },
    {
      "epoch": 8.238100806923168,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 35218
    },
    {
      "epoch": 8.238334697696176,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 35219
    },
    {
      "epoch": 8.238568588469185,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 35220
    },
    {
      "epoch": 8.238802479242194,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5929,
      "step": 35221
    },
    {
      "epoch": 8.239036370015203,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 35222
    },
    {
      "epoch": 8.239270260788212,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 35223
    },
    {
      "epoch": 8.23950415156122,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 35224
    },
    {
      "epoch": 8.23973804233423,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 35225
    },
    {
      "epoch": 8.239971933107238,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 35226
    },
    {
      "epoch": 8.240205823880247,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6795,
      "step": 35227
    },
    {
      "epoch": 8.240439714653258,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 35228
    },
    {
      "epoch": 8.240673605426267,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 35229
    },
    {
      "epoch": 8.240907496199275,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 35230
    },
    {
      "epoch": 8.241141386972284,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5532,
      "step": 35231
    },
    {
      "epoch": 8.241375277745293,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 35232
    },
    {
      "epoch": 8.241609168518302,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.656,
      "step": 35233
    },
    {
      "epoch": 8.24184305929131,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 35234
    },
    {
      "epoch": 8.24207695006432,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 35235
    },
    {
      "epoch": 8.242310840837328,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 35236
    },
    {
      "epoch": 8.242544731610337,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6018,
      "step": 35237
    },
    {
      "epoch": 8.242778622383346,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 35238
    },
    {
      "epoch": 8.243012513156357,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 35239
    },
    {
      "epoch": 8.243246403929366,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 35240
    },
    {
      "epoch": 8.243480294702374,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 35241
    },
    {
      "epoch": 8.243714185475383,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 35242
    },
    {
      "epoch": 8.243948076248392,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 35243
    },
    {
      "epoch": 8.244181967021401,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 35244
    },
    {
      "epoch": 8.24441585779441,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6142,
      "step": 35245
    },
    {
      "epoch": 8.244649748567419,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6607,
      "step": 35246
    },
    {
      "epoch": 8.244883639340427,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 35247
    },
    {
      "epoch": 8.245117530113436,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 35248
    },
    {
      "epoch": 8.245351420886447,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 35249
    },
    {
      "epoch": 8.245585311659456,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 35250
    },
    {
      "epoch": 8.245819202432465,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.559,
      "step": 35251
    },
    {
      "epoch": 8.246053093205473,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4363,
      "step": 35252
    },
    {
      "epoch": 8.246286983978482,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7914,
      "step": 35253
    },
    {
      "epoch": 8.246520874751491,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 35254
    },
    {
      "epoch": 8.2467547655245,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 35255
    },
    {
      "epoch": 8.246988656297509,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 35256
    },
    {
      "epoch": 8.247222547070518,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 35257
    },
    {
      "epoch": 8.247456437843526,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 35258
    },
    {
      "epoch": 8.247690328616535,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5071,
      "step": 35259
    },
    {
      "epoch": 8.247924219389546,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 35260
    },
    {
      "epoch": 8.248158110162555,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 35261
    },
    {
      "epoch": 8.248392000935564,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 35262
    },
    {
      "epoch": 8.248625891708572,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 35263
    },
    {
      "epoch": 8.248859782481581,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 35264
    },
    {
      "epoch": 8.24909367325459,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7597,
      "step": 35265
    },
    {
      "epoch": 8.249327564027599,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.546,
      "step": 35266
    },
    {
      "epoch": 8.249561454800608,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 35267
    },
    {
      "epoch": 8.249795345573617,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7615,
      "step": 35268
    },
    {
      "epoch": 8.250029236346625,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6689,
      "step": 35269
    },
    {
      "epoch": 8.250263127119634,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6111,
      "step": 35270
    },
    {
      "epoch": 8.250497017892645,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6841,
      "step": 35271
    },
    {
      "epoch": 8.250730908665654,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.4538,
      "step": 35272
    },
    {
      "epoch": 8.250964799438663,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 35273
    },
    {
      "epoch": 8.251198690211671,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 35274
    },
    {
      "epoch": 8.25143258098468,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 35275
    },
    {
      "epoch": 8.251666471757689,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.578,
      "step": 35276
    },
    {
      "epoch": 8.251900362530698,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 35277
    },
    {
      "epoch": 8.252134253303707,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6487,
      "step": 35278
    },
    {
      "epoch": 8.252368144076716,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7268,
      "step": 35279
    },
    {
      "epoch": 8.252602034849724,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 35280
    },
    {
      "epoch": 8.252835925622733,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 35281
    },
    {
      "epoch": 8.253069816395744,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 35282
    },
    {
      "epoch": 8.253303707168753,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 35283
    },
    {
      "epoch": 8.253537597941762,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 35284
    },
    {
      "epoch": 8.25377148871477,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 35285
    },
    {
      "epoch": 8.25400537948778,
      "grad_norm": 17.125,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 35286
    },
    {
      "epoch": 8.254239270260788,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 35287
    },
    {
      "epoch": 8.254473161033797,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 35288
    },
    {
      "epoch": 8.254707051806806,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8547,
      "step": 35289
    },
    {
      "epoch": 8.254940942579815,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6007,
      "step": 35290
    },
    {
      "epoch": 8.255174833352823,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 35291
    },
    {
      "epoch": 8.255408724125834,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.603,
      "step": 35292
    },
    {
      "epoch": 8.255642614898843,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 35293
    },
    {
      "epoch": 8.255876505671852,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7284,
      "step": 35294
    },
    {
      "epoch": 8.25611039644486,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 35295
    },
    {
      "epoch": 8.25634428721787,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.645,
      "step": 35296
    },
    {
      "epoch": 8.256578177990878,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7586,
      "step": 35297
    },
    {
      "epoch": 8.256812068763887,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.465,
      "step": 35298
    },
    {
      "epoch": 8.257045959536896,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5193,
      "step": 35299
    },
    {
      "epoch": 8.257279850309905,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 35300
    },
    {
      "epoch": 8.257279850309905,
      "eval_runtime": 4.5976,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 35300
    },
    {
      "epoch": 8.257513741082914,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 35301
    },
    {
      "epoch": 8.257747631855924,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 35302
    },
    {
      "epoch": 8.257981522628933,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5489,
      "step": 35303
    },
    {
      "epoch": 8.258215413401942,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6536,
      "step": 35304
    },
    {
      "epoch": 8.25844930417495,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 35305
    },
    {
      "epoch": 8.25868319494796,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.689,
      "step": 35306
    },
    {
      "epoch": 8.258917085720968,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 35307
    },
    {
      "epoch": 8.259150976493977,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 35308
    },
    {
      "epoch": 8.259384867266986,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.5965,
      "step": 35309
    },
    {
      "epoch": 8.259618758039995,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 35310
    },
    {
      "epoch": 8.259852648813004,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 35311
    },
    {
      "epoch": 8.260086539586013,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7485,
      "step": 35312
    },
    {
      "epoch": 8.260320430359023,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 35313
    },
    {
      "epoch": 8.260554321132032,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7554,
      "step": 35314
    },
    {
      "epoch": 8.26078821190504,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 35315
    },
    {
      "epoch": 8.26102210267805,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7413,
      "step": 35316
    },
    {
      "epoch": 8.261255993451059,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6971,
      "step": 35317
    },
    {
      "epoch": 8.261489884224067,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 35318
    },
    {
      "epoch": 8.261723774997076,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0575,
      "step": 35319
    },
    {
      "epoch": 8.261957665770085,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 35320
    },
    {
      "epoch": 8.262191556543094,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 35321
    },
    {
      "epoch": 8.262425447316103,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6733,
      "step": 35322
    },
    {
      "epoch": 8.262659338089112,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 35323
    },
    {
      "epoch": 8.262893228862122,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 35324
    },
    {
      "epoch": 8.263127119635131,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7377,
      "step": 35325
    },
    {
      "epoch": 8.26336101040814,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6248,
      "step": 35326
    },
    {
      "epoch": 8.263594901181149,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 35327
    },
    {
      "epoch": 8.263828791954158,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 35328
    },
    {
      "epoch": 8.264062682727166,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8487,
      "step": 35329
    },
    {
      "epoch": 8.264296573500175,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 35330
    },
    {
      "epoch": 8.264530464273184,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6758,
      "step": 35331
    },
    {
      "epoch": 8.264764355046193,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.5913,
      "step": 35332
    },
    {
      "epoch": 8.264998245819202,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 35333
    },
    {
      "epoch": 8.26523213659221,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.76,
      "step": 35334
    },
    {
      "epoch": 8.265466027365221,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4334,
      "step": 35335
    },
    {
      "epoch": 8.26569991813823,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 35336
    },
    {
      "epoch": 8.265933808911239,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.3921,
      "step": 35337
    },
    {
      "epoch": 8.266167699684248,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 35338
    },
    {
      "epoch": 8.266401590457257,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 35339
    },
    {
      "epoch": 8.266635481230265,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 35340
    },
    {
      "epoch": 8.266869372003274,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6616,
      "step": 35341
    },
    {
      "epoch": 8.267103262776283,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4935,
      "step": 35342
    },
    {
      "epoch": 8.267337153549292,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4999,
      "step": 35343
    },
    {
      "epoch": 8.2675710443223,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6554,
      "step": 35344
    },
    {
      "epoch": 8.267804935095311,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 35345
    },
    {
      "epoch": 8.26803882586832,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4188,
      "step": 35346
    },
    {
      "epoch": 8.268272716641329,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 35347
    },
    {
      "epoch": 8.268506607414338,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5852,
      "step": 35348
    },
    {
      "epoch": 8.268740498187347,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6213,
      "step": 35349
    },
    {
      "epoch": 8.268974388960356,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 35350
    },
    {
      "epoch": 8.269208279733364,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 35351
    },
    {
      "epoch": 8.269442170506373,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 35352
    },
    {
      "epoch": 8.269676061279382,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 35353
    },
    {
      "epoch": 8.26990995205239,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8347,
      "step": 35354
    },
    {
      "epoch": 8.2701438428254,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 35355
    },
    {
      "epoch": 8.27037773359841,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 35356
    },
    {
      "epoch": 8.27061162437142,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6726,
      "step": 35357
    },
    {
      "epoch": 8.270845515144428,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7085,
      "step": 35358
    },
    {
      "epoch": 8.271079405917437,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.3697,
      "step": 35359
    },
    {
      "epoch": 8.271313296690446,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3623,
      "step": 35360
    },
    {
      "epoch": 8.271547187463455,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 35361
    },
    {
      "epoch": 8.271781078236463,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.6247,
      "step": 35362
    },
    {
      "epoch": 8.272014969009472,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6104,
      "step": 35363
    },
    {
      "epoch": 8.272248859782481,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 35364
    },
    {
      "epoch": 8.27248275055549,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 35365
    },
    {
      "epoch": 8.2727166413285,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5298,
      "step": 35366
    },
    {
      "epoch": 8.27295053210151,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 35367
    },
    {
      "epoch": 8.273184422874518,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 35368
    },
    {
      "epoch": 8.273418313647527,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5226,
      "step": 35369
    },
    {
      "epoch": 8.273652204420536,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 35370
    },
    {
      "epoch": 8.273886095193545,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5387,
      "step": 35371
    },
    {
      "epoch": 8.274119985966554,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.666,
      "step": 35372
    },
    {
      "epoch": 8.274353876739562,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6856,
      "step": 35373
    },
    {
      "epoch": 8.274587767512571,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6203,
      "step": 35374
    },
    {
      "epoch": 8.27482165828558,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 35375
    },
    {
      "epoch": 8.275055549058589,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 35376
    },
    {
      "epoch": 8.2752894398316,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6639,
      "step": 35377
    },
    {
      "epoch": 8.275523330604608,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.661,
      "step": 35378
    },
    {
      "epoch": 8.275757221377617,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 35379
    },
    {
      "epoch": 8.275991112150626,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6852,
      "step": 35380
    },
    {
      "epoch": 8.276225002923635,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 35381
    },
    {
      "epoch": 8.276458893696644,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 35382
    },
    {
      "epoch": 8.276692784469653,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 35383
    },
    {
      "epoch": 8.276926675242661,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.3808,
      "step": 35384
    },
    {
      "epoch": 8.27716056601567,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6494,
      "step": 35385
    },
    {
      "epoch": 8.277394456788679,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 35386
    },
    {
      "epoch": 8.277628347561688,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 35387
    },
    {
      "epoch": 8.277862238334698,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 35388
    },
    {
      "epoch": 8.278096129107707,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 35389
    },
    {
      "epoch": 8.278330019880716,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.3045,
      "step": 35390
    },
    {
      "epoch": 8.278563910653725,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 35391
    },
    {
      "epoch": 8.278797801426734,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6251,
      "step": 35392
    },
    {
      "epoch": 8.279031692199743,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 35393
    },
    {
      "epoch": 8.279265582972752,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 35394
    },
    {
      "epoch": 8.27949947374576,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 35395
    },
    {
      "epoch": 8.27973336451877,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6853,
      "step": 35396
    },
    {
      "epoch": 8.279967255291778,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 35397
    },
    {
      "epoch": 8.280201146064787,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4511,
      "step": 35398
    },
    {
      "epoch": 8.280435036837797,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 35399
    },
    {
      "epoch": 8.280668927610806,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6384,
      "step": 35400
    },
    {
      "epoch": 8.280668927610806,
      "eval_runtime": 4.624,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 35400
    },
    {
      "epoch": 8.280902818383815,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 35401
    },
    {
      "epoch": 8.281136709156824,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 35402
    },
    {
      "epoch": 8.281370599929833,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 35403
    },
    {
      "epoch": 8.281604490702842,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 35404
    },
    {
      "epoch": 8.28183838147585,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 35405
    },
    {
      "epoch": 8.28207227224886,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8322,
      "step": 35406
    },
    {
      "epoch": 8.282306163021868,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 35407
    },
    {
      "epoch": 8.282540053794877,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 35408
    },
    {
      "epoch": 8.282773944567888,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5585,
      "step": 35409
    },
    {
      "epoch": 8.283007835340896,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7039,
      "step": 35410
    },
    {
      "epoch": 8.283241726113905,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6083,
      "step": 35411
    },
    {
      "epoch": 8.283475616886914,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6123,
      "step": 35412
    },
    {
      "epoch": 8.283709507659923,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5281,
      "step": 35413
    },
    {
      "epoch": 8.283943398432932,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 35414
    },
    {
      "epoch": 8.28417728920594,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6133,
      "step": 35415
    },
    {
      "epoch": 8.28441117997895,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6862,
      "step": 35416
    },
    {
      "epoch": 8.284645070751958,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6202,
      "step": 35417
    },
    {
      "epoch": 8.284878961524967,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 35418
    },
    {
      "epoch": 8.285112852297976,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8377,
      "step": 35419
    },
    {
      "epoch": 8.285346743070987,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7365,
      "step": 35420
    },
    {
      "epoch": 8.285580633843995,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 35421
    },
    {
      "epoch": 8.285814524617004,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7186,
      "step": 35422
    },
    {
      "epoch": 8.286048415390013,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 35423
    },
    {
      "epoch": 8.286282306163022,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6935,
      "step": 35424
    },
    {
      "epoch": 8.28651619693603,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3482,
      "step": 35425
    },
    {
      "epoch": 8.28675008770904,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.689,
      "step": 35426
    },
    {
      "epoch": 8.286983978482048,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 35427
    },
    {
      "epoch": 8.287217869255057,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7971,
      "step": 35428
    },
    {
      "epoch": 8.287451760028066,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.748,
      "step": 35429
    },
    {
      "epoch": 8.287685650801077,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7033,
      "step": 35430
    },
    {
      "epoch": 8.287919541574086,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.3832,
      "step": 35431
    },
    {
      "epoch": 8.288153432347094,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7997,
      "step": 35432
    },
    {
      "epoch": 8.288387323120103,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 35433
    },
    {
      "epoch": 8.288621213893112,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 35434
    },
    {
      "epoch": 8.288855104666121,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.5123,
      "step": 35435
    },
    {
      "epoch": 8.28908899543913,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 35436
    },
    {
      "epoch": 8.289322886212139,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 35437
    },
    {
      "epoch": 8.289556776985147,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7116,
      "step": 35438
    },
    {
      "epoch": 8.289790667758156,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 35439
    },
    {
      "epoch": 8.290024558531165,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8468,
      "step": 35440
    },
    {
      "epoch": 8.290258449304176,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.668,
      "step": 35441
    },
    {
      "epoch": 8.290492340077185,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6731,
      "step": 35442
    },
    {
      "epoch": 8.290726230850193,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 35443
    },
    {
      "epoch": 8.290960121623202,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 35444
    },
    {
      "epoch": 8.291194012396211,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 35445
    },
    {
      "epoch": 8.29142790316922,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 35446
    },
    {
      "epoch": 8.291661793942229,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 35447
    },
    {
      "epoch": 8.291895684715238,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 35448
    },
    {
      "epoch": 8.292129575488246,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 35449
    },
    {
      "epoch": 8.292363466261255,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1849,
      "step": 35450
    },
    {
      "epoch": 8.292597357034264,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 35451
    },
    {
      "epoch": 8.292831247807275,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5083,
      "step": 35452
    },
    {
      "epoch": 8.293065138580284,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6188,
      "step": 35453
    },
    {
      "epoch": 8.293299029353292,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 35454
    },
    {
      "epoch": 8.293532920126301,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 35455
    },
    {
      "epoch": 8.29376681089931,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 35456
    },
    {
      "epoch": 8.294000701672319,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 35457
    },
    {
      "epoch": 8.294234592445328,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6266,
      "step": 35458
    },
    {
      "epoch": 8.294468483218337,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 35459
    },
    {
      "epoch": 8.294702373991345,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 35460
    },
    {
      "epoch": 8.294936264764354,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 35461
    },
    {
      "epoch": 8.295170155537363,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 35462
    },
    {
      "epoch": 8.295404046310374,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9136,
      "step": 35463
    },
    {
      "epoch": 8.295637937083383,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6331,
      "step": 35464
    },
    {
      "epoch": 8.295871827856391,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 35465
    },
    {
      "epoch": 8.2961057186294,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6878,
      "step": 35466
    },
    {
      "epoch": 8.296339609402409,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 35467
    },
    {
      "epoch": 8.296573500175418,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4826,
      "step": 35468
    },
    {
      "epoch": 8.296807390948427,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5339,
      "step": 35469
    },
    {
      "epoch": 8.297041281721436,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7242,
      "step": 35470
    },
    {
      "epoch": 8.297275172494444,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 35471
    },
    {
      "epoch": 8.297509063267453,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 35472
    },
    {
      "epoch": 8.297742954040464,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 35473
    },
    {
      "epoch": 8.297976844813473,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8093,
      "step": 35474
    },
    {
      "epoch": 8.298210735586482,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 35475
    },
    {
      "epoch": 8.29844462635949,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 35476
    },
    {
      "epoch": 8.2986785171325,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6548,
      "step": 35477
    },
    {
      "epoch": 8.298912407905508,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 35478
    },
    {
      "epoch": 8.299146298678517,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 35479
    },
    {
      "epoch": 8.299380189451526,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6691,
      "step": 35480
    },
    {
      "epoch": 8.299614080224535,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5228,
      "step": 35481
    },
    {
      "epoch": 8.299847970997543,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8029,
      "step": 35482
    },
    {
      "epoch": 8.300081861770552,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5807,
      "step": 35483
    },
    {
      "epoch": 8.300315752543563,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4251,
      "step": 35484
    },
    {
      "epoch": 8.300549643316572,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 35485
    },
    {
      "epoch": 8.30078353408958,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 35486
    },
    {
      "epoch": 8.30101742486259,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.729,
      "step": 35487
    },
    {
      "epoch": 8.301251315635598,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6499,
      "step": 35488
    },
    {
      "epoch": 8.301485206408607,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.62,
      "step": 35489
    },
    {
      "epoch": 8.301719097181616,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.628,
      "step": 35490
    },
    {
      "epoch": 8.301952987954625,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 35491
    },
    {
      "epoch": 8.302186878727634,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.4823,
      "step": 35492
    },
    {
      "epoch": 8.302420769500642,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 35493
    },
    {
      "epoch": 8.302654660273653,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6001,
      "step": 35494
    },
    {
      "epoch": 8.302888551046662,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 35495
    },
    {
      "epoch": 8.30312244181967,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.525,
      "step": 35496
    },
    {
      "epoch": 8.30335633259268,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 35497
    },
    {
      "epoch": 8.303590223365688,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 35498
    },
    {
      "epoch": 8.303824114138697,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 35499
    },
    {
      "epoch": 8.304058004911706,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 35500
    },
    {
      "epoch": 8.304058004911706,
      "eval_runtime": 4.6367,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 35500
    },
    {
      "epoch": 8.304291895684715,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 35501
    },
    {
      "epoch": 8.304525786457724,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 35502
    },
    {
      "epoch": 8.304759677230733,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 35503
    },
    {
      "epoch": 8.304993568003741,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 35504
    },
    {
      "epoch": 8.305227458776752,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 35505
    },
    {
      "epoch": 8.305461349549761,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 35506
    },
    {
      "epoch": 8.30569524032277,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 35507
    },
    {
      "epoch": 8.305929131095779,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.718,
      "step": 35508
    },
    {
      "epoch": 8.306163021868787,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5401,
      "step": 35509
    },
    {
      "epoch": 8.306396912641796,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 35510
    },
    {
      "epoch": 8.306630803414805,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 35511
    },
    {
      "epoch": 8.306864694187814,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 35512
    },
    {
      "epoch": 8.307098584960823,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 35513
    },
    {
      "epoch": 8.307332475733832,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 35514
    },
    {
      "epoch": 8.30756636650684,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6323,
      "step": 35515
    },
    {
      "epoch": 8.307800257279851,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 35516
    },
    {
      "epoch": 8.30803414805286,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 35517
    },
    {
      "epoch": 8.308268038825869,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 35518
    },
    {
      "epoch": 8.308501929598878,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 35519
    },
    {
      "epoch": 8.308735820371886,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5972,
      "step": 35520
    },
    {
      "epoch": 8.308969711144895,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 35521
    },
    {
      "epoch": 8.309203601917904,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 35522
    },
    {
      "epoch": 8.309437492690913,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 35523
    },
    {
      "epoch": 8.309671383463922,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5531,
      "step": 35524
    },
    {
      "epoch": 8.30990527423693,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8576,
      "step": 35525
    },
    {
      "epoch": 8.310139165009941,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 35526
    },
    {
      "epoch": 8.31037305578295,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 35527
    },
    {
      "epoch": 8.310606946555959,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2843,
      "step": 35528
    },
    {
      "epoch": 8.310840837328968,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7554,
      "step": 35529
    },
    {
      "epoch": 8.311074728101977,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 35530
    },
    {
      "epoch": 8.311308618874985,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6887,
      "step": 35531
    },
    {
      "epoch": 8.311542509647994,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 35532
    },
    {
      "epoch": 8.311776400421003,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5727,
      "step": 35533
    },
    {
      "epoch": 8.312010291194012,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 35534
    },
    {
      "epoch": 8.31224418196702,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.6127,
      "step": 35535
    },
    {
      "epoch": 8.31247807274003,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6156,
      "step": 35536
    },
    {
      "epoch": 8.31271196351304,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6274,
      "step": 35537
    },
    {
      "epoch": 8.312945854286049,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6627,
      "step": 35538
    },
    {
      "epoch": 8.313179745059058,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4551,
      "step": 35539
    },
    {
      "epoch": 8.313413635832067,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5976,
      "step": 35540
    },
    {
      "epoch": 8.313647526605076,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 35541
    },
    {
      "epoch": 8.313881417378084,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 35542
    },
    {
      "epoch": 8.314115308151093,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 35543
    },
    {
      "epoch": 8.314349198924102,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5725,
      "step": 35544
    },
    {
      "epoch": 8.314583089697111,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7203,
      "step": 35545
    },
    {
      "epoch": 8.31481698047012,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4676,
      "step": 35546
    },
    {
      "epoch": 8.31505087124313,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 35547
    },
    {
      "epoch": 8.31528476201614,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7059,
      "step": 35548
    },
    {
      "epoch": 8.315518652789148,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 35549
    },
    {
      "epoch": 8.315752543562157,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 35550
    },
    {
      "epoch": 8.315986434335166,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.678,
      "step": 35551
    },
    {
      "epoch": 8.316220325108175,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5408,
      "step": 35552
    },
    {
      "epoch": 8.316454215881183,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5217,
      "step": 35553
    },
    {
      "epoch": 8.316688106654192,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 35554
    },
    {
      "epoch": 8.316921997427201,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 35555
    },
    {
      "epoch": 8.31715588820021,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6646,
      "step": 35556
    },
    {
      "epoch": 8.317389778973219,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6272,
      "step": 35557
    },
    {
      "epoch": 8.31762366974623,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 35558
    },
    {
      "epoch": 8.317857560519238,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 35559
    },
    {
      "epoch": 8.318091451292247,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 35560
    },
    {
      "epoch": 8.318325342065256,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7697,
      "step": 35561
    },
    {
      "epoch": 8.318559232838265,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7468,
      "step": 35562
    },
    {
      "epoch": 8.318793123611274,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 35563
    },
    {
      "epoch": 8.319027014384282,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6105,
      "step": 35564
    },
    {
      "epoch": 8.319260905157291,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6119,
      "step": 35565
    },
    {
      "epoch": 8.3194947959303,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 35566
    },
    {
      "epoch": 8.319728686703309,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4741,
      "step": 35567
    },
    {
      "epoch": 8.319962577476318,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 35568
    },
    {
      "epoch": 8.320196468249328,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 35569
    },
    {
      "epoch": 8.320430359022337,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 35570
    },
    {
      "epoch": 8.320664249795346,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5639,
      "step": 35571
    },
    {
      "epoch": 8.320898140568355,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 35572
    },
    {
      "epoch": 8.321132031341364,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 35573
    },
    {
      "epoch": 8.321365922114373,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6147,
      "step": 35574
    },
    {
      "epoch": 8.321599812887381,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 35575
    },
    {
      "epoch": 8.32183370366039,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7725,
      "step": 35576
    },
    {
      "epoch": 8.322067594433399,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 35577
    },
    {
      "epoch": 8.322301485206408,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 35578
    },
    {
      "epoch": 8.322535375979417,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 35579
    },
    {
      "epoch": 8.322769266752427,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7266,
      "step": 35580
    },
    {
      "epoch": 8.323003157525436,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 35581
    },
    {
      "epoch": 8.323237048298445,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 35582
    },
    {
      "epoch": 8.323470939071454,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.72,
      "step": 35583
    },
    {
      "epoch": 8.323704829844463,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7055,
      "step": 35584
    },
    {
      "epoch": 8.323938720617472,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 35585
    },
    {
      "epoch": 8.32417261139048,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6877,
      "step": 35586
    },
    {
      "epoch": 8.32440650216349,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 35587
    },
    {
      "epoch": 8.324640392936498,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 35588
    },
    {
      "epoch": 8.324874283709507,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 35589
    },
    {
      "epoch": 8.325108174482517,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 35590
    },
    {
      "epoch": 8.325342065255526,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8528,
      "step": 35591
    },
    {
      "epoch": 8.325575956028535,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7215,
      "step": 35592
    },
    {
      "epoch": 8.325809846801544,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 35593
    },
    {
      "epoch": 8.326043737574553,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 35594
    },
    {
      "epoch": 8.326277628347562,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 35595
    },
    {
      "epoch": 8.32651151912057,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 35596
    },
    {
      "epoch": 8.32674540989358,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 35597
    },
    {
      "epoch": 8.326979300666588,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 35598
    },
    {
      "epoch": 8.327213191439597,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6391,
      "step": 35599
    },
    {
      "epoch": 8.327447082212606,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 35600
    },
    {
      "epoch": 8.327447082212606,
      "eval_runtime": 4.6497,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 35600
    },
    {
      "epoch": 8.327680972985616,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 35601
    },
    {
      "epoch": 8.327914863758625,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 35602
    },
    {
      "epoch": 8.328148754531634,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7622,
      "step": 35603
    },
    {
      "epoch": 8.328382645304643,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 35604
    },
    {
      "epoch": 8.328616536077652,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4732,
      "step": 35605
    },
    {
      "epoch": 8.32885042685066,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 35606
    },
    {
      "epoch": 8.32908431762367,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5605,
      "step": 35607
    },
    {
      "epoch": 8.329318208396678,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 35608
    },
    {
      "epoch": 8.329552099169687,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4054,
      "step": 35609
    },
    {
      "epoch": 8.329785989942696,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 35610
    },
    {
      "epoch": 8.330019880715707,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7467,
      "step": 35611
    },
    {
      "epoch": 8.330253771488715,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5285,
      "step": 35612
    },
    {
      "epoch": 8.330487662261724,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7085,
      "step": 35613
    },
    {
      "epoch": 8.330721553034733,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5332,
      "step": 35614
    },
    {
      "epoch": 8.330955443807742,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.788,
      "step": 35615
    },
    {
      "epoch": 8.33118933458075,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.5086,
      "step": 35616
    },
    {
      "epoch": 8.33142322535376,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.5473,
      "step": 35617
    },
    {
      "epoch": 8.331657116126769,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8312,
      "step": 35618
    },
    {
      "epoch": 8.331891006899777,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.747,
      "step": 35619
    },
    {
      "epoch": 8.332124897672786,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 35620
    },
    {
      "epoch": 8.332358788445795,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 35621
    },
    {
      "epoch": 8.332592679218806,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5081,
      "step": 35622
    },
    {
      "epoch": 8.332826569991814,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7107,
      "step": 35623
    },
    {
      "epoch": 8.333060460764823,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 35624
    },
    {
      "epoch": 8.333294351537832,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6164,
      "step": 35625
    },
    {
      "epoch": 8.333528242310841,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6866,
      "step": 35626
    },
    {
      "epoch": 8.33376213308385,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 35627
    },
    {
      "epoch": 8.333996023856859,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 35628
    },
    {
      "epoch": 8.334229914629868,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5551,
      "step": 35629
    },
    {
      "epoch": 8.334463805402876,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.489,
      "step": 35630
    },
    {
      "epoch": 8.334697696175885,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 35631
    },
    {
      "epoch": 8.334931586948894,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.5875,
      "step": 35632
    },
    {
      "epoch": 8.335165477721905,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 35633
    },
    {
      "epoch": 8.335399368494913,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6698,
      "step": 35634
    },
    {
      "epoch": 8.335633259267922,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4338,
      "step": 35635
    },
    {
      "epoch": 8.335867150040931,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6231,
      "step": 35636
    },
    {
      "epoch": 8.33610104081394,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 35637
    },
    {
      "epoch": 8.336334931586949,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7283,
      "step": 35638
    },
    {
      "epoch": 8.336568822359958,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6571,
      "step": 35639
    },
    {
      "epoch": 8.336802713132967,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5411,
      "step": 35640
    },
    {
      "epoch": 8.337036603905975,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5234,
      "step": 35641
    },
    {
      "epoch": 8.337270494678984,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 35642
    },
    {
      "epoch": 8.337504385451993,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 35643
    },
    {
      "epoch": 8.337738276225004,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 35644
    },
    {
      "epoch": 8.337972166998012,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7547,
      "step": 35645
    },
    {
      "epoch": 8.338206057771021,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6273,
      "step": 35646
    },
    {
      "epoch": 8.33843994854403,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 35647
    },
    {
      "epoch": 8.338673839317039,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7205,
      "step": 35648
    },
    {
      "epoch": 8.338907730090048,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5724,
      "step": 35649
    },
    {
      "epoch": 8.339141620863057,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 35650
    },
    {
      "epoch": 8.339375511636065,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6348,
      "step": 35651
    },
    {
      "epoch": 8.339609402409074,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4939,
      "step": 35652
    },
    {
      "epoch": 8.339843293182083,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 35653
    },
    {
      "epoch": 8.340077183955094,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 35654
    },
    {
      "epoch": 8.340311074728103,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 35655
    },
    {
      "epoch": 8.340544965501111,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 35656
    },
    {
      "epoch": 8.34077885627412,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 35657
    },
    {
      "epoch": 8.34101274704713,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 35658
    },
    {
      "epoch": 8.341246637820138,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7552,
      "step": 35659
    },
    {
      "epoch": 8.341480528593147,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 35660
    },
    {
      "epoch": 8.341714419366156,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6118,
      "step": 35661
    },
    {
      "epoch": 8.341948310139164,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 35662
    },
    {
      "epoch": 8.342182200912173,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 35663
    },
    {
      "epoch": 8.342416091685182,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 35664
    },
    {
      "epoch": 8.342649982458193,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 35665
    },
    {
      "epoch": 8.342883873231202,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 35666
    },
    {
      "epoch": 8.34311776400421,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 35667
    },
    {
      "epoch": 8.34335165477722,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 35668
    },
    {
      "epoch": 8.343585545550228,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 35669
    },
    {
      "epoch": 8.343819436323237,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5065,
      "step": 35670
    },
    {
      "epoch": 8.344053327096246,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5997,
      "step": 35671
    },
    {
      "epoch": 8.344287217869255,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 35672
    },
    {
      "epoch": 8.344521108642263,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5813,
      "step": 35673
    },
    {
      "epoch": 8.344754999415272,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 35674
    },
    {
      "epoch": 8.344988890188283,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5195,
      "step": 35675
    },
    {
      "epoch": 8.345222780961292,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.525,
      "step": 35676
    },
    {
      "epoch": 8.3454566717343,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.523,
      "step": 35677
    },
    {
      "epoch": 8.34569056250731,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 35678
    },
    {
      "epoch": 8.345924453280318,
      "grad_norm": 2.484375,
      "learning_rate": 3e-05,
      "loss": 1.5342,
      "step": 35679
    },
    {
      "epoch": 8.346158344053327,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4764,
      "step": 35680
    },
    {
      "epoch": 8.346392234826336,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 35681
    },
    {
      "epoch": 8.346626125599345,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 35682
    },
    {
      "epoch": 8.346860016372354,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 35683
    },
    {
      "epoch": 8.347093907145362,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 35684
    },
    {
      "epoch": 8.347327797918371,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 35685
    },
    {
      "epoch": 8.347561688691382,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5071,
      "step": 35686
    },
    {
      "epoch": 8.34779557946439,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 35687
    },
    {
      "epoch": 8.3480294702374,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 35688
    },
    {
      "epoch": 8.348263361010408,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6065,
      "step": 35689
    },
    {
      "epoch": 8.348497251783417,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5335,
      "step": 35690
    },
    {
      "epoch": 8.348731142556426,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5925,
      "step": 35691
    },
    {
      "epoch": 8.348965033329435,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 35692
    },
    {
      "epoch": 8.349198924102444,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 35693
    },
    {
      "epoch": 8.349432814875453,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 35694
    },
    {
      "epoch": 8.349666705648461,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5898,
      "step": 35695
    },
    {
      "epoch": 8.34990059642147,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5559,
      "step": 35696
    },
    {
      "epoch": 8.350134487194481,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 35697
    },
    {
      "epoch": 8.35036837796749,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6701,
      "step": 35698
    },
    {
      "epoch": 8.350602268740499,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.626,
      "step": 35699
    },
    {
      "epoch": 8.350836159513507,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7023,
      "step": 35700
    },
    {
      "epoch": 8.350836159513507,
      "eval_runtime": 4.6229,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 35700
    },
    {
      "epoch": 8.351070050286516,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6275,
      "step": 35701
    },
    {
      "epoch": 8.351303941059525,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 35702
    },
    {
      "epoch": 8.351537831832534,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 35703
    },
    {
      "epoch": 8.351771722605543,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 35704
    },
    {
      "epoch": 8.352005613378552,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6406,
      "step": 35705
    },
    {
      "epoch": 8.35223950415156,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 35706
    },
    {
      "epoch": 8.352473394924571,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5949,
      "step": 35707
    },
    {
      "epoch": 8.35270728569758,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 35708
    },
    {
      "epoch": 8.352941176470589,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 35709
    },
    {
      "epoch": 8.353175067243598,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8247,
      "step": 35710
    },
    {
      "epoch": 8.353408958016606,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 35711
    },
    {
      "epoch": 8.353642848789615,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 35712
    },
    {
      "epoch": 8.353876739562624,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 35713
    },
    {
      "epoch": 8.354110630335633,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6045,
      "step": 35714
    },
    {
      "epoch": 8.354344521108642,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 35715
    },
    {
      "epoch": 8.35457841188165,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7855,
      "step": 35716
    },
    {
      "epoch": 8.35481230265466,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 35717
    },
    {
      "epoch": 8.35504619342767,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 35718
    },
    {
      "epoch": 8.355280084200679,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.472,
      "step": 35719
    },
    {
      "epoch": 8.355513974973688,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 35720
    },
    {
      "epoch": 8.355747865746697,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7826,
      "step": 35721
    },
    {
      "epoch": 8.355981756519705,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.4212,
      "step": 35722
    },
    {
      "epoch": 8.356215647292714,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6544,
      "step": 35723
    },
    {
      "epoch": 8.356449538065723,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 35724
    },
    {
      "epoch": 8.356683428838732,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6597,
      "step": 35725
    },
    {
      "epoch": 8.35691731961174,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 35726
    },
    {
      "epoch": 8.35715121038475,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 35727
    },
    {
      "epoch": 8.35738510115776,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 35728
    },
    {
      "epoch": 8.357618991930769,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 35729
    },
    {
      "epoch": 8.357852882703778,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6631,
      "step": 35730
    },
    {
      "epoch": 8.358086773476787,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7068,
      "step": 35731
    },
    {
      "epoch": 8.358320664249796,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 35732
    },
    {
      "epoch": 8.358554555022804,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 35733
    },
    {
      "epoch": 8.358788445795813,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9461,
      "step": 35734
    },
    {
      "epoch": 8.359022336568822,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 35735
    },
    {
      "epoch": 8.359256227341831,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 35736
    },
    {
      "epoch": 8.35949011811484,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6674,
      "step": 35737
    },
    {
      "epoch": 8.359724008887849,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6214,
      "step": 35738
    },
    {
      "epoch": 8.35995789966086,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5657,
      "step": 35739
    },
    {
      "epoch": 8.360191790433868,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 35740
    },
    {
      "epoch": 8.360425681206877,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 35741
    },
    {
      "epoch": 8.360659571979886,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.3826,
      "step": 35742
    },
    {
      "epoch": 8.360893462752895,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 35743
    },
    {
      "epoch": 8.361127353525903,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.701,
      "step": 35744
    },
    {
      "epoch": 8.361361244298912,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6868,
      "step": 35745
    },
    {
      "epoch": 8.361595135071921,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 35746
    },
    {
      "epoch": 8.36182902584493,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6359,
      "step": 35747
    },
    {
      "epoch": 8.362062916617939,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 35748
    },
    {
      "epoch": 8.362296807390948,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 35749
    },
    {
      "epoch": 8.362530698163958,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 35750
    },
    {
      "epoch": 8.362764588936967,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 35751
    },
    {
      "epoch": 8.362998479709976,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 35752
    },
    {
      "epoch": 8.363232370482985,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 35753
    },
    {
      "epoch": 8.363466261255994,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 35754
    },
    {
      "epoch": 8.363700152029002,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9204,
      "step": 35755
    },
    {
      "epoch": 8.363934042802011,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 35756
    },
    {
      "epoch": 8.36416793357502,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6327,
      "step": 35757
    },
    {
      "epoch": 8.364401824348029,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8657,
      "step": 35758
    },
    {
      "epoch": 8.364635715121038,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7769,
      "step": 35759
    },
    {
      "epoch": 8.364869605894047,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6291,
      "step": 35760
    },
    {
      "epoch": 8.365103496667057,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 35761
    },
    {
      "epoch": 8.365337387440066,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.2784,
      "step": 35762
    },
    {
      "epoch": 8.365571278213075,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 35763
    },
    {
      "epoch": 8.365805168986084,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 35764
    },
    {
      "epoch": 8.366039059759093,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 35765
    },
    {
      "epoch": 8.366272950532101,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 35766
    },
    {
      "epoch": 8.36650684130511,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6038,
      "step": 35767
    },
    {
      "epoch": 8.366740732078119,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7115,
      "step": 35768
    },
    {
      "epoch": 8.366974622851128,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 35769
    },
    {
      "epoch": 8.367208513624137,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 35770
    },
    {
      "epoch": 8.367442404397147,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 35771
    },
    {
      "epoch": 8.367676295170156,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5788,
      "step": 35772
    },
    {
      "epoch": 8.367910185943165,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6508,
      "step": 35773
    },
    {
      "epoch": 8.368144076716174,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 35774
    },
    {
      "epoch": 8.368377967489183,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 35775
    },
    {
      "epoch": 8.368611858262192,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 35776
    },
    {
      "epoch": 8.3688457490352,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.58,
      "step": 35777
    },
    {
      "epoch": 8.36907963980821,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7439,
      "step": 35778
    },
    {
      "epoch": 8.369313530581218,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 35779
    },
    {
      "epoch": 8.369547421354227,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 35780
    },
    {
      "epoch": 8.369781312127236,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6403,
      "step": 35781
    },
    {
      "epoch": 8.370015202900246,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 35782
    },
    {
      "epoch": 8.370249093673255,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 35783
    },
    {
      "epoch": 8.370482984446264,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5728,
      "step": 35784
    },
    {
      "epoch": 8.370716875219273,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8487,
      "step": 35785
    },
    {
      "epoch": 8.370950765992282,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9569,
      "step": 35786
    },
    {
      "epoch": 8.37118465676529,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.546,
      "step": 35787
    },
    {
      "epoch": 8.3714185475383,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 35788
    },
    {
      "epoch": 8.371652438311308,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7355,
      "step": 35789
    },
    {
      "epoch": 8.371886329084317,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 35790
    },
    {
      "epoch": 8.372120219857326,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 35791
    },
    {
      "epoch": 8.372354110630337,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5989,
      "step": 35792
    },
    {
      "epoch": 8.372588001403345,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 35793
    },
    {
      "epoch": 8.372821892176354,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6987,
      "step": 35794
    },
    {
      "epoch": 8.373055782949363,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7161,
      "step": 35795
    },
    {
      "epoch": 8.373289673722372,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.691,
      "step": 35796
    },
    {
      "epoch": 8.37352356449538,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 35797
    },
    {
      "epoch": 8.37375745526839,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 35798
    },
    {
      "epoch": 8.373991346041398,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 35799
    },
    {
      "epoch": 8.374225236814407,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6036,
      "step": 35800
    },
    {
      "epoch": 8.374225236814407,
      "eval_runtime": 4.685,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 35800
    },
    {
      "epoch": 8.374459127587416,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 35801
    },
    {
      "epoch": 8.374693018360425,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5643,
      "step": 35802
    },
    {
      "epoch": 8.374926909133436,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 35803
    },
    {
      "epoch": 8.375160799906444,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 35804
    },
    {
      "epoch": 8.375394690679453,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 35805
    },
    {
      "epoch": 8.375628581452462,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 35806
    },
    {
      "epoch": 8.37586247222547,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5102,
      "step": 35807
    },
    {
      "epoch": 8.37609636299848,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6077,
      "step": 35808
    },
    {
      "epoch": 8.376330253771489,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7518,
      "step": 35809
    },
    {
      "epoch": 8.376564144544497,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 35810
    },
    {
      "epoch": 8.376798035317506,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6817,
      "step": 35811
    },
    {
      "epoch": 8.377031926090515,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6233,
      "step": 35812
    },
    {
      "epoch": 8.377265816863524,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 35813
    },
    {
      "epoch": 8.377499707636535,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 35814
    },
    {
      "epoch": 8.377733598409543,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6952,
      "step": 35815
    },
    {
      "epoch": 8.377967489182552,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.552,
      "step": 35816
    },
    {
      "epoch": 8.378201379955561,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 35817
    },
    {
      "epoch": 8.37843527072857,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 35818
    },
    {
      "epoch": 8.378669161501579,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 35819
    },
    {
      "epoch": 8.378903052274588,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5363,
      "step": 35820
    },
    {
      "epoch": 8.379136943047596,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 35821
    },
    {
      "epoch": 8.379370833820605,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.832,
      "step": 35822
    },
    {
      "epoch": 8.379604724593614,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 35823
    },
    {
      "epoch": 8.379838615366623,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 35824
    },
    {
      "epoch": 8.380072506139634,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 35825
    },
    {
      "epoch": 8.380306396912642,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 35826
    },
    {
      "epoch": 8.380540287685651,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 35827
    },
    {
      "epoch": 8.38077417845866,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6112,
      "step": 35828
    },
    {
      "epoch": 8.381008069231669,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6125,
      "step": 35829
    },
    {
      "epoch": 8.381241960004678,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7767,
      "step": 35830
    },
    {
      "epoch": 8.381475850777687,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 35831
    },
    {
      "epoch": 8.381709741550695,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 35832
    },
    {
      "epoch": 8.381943632323704,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.528,
      "step": 35833
    },
    {
      "epoch": 8.382177523096713,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6332,
      "step": 35834
    },
    {
      "epoch": 8.382411413869724,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 35835
    },
    {
      "epoch": 8.382645304642732,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.783,
      "step": 35836
    },
    {
      "epoch": 8.382879195415741,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 35837
    },
    {
      "epoch": 8.38311308618875,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 35838
    },
    {
      "epoch": 8.383346976961759,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 35839
    },
    {
      "epoch": 8.383580867734768,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 35840
    },
    {
      "epoch": 8.383814758507777,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.3957,
      "step": 35841
    },
    {
      "epoch": 8.384048649280786,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 35842
    },
    {
      "epoch": 8.384282540053794,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 35843
    },
    {
      "epoch": 8.384516430826803,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 35844
    },
    {
      "epoch": 8.384750321599812,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 35845
    },
    {
      "epoch": 8.384984212372823,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8337,
      "step": 35846
    },
    {
      "epoch": 8.385218103145831,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.768,
      "step": 35847
    },
    {
      "epoch": 8.38545199391884,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 35848
    },
    {
      "epoch": 8.38568588469185,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 35849
    },
    {
      "epoch": 8.385919775464858,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 35850
    },
    {
      "epoch": 8.386153666237867,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 35851
    },
    {
      "epoch": 8.386387557010876,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 35852
    },
    {
      "epoch": 8.386621447783885,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6226,
      "step": 35853
    },
    {
      "epoch": 8.386855338556893,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 35854
    },
    {
      "epoch": 8.387089229329902,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6538,
      "step": 35855
    },
    {
      "epoch": 8.387323120102913,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 35856
    },
    {
      "epoch": 8.387557010875922,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.038,
      "step": 35857
    },
    {
      "epoch": 8.38779090164893,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 35858
    },
    {
      "epoch": 8.38802479242194,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8122,
      "step": 35859
    },
    {
      "epoch": 8.388258683194948,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6105,
      "step": 35860
    },
    {
      "epoch": 8.388492573967957,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 35861
    },
    {
      "epoch": 8.388726464740966,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 35862
    },
    {
      "epoch": 8.388960355513975,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 35863
    },
    {
      "epoch": 8.389194246286984,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5922,
      "step": 35864
    },
    {
      "epoch": 8.389428137059992,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6038,
      "step": 35865
    },
    {
      "epoch": 8.389662027833001,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 35866
    },
    {
      "epoch": 8.389895918606012,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 35867
    },
    {
      "epoch": 8.39012980937902,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5773,
      "step": 35868
    },
    {
      "epoch": 8.39036370015203,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5903,
      "step": 35869
    },
    {
      "epoch": 8.390597590925038,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 35870
    },
    {
      "epoch": 8.390831481698047,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5821,
      "step": 35871
    },
    {
      "epoch": 8.391065372471056,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 35872
    },
    {
      "epoch": 8.391299263244065,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 35873
    },
    {
      "epoch": 8.391533154017074,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 35874
    },
    {
      "epoch": 8.391767044790083,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 35875
    },
    {
      "epoch": 8.392000935563091,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5107,
      "step": 35876
    },
    {
      "epoch": 8.3922348263361,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 35877
    },
    {
      "epoch": 8.39246871710911,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6856,
      "step": 35878
    },
    {
      "epoch": 8.39270260788212,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 35879
    },
    {
      "epoch": 8.392936498655128,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 2.3146,
      "step": 35880
    },
    {
      "epoch": 8.393170389428137,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 35881
    },
    {
      "epoch": 8.393404280201146,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6975,
      "step": 35882
    },
    {
      "epoch": 8.393638170974155,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 35883
    },
    {
      "epoch": 8.393872061747164,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 35884
    },
    {
      "epoch": 8.394105952520173,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.716,
      "step": 35885
    },
    {
      "epoch": 8.394339843293182,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.78,
      "step": 35886
    },
    {
      "epoch": 8.39457373406619,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 35887
    },
    {
      "epoch": 8.3948076248392,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6368,
      "step": 35888
    },
    {
      "epoch": 8.39504151561221,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.761,
      "step": 35889
    },
    {
      "epoch": 8.395275406385219,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6477,
      "step": 35890
    },
    {
      "epoch": 8.395509297158227,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 35891
    },
    {
      "epoch": 8.395743187931236,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 35892
    },
    {
      "epoch": 8.395977078704245,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6784,
      "step": 35893
    },
    {
      "epoch": 8.396210969477254,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 35894
    },
    {
      "epoch": 8.396444860250263,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 35895
    },
    {
      "epoch": 8.396678751023272,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6772,
      "step": 35896
    },
    {
      "epoch": 8.39691264179628,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7318,
      "step": 35897
    },
    {
      "epoch": 8.39714653256929,
      "grad_norm": 8.125,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 35898
    },
    {
      "epoch": 8.3973804233423,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 35899
    },
    {
      "epoch": 8.397614314115309,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 35900
    },
    {
      "epoch": 8.397614314115309,
      "eval_runtime": 4.6641,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 35900
    },
    {
      "epoch": 8.397848204888318,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 35901
    },
    {
      "epoch": 8.398082095661326,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 35902
    },
    {
      "epoch": 8.398315986434335,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7219,
      "step": 35903
    },
    {
      "epoch": 8.398549877207344,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 35904
    },
    {
      "epoch": 8.398783767980353,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 35905
    },
    {
      "epoch": 8.399017658753362,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 35906
    },
    {
      "epoch": 8.39925154952637,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 35907
    },
    {
      "epoch": 8.39948544029938,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5305,
      "step": 35908
    },
    {
      "epoch": 8.39971933107239,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 35909
    },
    {
      "epoch": 8.399953221845399,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 35910
    },
    {
      "epoch": 8.400187112618408,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7655,
      "step": 35911
    },
    {
      "epoch": 8.400421003391417,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 35912
    },
    {
      "epoch": 8.400654894164425,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5573,
      "step": 35913
    },
    {
      "epoch": 8.400888784937434,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6394,
      "step": 35914
    },
    {
      "epoch": 8.401122675710443,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 35915
    },
    {
      "epoch": 8.401356566483452,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 35916
    },
    {
      "epoch": 8.40159045725646,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5164,
      "step": 35917
    },
    {
      "epoch": 8.40182434802947,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 35918
    },
    {
      "epoch": 8.402058238802478,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 35919
    },
    {
      "epoch": 8.402292129575489,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6872,
      "step": 35920
    },
    {
      "epoch": 8.402526020348498,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 35921
    },
    {
      "epoch": 8.402759911121507,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 35922
    },
    {
      "epoch": 8.402993801894516,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 35923
    },
    {
      "epoch": 8.403227692667524,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 35924
    },
    {
      "epoch": 8.403461583440533,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 35925
    },
    {
      "epoch": 8.403695474213542,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 35926
    },
    {
      "epoch": 8.403929364986551,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 35927
    },
    {
      "epoch": 8.40416325575956,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6237,
      "step": 35928
    },
    {
      "epoch": 8.404397146532569,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 35929
    },
    {
      "epoch": 8.404631037305577,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6423,
      "step": 35930
    },
    {
      "epoch": 8.404864928078588,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 35931
    },
    {
      "epoch": 8.405098818851597,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 35932
    },
    {
      "epoch": 8.405332709624606,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 35933
    },
    {
      "epoch": 8.405566600397615,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 35934
    },
    {
      "epoch": 8.405800491170623,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6388,
      "step": 35935
    },
    {
      "epoch": 8.406034381943632,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5683,
      "step": 35936
    },
    {
      "epoch": 8.406268272716641,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5088,
      "step": 35937
    },
    {
      "epoch": 8.40650216348965,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 35938
    },
    {
      "epoch": 8.406736054262659,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 35939
    },
    {
      "epoch": 8.406969945035668,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 35940
    },
    {
      "epoch": 8.407203835808676,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6205,
      "step": 35941
    },
    {
      "epoch": 8.407437726581687,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6416,
      "step": 35942
    },
    {
      "epoch": 8.407671617354696,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 35943
    },
    {
      "epoch": 8.407905508127705,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 35944
    },
    {
      "epoch": 8.408139398900714,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5393,
      "step": 35945
    },
    {
      "epoch": 8.408373289673722,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5993,
      "step": 35946
    },
    {
      "epoch": 8.408607180446731,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 35947
    },
    {
      "epoch": 8.40884107121974,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 35948
    },
    {
      "epoch": 8.409074961992749,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.5815,
      "step": 35949
    },
    {
      "epoch": 8.409308852765758,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.5703,
      "step": 35950
    },
    {
      "epoch": 8.409542743538767,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 35951
    },
    {
      "epoch": 8.409776634311777,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4556,
      "step": 35952
    },
    {
      "epoch": 8.410010525084786,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5402,
      "step": 35953
    },
    {
      "epoch": 8.410244415857795,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.3905,
      "step": 35954
    },
    {
      "epoch": 8.410478306630804,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8242,
      "step": 35955
    },
    {
      "epoch": 8.410712197403813,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 35956
    },
    {
      "epoch": 8.410946088176821,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6052,
      "step": 35957
    },
    {
      "epoch": 8.41117997894983,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 35958
    },
    {
      "epoch": 8.411413869722839,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 35959
    },
    {
      "epoch": 8.411647760495848,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 35960
    },
    {
      "epoch": 8.411881651268857,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 35961
    },
    {
      "epoch": 8.412115542041866,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 35962
    },
    {
      "epoch": 8.412349432814876,
      "grad_norm": 2.453125,
      "learning_rate": 3e-05,
      "loss": 1.4306,
      "step": 35963
    },
    {
      "epoch": 8.412583323587885,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7597,
      "step": 35964
    },
    {
      "epoch": 8.412817214360894,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 35965
    },
    {
      "epoch": 8.413051105133903,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 35966
    },
    {
      "epoch": 8.413284995906912,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 35967
    },
    {
      "epoch": 8.41351888667992,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5004,
      "step": 35968
    },
    {
      "epoch": 8.41375277745293,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 35969
    },
    {
      "epoch": 8.413986668225938,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 35970
    },
    {
      "epoch": 8.414220558998947,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4916,
      "step": 35971
    },
    {
      "epoch": 8.414454449771956,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 35972
    },
    {
      "epoch": 8.414688340544966,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7208,
      "step": 35973
    },
    {
      "epoch": 8.414922231317975,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 35974
    },
    {
      "epoch": 8.415156122090984,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 35975
    },
    {
      "epoch": 8.415390012863993,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5664,
      "step": 35976
    },
    {
      "epoch": 8.415623903637002,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 35977
    },
    {
      "epoch": 8.41585779441001,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 35978
    },
    {
      "epoch": 8.41609168518302,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6449,
      "step": 35979
    },
    {
      "epoch": 8.416325575956028,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5918,
      "step": 35980
    },
    {
      "epoch": 8.416559466729037,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 35981
    },
    {
      "epoch": 8.416793357502046,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4029,
      "step": 35982
    },
    {
      "epoch": 8.417027248275055,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 35983
    },
    {
      "epoch": 8.417261139048065,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 35984
    },
    {
      "epoch": 8.417495029821074,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6883,
      "step": 35985
    },
    {
      "epoch": 8.417728920594083,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 35986
    },
    {
      "epoch": 8.417962811367092,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6031,
      "step": 35987
    },
    {
      "epoch": 8.4181967021401,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6029,
      "step": 35988
    },
    {
      "epoch": 8.41843059291311,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8532,
      "step": 35989
    },
    {
      "epoch": 8.418664483686118,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 35990
    },
    {
      "epoch": 8.418898374459127,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 35991
    },
    {
      "epoch": 8.419132265232136,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 35992
    },
    {
      "epoch": 8.419366156005145,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 35993
    },
    {
      "epoch": 8.419600046778154,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 35994
    },
    {
      "epoch": 8.419833937551164,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 35995
    },
    {
      "epoch": 8.420067828324173,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8838,
      "step": 35996
    },
    {
      "epoch": 8.420301719097182,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 35997
    },
    {
      "epoch": 8.420535609870191,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 35998
    },
    {
      "epoch": 8.4207695006432,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 35999
    },
    {
      "epoch": 8.421003391416209,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 36000
    },
    {
      "epoch": 8.421003391416209,
      "eval_runtime": 4.5985,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 36000
    },
    {
      "epoch": 8.421237282189217,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6318,
      "step": 36001
    },
    {
      "epoch": 8.421471172962226,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5634,
      "step": 36002
    },
    {
      "epoch": 8.421705063735235,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 36003
    },
    {
      "epoch": 8.421938954508244,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4871,
      "step": 36004
    },
    {
      "epoch": 8.422172845281253,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8693,
      "step": 36005
    },
    {
      "epoch": 8.422406736054263,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 36006
    },
    {
      "epoch": 8.422640626827272,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 36007
    },
    {
      "epoch": 8.422874517600281,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5941,
      "step": 36008
    },
    {
      "epoch": 8.42310840837329,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6475,
      "step": 36009
    },
    {
      "epoch": 8.423342299146299,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.63,
      "step": 36010
    },
    {
      "epoch": 8.423576189919308,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 36011
    },
    {
      "epoch": 8.423810080692316,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 36012
    },
    {
      "epoch": 8.424043971465325,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.819,
      "step": 36013
    },
    {
      "epoch": 8.424277862238334,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 36014
    },
    {
      "epoch": 8.424511753011343,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 36015
    },
    {
      "epoch": 8.424745643784354,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7259,
      "step": 36016
    },
    {
      "epoch": 8.424979534557362,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.741,
      "step": 36017
    },
    {
      "epoch": 8.425213425330371,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 36018
    },
    {
      "epoch": 8.42544731610338,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.4949,
      "step": 36019
    },
    {
      "epoch": 8.425681206876389,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 36020
    },
    {
      "epoch": 8.425915097649398,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 36021
    },
    {
      "epoch": 8.426148988422407,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8308,
      "step": 36022
    },
    {
      "epoch": 8.426382879195415,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.3204,
      "step": 36023
    },
    {
      "epoch": 8.426616769968424,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 36024
    },
    {
      "epoch": 8.426850660741433,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6858,
      "step": 36025
    },
    {
      "epoch": 8.427084551514442,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6067,
      "step": 36026
    },
    {
      "epoch": 8.427318442287453,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 36027
    },
    {
      "epoch": 8.427552333060461,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6558,
      "step": 36028
    },
    {
      "epoch": 8.42778622383347,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 36029
    },
    {
      "epoch": 8.428020114606479,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.545,
      "step": 36030
    },
    {
      "epoch": 8.428254005379488,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.363,
      "step": 36031
    },
    {
      "epoch": 8.428487896152497,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.573,
      "step": 36032
    },
    {
      "epoch": 8.428721786925506,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7153,
      "step": 36033
    },
    {
      "epoch": 8.428955677698514,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.726,
      "step": 36034
    },
    {
      "epoch": 8.429189568471523,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.686,
      "step": 36035
    },
    {
      "epoch": 8.429423459244532,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4013,
      "step": 36036
    },
    {
      "epoch": 8.429657350017543,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.885,
      "step": 36037
    },
    {
      "epoch": 8.429891240790552,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7947,
      "step": 36038
    },
    {
      "epoch": 8.43012513156356,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5664,
      "step": 36039
    },
    {
      "epoch": 8.43035902233657,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 36040
    },
    {
      "epoch": 8.430592913109578,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 36041
    },
    {
      "epoch": 8.430826803882587,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 36042
    },
    {
      "epoch": 8.431060694655596,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 36043
    },
    {
      "epoch": 8.431294585428605,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 36044
    },
    {
      "epoch": 8.431528476201613,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 36045
    },
    {
      "epoch": 8.431762366974622,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6499,
      "step": 36046
    },
    {
      "epoch": 8.431996257747631,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 36047
    },
    {
      "epoch": 8.432230148520642,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 36048
    },
    {
      "epoch": 8.43246403929365,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 36049
    },
    {
      "epoch": 8.43269793006666,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 36050
    },
    {
      "epoch": 8.432931820839668,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 36051
    },
    {
      "epoch": 8.433165711612677,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.621,
      "step": 36052
    },
    {
      "epoch": 8.433399602385686,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.556,
      "step": 36053
    },
    {
      "epoch": 8.433633493158695,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 36054
    },
    {
      "epoch": 8.433867383931704,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 36055
    },
    {
      "epoch": 8.434101274704712,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 36056
    },
    {
      "epoch": 8.434335165477721,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 36057
    },
    {
      "epoch": 8.43456905625073,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 36058
    },
    {
      "epoch": 8.43480294702374,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 36059
    },
    {
      "epoch": 8.43503683779675,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 36060
    },
    {
      "epoch": 8.435270728569758,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 36061
    },
    {
      "epoch": 8.435504619342767,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 36062
    },
    {
      "epoch": 8.435738510115776,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 36063
    },
    {
      "epoch": 8.435972400888785,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.578,
      "step": 36064
    },
    {
      "epoch": 8.436206291661794,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.637,
      "step": 36065
    },
    {
      "epoch": 8.436440182434803,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 36066
    },
    {
      "epoch": 8.436674073207811,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 36067
    },
    {
      "epoch": 8.43690796398082,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5603,
      "step": 36068
    },
    {
      "epoch": 8.437141854753829,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 36069
    },
    {
      "epoch": 8.43737574552684,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.5285,
      "step": 36070
    },
    {
      "epoch": 8.437609636299849,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 36071
    },
    {
      "epoch": 8.437843527072857,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6142,
      "step": 36072
    },
    {
      "epoch": 8.438077417845866,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6044,
      "step": 36073
    },
    {
      "epoch": 8.438311308618875,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5399,
      "step": 36074
    },
    {
      "epoch": 8.438545199391884,
      "grad_norm": 7.5625,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 36075
    },
    {
      "epoch": 8.438779090164893,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.614,
      "step": 36076
    },
    {
      "epoch": 8.439012980937902,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4796,
      "step": 36077
    },
    {
      "epoch": 8.43924687171091,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5263,
      "step": 36078
    },
    {
      "epoch": 8.43948076248392,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 36079
    },
    {
      "epoch": 8.43971465325693,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6602,
      "step": 36080
    },
    {
      "epoch": 8.439948544029939,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 36081
    },
    {
      "epoch": 8.440182434802947,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 36082
    },
    {
      "epoch": 8.440416325575956,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7975,
      "step": 36083
    },
    {
      "epoch": 8.440650216348965,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 36084
    },
    {
      "epoch": 8.440884107121974,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 36085
    },
    {
      "epoch": 8.441117997894983,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 36086
    },
    {
      "epoch": 8.441351888667992,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 36087
    },
    {
      "epoch": 8.441585779441,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5688,
      "step": 36088
    },
    {
      "epoch": 8.44181967021401,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7307,
      "step": 36089
    },
    {
      "epoch": 8.44205356098702,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4854,
      "step": 36090
    },
    {
      "epoch": 8.442287451760029,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6356,
      "step": 36091
    },
    {
      "epoch": 8.442521342533038,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 36092
    },
    {
      "epoch": 8.442755233306046,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.7103,
      "step": 36093
    },
    {
      "epoch": 8.442989124079055,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9442,
      "step": 36094
    },
    {
      "epoch": 8.443223014852064,
      "grad_norm": 8.8125,
      "learning_rate": 3e-05,
      "loss": 1.7194,
      "step": 36095
    },
    {
      "epoch": 8.443456905625073,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5532,
      "step": 36096
    },
    {
      "epoch": 8.443690796398082,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7127,
      "step": 36097
    },
    {
      "epoch": 8.44392468717109,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5909,
      "step": 36098
    },
    {
      "epoch": 8.4441585779441,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 36099
    },
    {
      "epoch": 8.444392468717108,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.494,
      "step": 36100
    },
    {
      "epoch": 8.444392468717108,
      "eval_runtime": 4.6268,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 36100
    },
    {
      "epoch": 8.444626359490119,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6657,
      "step": 36101
    },
    {
      "epoch": 8.444860250263128,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 36102
    },
    {
      "epoch": 8.445094141036137,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5325,
      "step": 36103
    },
    {
      "epoch": 8.445328031809145,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 36104
    },
    {
      "epoch": 8.445561922582154,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 36105
    },
    {
      "epoch": 8.445795813355163,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 36106
    },
    {
      "epoch": 8.446029704128172,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 36107
    },
    {
      "epoch": 8.44626359490118,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 36108
    },
    {
      "epoch": 8.44649748567419,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5941,
      "step": 36109
    },
    {
      "epoch": 8.446731376447199,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 36110
    },
    {
      "epoch": 8.446965267220207,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 36111
    },
    {
      "epoch": 8.447199157993218,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 36112
    },
    {
      "epoch": 8.447433048766227,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5472,
      "step": 36113
    },
    {
      "epoch": 8.447666939539236,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 36114
    },
    {
      "epoch": 8.447900830312244,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 36115
    },
    {
      "epoch": 8.448134721085253,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 36116
    },
    {
      "epoch": 8.448368611858262,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 36117
    },
    {
      "epoch": 8.448602502631271,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7437,
      "step": 36118
    },
    {
      "epoch": 8.44883639340428,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 36119
    },
    {
      "epoch": 8.449070284177289,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 36120
    },
    {
      "epoch": 8.449304174950298,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 36121
    },
    {
      "epoch": 8.449538065723306,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 36122
    },
    {
      "epoch": 8.449771956496317,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 36123
    },
    {
      "epoch": 8.450005847269326,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5978,
      "step": 36124
    },
    {
      "epoch": 8.450239738042335,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 36125
    },
    {
      "epoch": 8.450473628815343,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 36126
    },
    {
      "epoch": 8.450707519588352,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4981,
      "step": 36127
    },
    {
      "epoch": 8.450941410361361,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 36128
    },
    {
      "epoch": 8.45117530113437,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6651,
      "step": 36129
    },
    {
      "epoch": 8.451409191907379,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 36130
    },
    {
      "epoch": 8.451643082680388,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 36131
    },
    {
      "epoch": 8.451876973453397,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7589,
      "step": 36132
    },
    {
      "epoch": 8.452110864226407,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 36133
    },
    {
      "epoch": 8.452344754999416,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6226,
      "step": 36134
    },
    {
      "epoch": 8.452578645772425,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 36135
    },
    {
      "epoch": 8.452812536545434,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 36136
    },
    {
      "epoch": 8.453046427318442,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 36137
    },
    {
      "epoch": 8.453280318091451,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7757,
      "step": 36138
    },
    {
      "epoch": 8.45351420886446,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 36139
    },
    {
      "epoch": 8.453748099637469,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.3496,
      "step": 36140
    },
    {
      "epoch": 8.453981990410478,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 36141
    },
    {
      "epoch": 8.454215881183487,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 36142
    },
    {
      "epoch": 8.454449771956495,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 36143
    },
    {
      "epoch": 8.454683662729506,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 36144
    },
    {
      "epoch": 8.454917553502515,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 36145
    },
    {
      "epoch": 8.455151444275524,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.4454,
      "step": 36146
    },
    {
      "epoch": 8.455385335048533,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 36147
    },
    {
      "epoch": 8.455619225821541,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.5217,
      "step": 36148
    },
    {
      "epoch": 8.45585311659455,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4636,
      "step": 36149
    },
    {
      "epoch": 8.45608700736756,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 36150
    },
    {
      "epoch": 8.456320898140568,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 36151
    },
    {
      "epoch": 8.456554788913577,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 36152
    },
    {
      "epoch": 8.456788679686586,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5325,
      "step": 36153
    },
    {
      "epoch": 8.457022570459596,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4686,
      "step": 36154
    },
    {
      "epoch": 8.457256461232605,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6901,
      "step": 36155
    },
    {
      "epoch": 8.457490352005614,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.731,
      "step": 36156
    },
    {
      "epoch": 8.457724242778623,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 36157
    },
    {
      "epoch": 8.457958133551632,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 36158
    },
    {
      "epoch": 8.45819202432464,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 36159
    },
    {
      "epoch": 8.45842591509765,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 36160
    },
    {
      "epoch": 8.458659805870658,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7872,
      "step": 36161
    },
    {
      "epoch": 8.458893696643667,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 36162
    },
    {
      "epoch": 8.459127587416676,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 36163
    },
    {
      "epoch": 8.459361478189685,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 36164
    },
    {
      "epoch": 8.459595368962695,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0835,
      "step": 36165
    },
    {
      "epoch": 8.459829259735704,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 36166
    },
    {
      "epoch": 8.460063150508713,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 36167
    },
    {
      "epoch": 8.460297041281722,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 36168
    },
    {
      "epoch": 8.46053093205473,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 36169
    },
    {
      "epoch": 8.46076482282774,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 36170
    },
    {
      "epoch": 8.460998713600748,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7536,
      "step": 36171
    },
    {
      "epoch": 8.461232604373757,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 36172
    },
    {
      "epoch": 8.461466495146766,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.5075,
      "step": 36173
    },
    {
      "epoch": 8.461700385919775,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6258,
      "step": 36174
    },
    {
      "epoch": 8.461934276692784,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 36175
    },
    {
      "epoch": 8.462168167465794,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 36176
    },
    {
      "epoch": 8.462402058238803,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 36177
    },
    {
      "epoch": 8.462635949011812,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 36178
    },
    {
      "epoch": 8.46286983978482,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 36179
    },
    {
      "epoch": 8.46310373055783,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 36180
    },
    {
      "epoch": 8.463337621330838,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 36181
    },
    {
      "epoch": 8.463571512103847,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 36182
    },
    {
      "epoch": 8.463805402876856,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 36183
    },
    {
      "epoch": 8.464039293649865,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6938,
      "step": 36184
    },
    {
      "epoch": 8.464273184422874,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7379,
      "step": 36185
    },
    {
      "epoch": 8.464507075195883,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 36186
    },
    {
      "epoch": 8.464740965968893,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 36187
    },
    {
      "epoch": 8.464974856741902,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.813,
      "step": 36188
    },
    {
      "epoch": 8.465208747514911,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 36189
    },
    {
      "epoch": 8.46544263828792,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5295,
      "step": 36190
    },
    {
      "epoch": 8.465676529060929,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 36191
    },
    {
      "epoch": 8.465910419833937,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6397,
      "step": 36192
    },
    {
      "epoch": 8.466144310606946,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6285,
      "step": 36193
    },
    {
      "epoch": 8.466378201379955,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5276,
      "step": 36194
    },
    {
      "epoch": 8.466612092152964,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 36195
    },
    {
      "epoch": 8.466845982925973,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7883,
      "step": 36196
    },
    {
      "epoch": 8.467079873698983,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 36197
    },
    {
      "epoch": 8.467313764471992,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 36198
    },
    {
      "epoch": 8.467547655245001,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6826,
      "step": 36199
    },
    {
      "epoch": 8.46778154601801,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.533,
      "step": 36200
    },
    {
      "epoch": 8.46778154601801,
      "eval_runtime": 4.6216,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 36200
    },
    {
      "epoch": 8.468015436791019,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 36201
    },
    {
      "epoch": 8.468249327564028,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 36202
    },
    {
      "epoch": 8.468483218337036,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 36203
    },
    {
      "epoch": 8.468717109110045,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 36204
    },
    {
      "epoch": 8.468950999883054,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6002,
      "step": 36205
    },
    {
      "epoch": 8.469184890656063,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 36206
    },
    {
      "epoch": 8.469418781429072,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 36207
    },
    {
      "epoch": 8.469652672202082,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 36208
    },
    {
      "epoch": 8.469886562975091,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 36209
    },
    {
      "epoch": 8.4701204537481,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 36210
    },
    {
      "epoch": 8.470354344521109,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6446,
      "step": 36211
    },
    {
      "epoch": 8.470588235294118,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 36212
    },
    {
      "epoch": 8.470822126067127,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 36213
    },
    {
      "epoch": 8.471056016840135,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 36214
    },
    {
      "epoch": 8.471289907613144,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5549,
      "step": 36215
    },
    {
      "epoch": 8.471523798386153,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6483,
      "step": 36216
    },
    {
      "epoch": 8.471757689159162,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 36217
    },
    {
      "epoch": 8.471991579932173,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5852,
      "step": 36218
    },
    {
      "epoch": 8.472225470705181,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7699,
      "step": 36219
    },
    {
      "epoch": 8.47245936147819,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 36220
    },
    {
      "epoch": 8.472693252251199,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 36221
    },
    {
      "epoch": 8.472927143024208,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7595,
      "step": 36222
    },
    {
      "epoch": 8.473161033797217,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 36223
    },
    {
      "epoch": 8.473394924570226,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7725,
      "step": 36224
    },
    {
      "epoch": 8.473628815343234,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 36225
    },
    {
      "epoch": 8.473862706116243,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 36226
    },
    {
      "epoch": 8.474096596889252,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.776,
      "step": 36227
    },
    {
      "epoch": 8.474330487662261,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 36228
    },
    {
      "epoch": 8.474564378435272,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 36229
    },
    {
      "epoch": 8.47479826920828,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5959,
      "step": 36230
    },
    {
      "epoch": 8.47503215998129,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.495,
      "step": 36231
    },
    {
      "epoch": 8.475266050754298,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5041,
      "step": 36232
    },
    {
      "epoch": 8.475499941527307,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4649,
      "step": 36233
    },
    {
      "epoch": 8.475733832300316,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 36234
    },
    {
      "epoch": 8.475967723073325,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 36235
    },
    {
      "epoch": 8.476201613846333,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 36236
    },
    {
      "epoch": 8.476435504619342,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 36237
    },
    {
      "epoch": 8.476669395392351,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8363,
      "step": 36238
    },
    {
      "epoch": 8.47690328616536,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8203,
      "step": 36239
    },
    {
      "epoch": 8.47713717693837,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 36240
    },
    {
      "epoch": 8.47737106771138,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 36241
    },
    {
      "epoch": 8.477604958484388,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.4658,
      "step": 36242
    },
    {
      "epoch": 8.477838849257397,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 36243
    },
    {
      "epoch": 8.478072740030406,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6623,
      "step": 36244
    },
    {
      "epoch": 8.478306630803415,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 36245
    },
    {
      "epoch": 8.478540521576424,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 36246
    },
    {
      "epoch": 8.478774412349432,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 36247
    },
    {
      "epoch": 8.479008303122441,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 36248
    },
    {
      "epoch": 8.47924219389545,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6019,
      "step": 36249
    },
    {
      "epoch": 8.479476084668459,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 36250
    },
    {
      "epoch": 8.47970997544147,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 36251
    },
    {
      "epoch": 8.479943866214478,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6867,
      "step": 36252
    },
    {
      "epoch": 8.480177756987487,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.603,
      "step": 36253
    },
    {
      "epoch": 8.480411647760496,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5995,
      "step": 36254
    },
    {
      "epoch": 8.480645538533505,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5939,
      "step": 36255
    },
    {
      "epoch": 8.480879429306514,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 36256
    },
    {
      "epoch": 8.481113320079523,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 36257
    },
    {
      "epoch": 8.481347210852531,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 36258
    },
    {
      "epoch": 8.48158110162554,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 36259
    },
    {
      "epoch": 8.481814992398549,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 36260
    },
    {
      "epoch": 8.48204888317156,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 36261
    },
    {
      "epoch": 8.482282773944569,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 36262
    },
    {
      "epoch": 8.482516664717577,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6164,
      "step": 36263
    },
    {
      "epoch": 8.482750555490586,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7278,
      "step": 36264
    },
    {
      "epoch": 8.482984446263595,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.591,
      "step": 36265
    },
    {
      "epoch": 8.483218337036604,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 36266
    },
    {
      "epoch": 8.483452227809613,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 36267
    },
    {
      "epoch": 8.483686118582622,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6561,
      "step": 36268
    },
    {
      "epoch": 8.48392000935563,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 36269
    },
    {
      "epoch": 8.48415390012864,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 36270
    },
    {
      "epoch": 8.484387790901648,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4797,
      "step": 36271
    },
    {
      "epoch": 8.484621681674659,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7258,
      "step": 36272
    },
    {
      "epoch": 8.484855572447668,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7092,
      "step": 36273
    },
    {
      "epoch": 8.485089463220676,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 36274
    },
    {
      "epoch": 8.485323353993685,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 36275
    },
    {
      "epoch": 8.485557244766694,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 36276
    },
    {
      "epoch": 8.485791135539703,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7483,
      "step": 36277
    },
    {
      "epoch": 8.486025026312712,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 36278
    },
    {
      "epoch": 8.48625891708572,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 36279
    },
    {
      "epoch": 8.48649280785873,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 36280
    },
    {
      "epoch": 8.486726698631738,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7241,
      "step": 36281
    },
    {
      "epoch": 8.486960589404749,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 36282
    },
    {
      "epoch": 8.487194480177758,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 36283
    },
    {
      "epoch": 8.487428370950767,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.614,
      "step": 36284
    },
    {
      "epoch": 8.487662261723775,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 36285
    },
    {
      "epoch": 8.487896152496784,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 36286
    },
    {
      "epoch": 8.488130043269793,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 36287
    },
    {
      "epoch": 8.488363934042802,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 36288
    },
    {
      "epoch": 8.48859782481581,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8183,
      "step": 36289
    },
    {
      "epoch": 8.48883171558882,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.3931,
      "step": 36290
    },
    {
      "epoch": 8.489065606361828,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.684,
      "step": 36291
    },
    {
      "epoch": 8.489299497134837,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 36292
    },
    {
      "epoch": 8.489533387907848,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.606,
      "step": 36293
    },
    {
      "epoch": 8.489767278680857,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6513,
      "step": 36294
    },
    {
      "epoch": 8.490001169453866,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6993,
      "step": 36295
    },
    {
      "epoch": 8.490235060226874,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 36296
    },
    {
      "epoch": 8.490468950999883,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.7266,
      "step": 36297
    },
    {
      "epoch": 8.490702841772892,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.15,
      "step": 36298
    },
    {
      "epoch": 8.4909367325459,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 36299
    },
    {
      "epoch": 8.49117062331891,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7453,
      "step": 36300
    },
    {
      "epoch": 8.49117062331891,
      "eval_runtime": 4.6154,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 36300
    },
    {
      "epoch": 8.491404514091919,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 36301
    },
    {
      "epoch": 8.491638404864927,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7443,
      "step": 36302
    },
    {
      "epoch": 8.491872295637936,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 36303
    },
    {
      "epoch": 8.492106186410947,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5096,
      "step": 36304
    },
    {
      "epoch": 8.492340077183956,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.3774,
      "step": 36305
    },
    {
      "epoch": 8.492573967956965,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 36306
    },
    {
      "epoch": 8.492807858729973,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 36307
    },
    {
      "epoch": 8.493041749502982,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5682,
      "step": 36308
    },
    {
      "epoch": 8.493275640275991,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5872,
      "step": 36309
    },
    {
      "epoch": 8.493509531049,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 36310
    },
    {
      "epoch": 8.493743421822009,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 36311
    },
    {
      "epoch": 8.493977312595018,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 36312
    },
    {
      "epoch": 8.494211203368026,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 36313
    },
    {
      "epoch": 8.494445094141037,
      "grad_norm": 8.5,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 36314
    },
    {
      "epoch": 8.494678984914046,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6389,
      "step": 36315
    },
    {
      "epoch": 8.494912875687055,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7597,
      "step": 36316
    },
    {
      "epoch": 8.495146766460064,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5139,
      "step": 36317
    },
    {
      "epoch": 8.495380657233072,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 36318
    },
    {
      "epoch": 8.495614548006081,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5641,
      "step": 36319
    },
    {
      "epoch": 8.49584843877909,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 36320
    },
    {
      "epoch": 8.496082329552099,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5338,
      "step": 36321
    },
    {
      "epoch": 8.496316220325108,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 36322
    },
    {
      "epoch": 8.496550111098117,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 36323
    },
    {
      "epoch": 8.496784001871125,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7688,
      "step": 36324
    },
    {
      "epoch": 8.497017892644136,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.5426,
      "step": 36325
    },
    {
      "epoch": 8.497251783417145,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6095,
      "step": 36326
    },
    {
      "epoch": 8.497485674190154,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6192,
      "step": 36327
    },
    {
      "epoch": 8.497719564963162,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 36328
    },
    {
      "epoch": 8.497953455736171,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 36329
    },
    {
      "epoch": 8.49818734650918,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6084,
      "step": 36330
    },
    {
      "epoch": 8.498421237282189,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7426,
      "step": 36331
    },
    {
      "epoch": 8.498655128055198,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 36332
    },
    {
      "epoch": 8.498889018828207,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5047,
      "step": 36333
    },
    {
      "epoch": 8.499122909601216,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7421,
      "step": 36334
    },
    {
      "epoch": 8.499356800374226,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6326,
      "step": 36335
    },
    {
      "epoch": 8.499590691147235,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7021,
      "step": 36336
    },
    {
      "epoch": 8.499824581920244,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 36337
    },
    {
      "epoch": 8.500058472693253,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5807,
      "step": 36338
    },
    {
      "epoch": 8.500292363466261,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 36339
    },
    {
      "epoch": 8.50052625423927,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 36340
    },
    {
      "epoch": 8.50076014501228,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4528,
      "step": 36341
    },
    {
      "epoch": 8.500994035785288,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5745,
      "step": 36342
    },
    {
      "epoch": 8.501227926558297,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 36343
    },
    {
      "epoch": 8.501461817331306,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5347,
      "step": 36344
    },
    {
      "epoch": 8.501695708104315,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 36345
    },
    {
      "epoch": 8.501929598877325,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5569,
      "step": 36346
    },
    {
      "epoch": 8.502163489650334,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5342,
      "step": 36347
    },
    {
      "epoch": 8.502397380423343,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5355,
      "step": 36348
    },
    {
      "epoch": 8.502631271196352,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 36349
    },
    {
      "epoch": 8.50286516196936,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 36350
    },
    {
      "epoch": 8.50309905274237,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 36351
    },
    {
      "epoch": 8.503332943515378,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 36352
    },
    {
      "epoch": 8.503566834288387,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6943,
      "step": 36353
    },
    {
      "epoch": 8.503800725061396,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 36354
    },
    {
      "epoch": 8.504034615834405,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 36355
    },
    {
      "epoch": 8.504268506607414,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 36356
    },
    {
      "epoch": 8.504502397380424,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5639,
      "step": 36357
    },
    {
      "epoch": 8.504736288153433,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 36358
    },
    {
      "epoch": 8.504970178926442,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6342,
      "step": 36359
    },
    {
      "epoch": 8.50520406969945,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6279,
      "step": 36360
    },
    {
      "epoch": 8.50543796047246,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.4549,
      "step": 36361
    },
    {
      "epoch": 8.505671851245468,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6502,
      "step": 36362
    },
    {
      "epoch": 8.505905742018477,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5981,
      "step": 36363
    },
    {
      "epoch": 8.506139632791486,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6273,
      "step": 36364
    },
    {
      "epoch": 8.506373523564495,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6525,
      "step": 36365
    },
    {
      "epoch": 8.506607414337504,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6107,
      "step": 36366
    },
    {
      "epoch": 8.506841305110513,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 36367
    },
    {
      "epoch": 8.507075195883523,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.3658,
      "step": 36368
    },
    {
      "epoch": 8.507309086656532,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 36369
    },
    {
      "epoch": 8.50754297742954,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.5213,
      "step": 36370
    },
    {
      "epoch": 8.50777686820255,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6323,
      "step": 36371
    },
    {
      "epoch": 8.508010758975558,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9196,
      "step": 36372
    },
    {
      "epoch": 8.508244649748567,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6257,
      "step": 36373
    },
    {
      "epoch": 8.508478540521576,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 36374
    },
    {
      "epoch": 8.508712431294585,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 36375
    },
    {
      "epoch": 8.508946322067594,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5846,
      "step": 36376
    },
    {
      "epoch": 8.509180212840603,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 36377
    },
    {
      "epoch": 8.509414103613612,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 36378
    },
    {
      "epoch": 8.509647994386622,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4637,
      "step": 36379
    },
    {
      "epoch": 8.509881885159631,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 36380
    },
    {
      "epoch": 8.51011577593264,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 36381
    },
    {
      "epoch": 8.510349666705649,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6032,
      "step": 36382
    },
    {
      "epoch": 8.510583557478657,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6835,
      "step": 36383
    },
    {
      "epoch": 8.510817448251666,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 36384
    },
    {
      "epoch": 8.511051339024675,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 36385
    },
    {
      "epoch": 8.511285229797684,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4627,
      "step": 36386
    },
    {
      "epoch": 8.511519120570693,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6323,
      "step": 36387
    },
    {
      "epoch": 8.511753011343703,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 36388
    },
    {
      "epoch": 8.511986902116712,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6275,
      "step": 36389
    },
    {
      "epoch": 8.512220792889721,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 36390
    },
    {
      "epoch": 8.51245468366273,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 36391
    },
    {
      "epoch": 8.512688574435739,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.635,
      "step": 36392
    },
    {
      "epoch": 8.512922465208748,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 36393
    },
    {
      "epoch": 8.513156355981756,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 36394
    },
    {
      "epoch": 8.513390246754765,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.4427,
      "step": 36395
    },
    {
      "epoch": 8.513624137527774,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6322,
      "step": 36396
    },
    {
      "epoch": 8.513858028300783,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6143,
      "step": 36397
    },
    {
      "epoch": 8.514091919073792,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 36398
    },
    {
      "epoch": 8.514325809846802,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 36399
    },
    {
      "epoch": 8.514559700619811,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6579,
      "step": 36400
    },
    {
      "epoch": 8.514559700619811,
      "eval_runtime": 4.6176,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 36400
    },
    {
      "epoch": 8.51479359139282,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 36401
    },
    {
      "epoch": 8.515027482165829,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6828,
      "step": 36402
    },
    {
      "epoch": 8.515261372938838,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 36403
    },
    {
      "epoch": 8.515495263711847,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 36404
    },
    {
      "epoch": 8.515729154484855,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.2499,
      "step": 36405
    },
    {
      "epoch": 8.515963045257864,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4094,
      "step": 36406
    },
    {
      "epoch": 8.516196936030873,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 36407
    },
    {
      "epoch": 8.516430826803882,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6169,
      "step": 36408
    },
    {
      "epoch": 8.51666471757689,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 36409
    },
    {
      "epoch": 8.516898608349901,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 36410
    },
    {
      "epoch": 8.51713249912291,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7186,
      "step": 36411
    },
    {
      "epoch": 8.517366389895919,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 36412
    },
    {
      "epoch": 8.517600280668928,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 36413
    },
    {
      "epoch": 8.517834171441937,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 36414
    },
    {
      "epoch": 8.518068062214946,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6319,
      "step": 36415
    },
    {
      "epoch": 8.518301952987954,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6027,
      "step": 36416
    },
    {
      "epoch": 8.518535843760963,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 36417
    },
    {
      "epoch": 8.518769734533972,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 36418
    },
    {
      "epoch": 8.519003625306981,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 36419
    },
    {
      "epoch": 8.51923751607999,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 36420
    },
    {
      "epoch": 8.519471406853,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 36421
    },
    {
      "epoch": 8.51970529762601,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 36422
    },
    {
      "epoch": 8.519939188399018,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 36423
    },
    {
      "epoch": 8.520173079172027,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6118,
      "step": 36424
    },
    {
      "epoch": 8.520406969945036,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 36425
    },
    {
      "epoch": 8.520640860718045,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 36426
    },
    {
      "epoch": 8.520874751491053,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 36427
    },
    {
      "epoch": 8.521108642264062,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6923,
      "step": 36428
    },
    {
      "epoch": 8.521342533037071,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7207,
      "step": 36429
    },
    {
      "epoch": 8.52157642381008,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7037,
      "step": 36430
    },
    {
      "epoch": 8.521810314583089,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6437,
      "step": 36431
    },
    {
      "epoch": 8.5220442053561,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6548,
      "step": 36432
    },
    {
      "epoch": 8.522278096129108,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 36433
    },
    {
      "epoch": 8.522511986902117,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7687,
      "step": 36434
    },
    {
      "epoch": 8.522745877675126,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.5966,
      "step": 36435
    },
    {
      "epoch": 8.522979768448135,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 36436
    },
    {
      "epoch": 8.523213659221144,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 36437
    },
    {
      "epoch": 8.523447549994152,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6337,
      "step": 36438
    },
    {
      "epoch": 8.523681440767161,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 36439
    },
    {
      "epoch": 8.52391533154017,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 36440
    },
    {
      "epoch": 8.524149222313179,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 36441
    },
    {
      "epoch": 8.52438311308619,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.681,
      "step": 36442
    },
    {
      "epoch": 8.524617003859198,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 36443
    },
    {
      "epoch": 8.524850894632207,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 36444
    },
    {
      "epoch": 8.525084785405216,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 36445
    },
    {
      "epoch": 8.525318676178225,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 36446
    },
    {
      "epoch": 8.525552566951234,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6467,
      "step": 36447
    },
    {
      "epoch": 8.525786457724243,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 36448
    },
    {
      "epoch": 8.526020348497251,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 36449
    },
    {
      "epoch": 8.52625423927026,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 36450
    },
    {
      "epoch": 8.526488130043269,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5759,
      "step": 36451
    },
    {
      "epoch": 8.52672202081628,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7846,
      "step": 36452
    },
    {
      "epoch": 8.526955911589289,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 36453
    },
    {
      "epoch": 8.527189802362297,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 36454
    },
    {
      "epoch": 8.527423693135306,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7099,
      "step": 36455
    },
    {
      "epoch": 8.527657583908315,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4479,
      "step": 36456
    },
    {
      "epoch": 8.527891474681324,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 36457
    },
    {
      "epoch": 8.528125365454333,
      "grad_norm": 7.375,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 36458
    },
    {
      "epoch": 8.528359256227342,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8501,
      "step": 36459
    },
    {
      "epoch": 8.52859314700035,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 36460
    },
    {
      "epoch": 8.52882703777336,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 36461
    },
    {
      "epoch": 8.529060928546368,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5961,
      "step": 36462
    },
    {
      "epoch": 8.529294819319379,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7349,
      "step": 36463
    },
    {
      "epoch": 8.529528710092388,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5007,
      "step": 36464
    },
    {
      "epoch": 8.529762600865396,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 36465
    },
    {
      "epoch": 8.529996491638405,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 36466
    },
    {
      "epoch": 8.530230382411414,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5959,
      "step": 36467
    },
    {
      "epoch": 8.530464273184423,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 36468
    },
    {
      "epoch": 8.530698163957432,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 36469
    },
    {
      "epoch": 8.53093205473044,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6748,
      "step": 36470
    },
    {
      "epoch": 8.53116594550345,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 36471
    },
    {
      "epoch": 8.531399836276458,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 36472
    },
    {
      "epoch": 8.531633727049467,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 36473
    },
    {
      "epoch": 8.531867617822478,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 36474
    },
    {
      "epoch": 8.532101508595487,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 36475
    },
    {
      "epoch": 8.532335399368495,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.634,
      "step": 36476
    },
    {
      "epoch": 8.532569290141504,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 36477
    },
    {
      "epoch": 8.532803180914513,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6901,
      "step": 36478
    },
    {
      "epoch": 8.533037071687522,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 36479
    },
    {
      "epoch": 8.53327096246053,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 36480
    },
    {
      "epoch": 8.53350485323354,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 36481
    },
    {
      "epoch": 8.533738744006548,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 36482
    },
    {
      "epoch": 8.533972634779557,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5816,
      "step": 36483
    },
    {
      "epoch": 8.534206525552566,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 36484
    },
    {
      "epoch": 8.534440416325577,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5074,
      "step": 36485
    },
    {
      "epoch": 8.534674307098586,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 36486
    },
    {
      "epoch": 8.534908197871594,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 36487
    },
    {
      "epoch": 8.535142088644603,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 36488
    },
    {
      "epoch": 8.535375979417612,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 36489
    },
    {
      "epoch": 8.535609870190621,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7037,
      "step": 36490
    },
    {
      "epoch": 8.53584376096363,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 36491
    },
    {
      "epoch": 8.536077651736639,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 36492
    },
    {
      "epoch": 8.536311542509647,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.502,
      "step": 36493
    },
    {
      "epoch": 8.536545433282656,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 36494
    },
    {
      "epoch": 8.536779324055665,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8902,
      "step": 36495
    },
    {
      "epoch": 8.537013214828676,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6997,
      "step": 36496
    },
    {
      "epoch": 8.537247105601685,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 36497
    },
    {
      "epoch": 8.537480996374693,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 36498
    },
    {
      "epoch": 8.537714887147702,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 36499
    },
    {
      "epoch": 8.537948777920711,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7816,
      "step": 36500
    },
    {
      "epoch": 8.537948777920711,
      "eval_runtime": 4.6476,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 36500
    },
    {
      "epoch": 8.53818266869372,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5546,
      "step": 36501
    },
    {
      "epoch": 8.538416559466729,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7764,
      "step": 36502
    },
    {
      "epoch": 8.538650450239738,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8891,
      "step": 36503
    },
    {
      "epoch": 8.538884341012746,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.4614,
      "step": 36504
    },
    {
      "epoch": 8.539118231785755,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6123,
      "step": 36505
    },
    {
      "epoch": 8.539352122558766,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 36506
    },
    {
      "epoch": 8.539586013331775,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 36507
    },
    {
      "epoch": 8.539819904104784,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 36508
    },
    {
      "epoch": 8.540053794877792,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 36509
    },
    {
      "epoch": 8.540287685650801,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6645,
      "step": 36510
    },
    {
      "epoch": 8.54052157642381,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 36511
    },
    {
      "epoch": 8.540755467196819,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 36512
    },
    {
      "epoch": 8.540989357969828,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.3752,
      "step": 36513
    },
    {
      "epoch": 8.541223248742837,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.651,
      "step": 36514
    },
    {
      "epoch": 8.541457139515845,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.7241,
      "step": 36515
    },
    {
      "epoch": 8.541691030288856,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7007,
      "step": 36516
    },
    {
      "epoch": 8.541924921061865,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6387,
      "step": 36517
    },
    {
      "epoch": 8.542158811834874,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 36518
    },
    {
      "epoch": 8.542392702607883,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 36519
    },
    {
      "epoch": 8.542626593380891,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 36520
    },
    {
      "epoch": 8.5428604841539,
      "grad_norm": 8.125,
      "learning_rate": 3e-05,
      "loss": 1.4864,
      "step": 36521
    },
    {
      "epoch": 8.543094374926909,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.5704,
      "step": 36522
    },
    {
      "epoch": 8.543328265699918,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6725,
      "step": 36523
    },
    {
      "epoch": 8.543562156472927,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 36524
    },
    {
      "epoch": 8.543796047245936,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 36525
    },
    {
      "epoch": 8.544029938018944,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 36526
    },
    {
      "epoch": 8.544263828791955,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 36527
    },
    {
      "epoch": 8.544497719564964,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 36528
    },
    {
      "epoch": 8.544731610337973,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 36529
    },
    {
      "epoch": 8.544965501110982,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 36530
    },
    {
      "epoch": 8.54519939188399,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5859,
      "step": 36531
    },
    {
      "epoch": 8.545433282657,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 36532
    },
    {
      "epoch": 8.545667173430008,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7536,
      "step": 36533
    },
    {
      "epoch": 8.545901064203017,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8257,
      "step": 36534
    },
    {
      "epoch": 8.546134954976026,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 36535
    },
    {
      "epoch": 8.546368845749035,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.865,
      "step": 36536
    },
    {
      "epoch": 8.546602736522043,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 36537
    },
    {
      "epoch": 8.546836627295054,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 36538
    },
    {
      "epoch": 8.547070518068063,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 36539
    },
    {
      "epoch": 8.547304408841072,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 36540
    },
    {
      "epoch": 8.54753829961408,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 36541
    },
    {
      "epoch": 8.54777219038709,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5656,
      "step": 36542
    },
    {
      "epoch": 8.548006081160098,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 36543
    },
    {
      "epoch": 8.548239971933107,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 36544
    },
    {
      "epoch": 8.548473862706116,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 36545
    },
    {
      "epoch": 8.548707753479125,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 36546
    },
    {
      "epoch": 8.548941644252134,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8369,
      "step": 36547
    },
    {
      "epoch": 8.549175535025142,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 36548
    },
    {
      "epoch": 8.549409425798153,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6231,
      "step": 36549
    },
    {
      "epoch": 8.549643316571162,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 36550
    },
    {
      "epoch": 8.54987720734417,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.56,
      "step": 36551
    },
    {
      "epoch": 8.55011109811718,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8476,
      "step": 36552
    },
    {
      "epoch": 8.550344988890188,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 36553
    },
    {
      "epoch": 8.550578879663197,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 36554
    },
    {
      "epoch": 8.550812770436206,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6202,
      "step": 36555
    },
    {
      "epoch": 8.551046661209215,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 36556
    },
    {
      "epoch": 8.551280551982224,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 36557
    },
    {
      "epoch": 8.551514442755233,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 36558
    },
    {
      "epoch": 8.551748333528241,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5406,
      "step": 36559
    },
    {
      "epoch": 8.551982224301252,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 36560
    },
    {
      "epoch": 8.55221611507426,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 36561
    },
    {
      "epoch": 8.55245000584727,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5898,
      "step": 36562
    },
    {
      "epoch": 8.552683896620279,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 36563
    },
    {
      "epoch": 8.552917787393287,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6524,
      "step": 36564
    },
    {
      "epoch": 8.553151678166296,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 36565
    },
    {
      "epoch": 8.553385568939305,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6216,
      "step": 36566
    },
    {
      "epoch": 8.553619459712314,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 36567
    },
    {
      "epoch": 8.553853350485323,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 36568
    },
    {
      "epoch": 8.554087241258332,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 36569
    },
    {
      "epoch": 8.554321132031342,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5529,
      "step": 36570
    },
    {
      "epoch": 8.554555022804351,
      "grad_norm": 11.3125,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 36571
    },
    {
      "epoch": 8.55478891357736,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 36572
    },
    {
      "epoch": 8.555022804350369,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 36573
    },
    {
      "epoch": 8.555256695123377,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 36574
    },
    {
      "epoch": 8.555490585896386,
      "grad_norm": 2.5625,
      "learning_rate": 3e-05,
      "loss": 1.3282,
      "step": 36575
    },
    {
      "epoch": 8.555724476669395,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 36576
    },
    {
      "epoch": 8.555958367442404,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 36577
    },
    {
      "epoch": 8.556192258215413,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7384,
      "step": 36578
    },
    {
      "epoch": 8.556426148988422,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6925,
      "step": 36579
    },
    {
      "epoch": 8.556660039761432,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 36580
    },
    {
      "epoch": 8.556893930534441,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 36581
    },
    {
      "epoch": 8.55712782130745,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8223,
      "step": 36582
    },
    {
      "epoch": 8.557361712080459,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7698,
      "step": 36583
    },
    {
      "epoch": 8.557595602853468,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5235,
      "step": 36584
    },
    {
      "epoch": 8.557829493626476,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6699,
      "step": 36585
    },
    {
      "epoch": 8.558063384399485,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6137,
      "step": 36586
    },
    {
      "epoch": 8.558297275172494,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.563,
      "step": 36587
    },
    {
      "epoch": 8.558531165945503,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 36588
    },
    {
      "epoch": 8.558765056718512,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 36589
    },
    {
      "epoch": 8.55899894749152,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 36590
    },
    {
      "epoch": 8.559232838264531,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 36591
    },
    {
      "epoch": 8.55946672903754,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6654,
      "step": 36592
    },
    {
      "epoch": 8.559700619810549,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6259,
      "step": 36593
    },
    {
      "epoch": 8.559934510583558,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.4822,
      "step": 36594
    },
    {
      "epoch": 8.560168401356567,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7335,
      "step": 36595
    },
    {
      "epoch": 8.560402292129575,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 36596
    },
    {
      "epoch": 8.560636182902584,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7239,
      "step": 36597
    },
    {
      "epoch": 8.560870073675593,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6374,
      "step": 36598
    },
    {
      "epoch": 8.561103964448602,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 36599
    },
    {
      "epoch": 8.56133785522161,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 36600
    },
    {
      "epoch": 8.56133785522161,
      "eval_runtime": 4.6522,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 36600
    },
    {
      "epoch": 8.56157174599462,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7888,
      "step": 36601
    },
    {
      "epoch": 8.56180563676763,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 36602
    },
    {
      "epoch": 8.56203952754064,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5549,
      "step": 36603
    },
    {
      "epoch": 8.562273418313648,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 36604
    },
    {
      "epoch": 8.562507309086657,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 36605
    },
    {
      "epoch": 8.562741199859666,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 36606
    },
    {
      "epoch": 8.562975090632674,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7101,
      "step": 36607
    },
    {
      "epoch": 8.563208981405683,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5504,
      "step": 36608
    },
    {
      "epoch": 8.563442872178692,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4732,
      "step": 36609
    },
    {
      "epoch": 8.563676762951701,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7468,
      "step": 36610
    },
    {
      "epoch": 8.56391065372471,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6403,
      "step": 36611
    },
    {
      "epoch": 8.564144544497719,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 36612
    },
    {
      "epoch": 8.56437843527073,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 36613
    },
    {
      "epoch": 8.564612326043738,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6461,
      "step": 36614
    },
    {
      "epoch": 8.564846216816747,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 36615
    },
    {
      "epoch": 8.565080107589756,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 36616
    },
    {
      "epoch": 8.565313998362765,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 36617
    },
    {
      "epoch": 8.565547889135773,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7104,
      "step": 36618
    },
    {
      "epoch": 8.565781779908782,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5353,
      "step": 36619
    },
    {
      "epoch": 8.566015670681791,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 36620
    },
    {
      "epoch": 8.5662495614548,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 36621
    },
    {
      "epoch": 8.566483452227809,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 36622
    },
    {
      "epoch": 8.56671734300082,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 36623
    },
    {
      "epoch": 8.566951233773828,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0396,
      "step": 36624
    },
    {
      "epoch": 8.567185124546837,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 36625
    },
    {
      "epoch": 8.567419015319846,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7007,
      "step": 36626
    },
    {
      "epoch": 8.567652906092855,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7927,
      "step": 36627
    },
    {
      "epoch": 8.567886796865864,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6841,
      "step": 36628
    },
    {
      "epoch": 8.568120687638872,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 36629
    },
    {
      "epoch": 8.568354578411881,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 36630
    },
    {
      "epoch": 8.56858846918489,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 36631
    },
    {
      "epoch": 8.568822359957899,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 36632
    },
    {
      "epoch": 8.56905625073091,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.655,
      "step": 36633
    },
    {
      "epoch": 8.569290141503918,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7663,
      "step": 36634
    },
    {
      "epoch": 8.569524032276927,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4389,
      "step": 36635
    },
    {
      "epoch": 8.569757923049936,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6793,
      "step": 36636
    },
    {
      "epoch": 8.569991813822945,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5302,
      "step": 36637
    },
    {
      "epoch": 8.570225704595954,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6384,
      "step": 36638
    },
    {
      "epoch": 8.570459595368963,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7238,
      "step": 36639
    },
    {
      "epoch": 8.570693486141971,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 36640
    },
    {
      "epoch": 8.57092737691498,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 36641
    },
    {
      "epoch": 8.57116126768799,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 36642
    },
    {
      "epoch": 8.571395158460998,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 36643
    },
    {
      "epoch": 8.571629049234009,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 36644
    },
    {
      "epoch": 8.571862940007017,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 36645
    },
    {
      "epoch": 8.572096830780026,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 36646
    },
    {
      "epoch": 8.572330721553035,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 36647
    },
    {
      "epoch": 8.572564612326044,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 36648
    },
    {
      "epoch": 8.572798503099053,
      "grad_norm": 7.96875,
      "learning_rate": 3e-05,
      "loss": 1.6981,
      "step": 36649
    },
    {
      "epoch": 8.573032393872062,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.4002,
      "step": 36650
    },
    {
      "epoch": 8.57326628464507,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 36651
    },
    {
      "epoch": 8.57350017541808,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5458,
      "step": 36652
    },
    {
      "epoch": 8.573734066191088,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 36653
    },
    {
      "epoch": 8.573967956964097,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 36654
    },
    {
      "epoch": 8.574201847737108,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 36655
    },
    {
      "epoch": 8.574435738510116,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7868,
      "step": 36656
    },
    {
      "epoch": 8.574669629283125,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 36657
    },
    {
      "epoch": 8.574903520056134,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5787,
      "step": 36658
    },
    {
      "epoch": 8.575137410829143,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 36659
    },
    {
      "epoch": 8.575371301602152,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.725,
      "step": 36660
    },
    {
      "epoch": 8.57560519237516,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.4639,
      "step": 36661
    },
    {
      "epoch": 8.57583908314817,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 36662
    },
    {
      "epoch": 8.576072973921178,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 36663
    },
    {
      "epoch": 8.576306864694187,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 36664
    },
    {
      "epoch": 8.576540755467196,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 36665
    },
    {
      "epoch": 8.576774646240207,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6799,
      "step": 36666
    },
    {
      "epoch": 8.577008537013215,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6564,
      "step": 36667
    },
    {
      "epoch": 8.577242427786224,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6057,
      "step": 36668
    },
    {
      "epoch": 8.577476318559233,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 36669
    },
    {
      "epoch": 8.577710209332242,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 36670
    },
    {
      "epoch": 8.57794410010525,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.6325,
      "step": 36671
    },
    {
      "epoch": 8.57817799087826,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.4945,
      "step": 36672
    },
    {
      "epoch": 8.578411881651268,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6474,
      "step": 36673
    },
    {
      "epoch": 8.578645772424277,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8141,
      "step": 36674
    },
    {
      "epoch": 8.578879663197286,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 36675
    },
    {
      "epoch": 8.579113553970295,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5621,
      "step": 36676
    },
    {
      "epoch": 8.579347444743306,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 36677
    },
    {
      "epoch": 8.579581335516314,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.634,
      "step": 36678
    },
    {
      "epoch": 8.579815226289323,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6575,
      "step": 36679
    },
    {
      "epoch": 8.580049117062332,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 36680
    },
    {
      "epoch": 8.580283007835341,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5562,
      "step": 36681
    },
    {
      "epoch": 8.58051689860835,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6338,
      "step": 36682
    },
    {
      "epoch": 8.580750789381359,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 36683
    },
    {
      "epoch": 8.580984680154367,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7343,
      "step": 36684
    },
    {
      "epoch": 8.581218570927376,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.6026,
      "step": 36685
    },
    {
      "epoch": 8.581452461700385,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 36686
    },
    {
      "epoch": 8.581686352473396,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 36687
    },
    {
      "epoch": 8.581920243246405,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7242,
      "step": 36688
    },
    {
      "epoch": 8.582154134019413,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 36689
    },
    {
      "epoch": 8.582388024792422,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 36690
    },
    {
      "epoch": 8.582621915565431,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 36691
    },
    {
      "epoch": 8.58285580633844,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6978,
      "step": 36692
    },
    {
      "epoch": 8.583089697111449,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 36693
    },
    {
      "epoch": 8.583323587884458,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 36694
    },
    {
      "epoch": 8.583557478657466,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 36695
    },
    {
      "epoch": 8.583791369430475,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5578,
      "step": 36696
    },
    {
      "epoch": 8.584025260203486,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 36697
    },
    {
      "epoch": 8.584259150976495,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7772,
      "step": 36698
    },
    {
      "epoch": 8.584493041749504,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.4568,
      "step": 36699
    },
    {
      "epoch": 8.584726932522512,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 36700
    },
    {
      "epoch": 8.584726932522512,
      "eval_runtime": 4.6381,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 36700
    },
    {
      "epoch": 8.584960823295521,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5904,
      "step": 36701
    },
    {
      "epoch": 8.58519471406853,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5616,
      "step": 36702
    },
    {
      "epoch": 8.585428604841539,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 36703
    },
    {
      "epoch": 8.585662495614548,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 36704
    },
    {
      "epoch": 8.585896386387557,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 36705
    },
    {
      "epoch": 8.586130277160565,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.7335,
      "step": 36706
    },
    {
      "epoch": 8.586364167933574,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.458,
      "step": 36707
    },
    {
      "epoch": 8.586598058706585,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6083,
      "step": 36708
    },
    {
      "epoch": 8.586831949479594,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5532,
      "step": 36709
    },
    {
      "epoch": 8.587065840252603,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 36710
    },
    {
      "epoch": 8.587299731025611,
      "grad_norm": 11.0625,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 36711
    },
    {
      "epoch": 8.58753362179862,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7773,
      "step": 36712
    },
    {
      "epoch": 8.587767512571629,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 36713
    },
    {
      "epoch": 8.588001403344638,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.4034,
      "step": 36714
    },
    {
      "epoch": 8.588235294117647,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 36715
    },
    {
      "epoch": 8.588469184890656,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7814,
      "step": 36716
    },
    {
      "epoch": 8.588703075663664,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6254,
      "step": 36717
    },
    {
      "epoch": 8.588936966436673,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5336,
      "step": 36718
    },
    {
      "epoch": 8.589170857209684,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 36719
    },
    {
      "epoch": 8.589404747982693,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 36720
    },
    {
      "epoch": 8.589638638755702,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6234,
      "step": 36721
    },
    {
      "epoch": 8.58987252952871,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 36722
    },
    {
      "epoch": 8.59010642030172,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6338,
      "step": 36723
    },
    {
      "epoch": 8.590340311074728,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 36724
    },
    {
      "epoch": 8.590574201847737,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7327,
      "step": 36725
    },
    {
      "epoch": 8.590808092620746,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 36726
    },
    {
      "epoch": 8.591041983393755,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 36727
    },
    {
      "epoch": 8.591275874166763,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 36728
    },
    {
      "epoch": 8.591509764939772,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5498,
      "step": 36729
    },
    {
      "epoch": 8.591743655712783,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4946,
      "step": 36730
    },
    {
      "epoch": 8.591977546485792,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8385,
      "step": 36731
    },
    {
      "epoch": 8.5922114372588,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6597,
      "step": 36732
    },
    {
      "epoch": 8.59244532803181,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 36733
    },
    {
      "epoch": 8.592679218804818,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 36734
    },
    {
      "epoch": 8.592913109577827,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0391,
      "step": 36735
    },
    {
      "epoch": 8.593147000350836,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6463,
      "step": 36736
    },
    {
      "epoch": 8.593380891123845,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.619,
      "step": 36737
    },
    {
      "epoch": 8.593614781896854,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 36738
    },
    {
      "epoch": 8.593848672669862,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.3312,
      "step": 36739
    },
    {
      "epoch": 8.594082563442871,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 36740
    },
    {
      "epoch": 8.594316454215882,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7398,
      "step": 36741
    },
    {
      "epoch": 8.59455034498889,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 36742
    },
    {
      "epoch": 8.5947842357619,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 36743
    },
    {
      "epoch": 8.595018126534908,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 36744
    },
    {
      "epoch": 8.595252017307917,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 36745
    },
    {
      "epoch": 8.595485908080926,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.495,
      "step": 36746
    },
    {
      "epoch": 8.595719798853935,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5968,
      "step": 36747
    },
    {
      "epoch": 8.595953689626944,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5901,
      "step": 36748
    },
    {
      "epoch": 8.596187580399953,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 36749
    },
    {
      "epoch": 8.596421471172961,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 36750
    },
    {
      "epoch": 8.596655361945972,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 36751
    },
    {
      "epoch": 8.59688925271898,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 36752
    },
    {
      "epoch": 8.59712314349199,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 36753
    },
    {
      "epoch": 8.597357034264999,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7025,
      "step": 36754
    },
    {
      "epoch": 8.597590925038007,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 36755
    },
    {
      "epoch": 8.597824815811016,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 36756
    },
    {
      "epoch": 8.598058706584025,
      "grad_norm": 8.125,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 36757
    },
    {
      "epoch": 8.598292597357034,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5125,
      "step": 36758
    },
    {
      "epoch": 8.598526488130043,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 36759
    },
    {
      "epoch": 8.598760378903052,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 36760
    },
    {
      "epoch": 8.598994269676062,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6339,
      "step": 36761
    },
    {
      "epoch": 8.599228160449071,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 36762
    },
    {
      "epoch": 8.59946205122208,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 36763
    },
    {
      "epoch": 8.599695941995089,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 36764
    },
    {
      "epoch": 8.599929832768098,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 36765
    },
    {
      "epoch": 8.600163723541106,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.555,
      "step": 36766
    },
    {
      "epoch": 8.600397614314115,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 36767
    },
    {
      "epoch": 8.600631505087124,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4324,
      "step": 36768
    },
    {
      "epoch": 8.600865395860133,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 36769
    },
    {
      "epoch": 8.601099286633142,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7407,
      "step": 36770
    },
    {
      "epoch": 8.60133317740615,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 36771
    },
    {
      "epoch": 8.601567068179161,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 36772
    },
    {
      "epoch": 8.60180095895217,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6833,
      "step": 36773
    },
    {
      "epoch": 8.602034849725179,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7483,
      "step": 36774
    },
    {
      "epoch": 8.602268740498188,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 36775
    },
    {
      "epoch": 8.602502631271197,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 36776
    },
    {
      "epoch": 8.602736522044205,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 36777
    },
    {
      "epoch": 8.602970412817214,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5222,
      "step": 36778
    },
    {
      "epoch": 8.603204303590223,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6249,
      "step": 36779
    },
    {
      "epoch": 8.603438194363232,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 36780
    },
    {
      "epoch": 8.60367208513624,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 36781
    },
    {
      "epoch": 8.60390597590925,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6645,
      "step": 36782
    },
    {
      "epoch": 8.60413986668226,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 36783
    },
    {
      "epoch": 8.604373757455269,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 36784
    },
    {
      "epoch": 8.604607648228278,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 36785
    },
    {
      "epoch": 8.604841539001287,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 36786
    },
    {
      "epoch": 8.605075429774296,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 36787
    },
    {
      "epoch": 8.605309320547304,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 36788
    },
    {
      "epoch": 8.605543211320313,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6505,
      "step": 36789
    },
    {
      "epoch": 8.605777102093322,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 36790
    },
    {
      "epoch": 8.60601099286633,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 36791
    },
    {
      "epoch": 8.60624488363934,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 36792
    },
    {
      "epoch": 8.606478774412349,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 36793
    },
    {
      "epoch": 8.60671266518536,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.39,
      "step": 36794
    },
    {
      "epoch": 8.606946555958368,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 36795
    },
    {
      "epoch": 8.607180446731377,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 36796
    },
    {
      "epoch": 8.607414337504386,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5313,
      "step": 36797
    },
    {
      "epoch": 8.607648228277395,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5081,
      "step": 36798
    },
    {
      "epoch": 8.607882119050403,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5531,
      "step": 36799
    },
    {
      "epoch": 8.608116009823412,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6773,
      "step": 36800
    },
    {
      "epoch": 8.608116009823412,
      "eval_runtime": 4.6071,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 36800
    },
    {
      "epoch": 8.608349900596421,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6081,
      "step": 36801
    },
    {
      "epoch": 8.60858379136943,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 36802
    },
    {
      "epoch": 8.608817682142439,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.5176,
      "step": 36803
    },
    {
      "epoch": 8.60905157291545,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 36804
    },
    {
      "epoch": 8.609285463688458,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.916,
      "step": 36805
    },
    {
      "epoch": 8.609519354461467,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 36806
    },
    {
      "epoch": 8.609753245234476,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 36807
    },
    {
      "epoch": 8.609987136007485,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 36808
    },
    {
      "epoch": 8.610221026780494,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 36809
    },
    {
      "epoch": 8.610454917553502,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 36810
    },
    {
      "epoch": 8.610688808326511,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 36811
    },
    {
      "epoch": 8.61092269909952,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 36812
    },
    {
      "epoch": 8.611156589872529,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6387,
      "step": 36813
    },
    {
      "epoch": 8.61139048064554,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 36814
    },
    {
      "epoch": 8.611624371418548,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.5944,
      "step": 36815
    },
    {
      "epoch": 8.611858262191557,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5724,
      "step": 36816
    },
    {
      "epoch": 8.612092152964566,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7305,
      "step": 36817
    },
    {
      "epoch": 8.612326043737575,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 36818
    },
    {
      "epoch": 8.612559934510584,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7268,
      "step": 36819
    },
    {
      "epoch": 8.612793825283592,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5508,
      "step": 36820
    },
    {
      "epoch": 8.613027716056601,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 36821
    },
    {
      "epoch": 8.61326160682961,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.752,
      "step": 36822
    },
    {
      "epoch": 8.613495497602619,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 36823
    },
    {
      "epoch": 8.613729388375628,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.267,
      "step": 36824
    },
    {
      "epoch": 8.613963279148638,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6148,
      "step": 36825
    },
    {
      "epoch": 8.614197169921647,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 36826
    },
    {
      "epoch": 8.614431060694656,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7281,
      "step": 36827
    },
    {
      "epoch": 8.614664951467665,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5347,
      "step": 36828
    },
    {
      "epoch": 8.614898842240674,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 36829
    },
    {
      "epoch": 8.615132733013683,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6364,
      "step": 36830
    },
    {
      "epoch": 8.615366623786691,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 36831
    },
    {
      "epoch": 8.6156005145597,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7501,
      "step": 36832
    },
    {
      "epoch": 8.61583440533271,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 36833
    },
    {
      "epoch": 8.616068296105718,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 36834
    },
    {
      "epoch": 8.616302186878727,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 36835
    },
    {
      "epoch": 8.616536077651737,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 36836
    },
    {
      "epoch": 8.616769968424746,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 36837
    },
    {
      "epoch": 8.617003859197755,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 36838
    },
    {
      "epoch": 8.617237749970764,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 36839
    },
    {
      "epoch": 8.617471640743773,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 36840
    },
    {
      "epoch": 8.617705531516782,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 36841
    },
    {
      "epoch": 8.61793942228979,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5279,
      "step": 36842
    },
    {
      "epoch": 8.6181733130628,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 36843
    },
    {
      "epoch": 8.618407203835808,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 36844
    },
    {
      "epoch": 8.618641094608817,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 36845
    },
    {
      "epoch": 8.618874985381826,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 36846
    },
    {
      "epoch": 8.619108876154836,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5827,
      "step": 36847
    },
    {
      "epoch": 8.619342766927845,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 36848
    },
    {
      "epoch": 8.619576657700854,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 36849
    },
    {
      "epoch": 8.619810548473863,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5967,
      "step": 36850
    },
    {
      "epoch": 8.620044439246872,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 36851
    },
    {
      "epoch": 8.62027833001988,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7096,
      "step": 36852
    },
    {
      "epoch": 8.62051222079289,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 36853
    },
    {
      "epoch": 8.620746111565898,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 36854
    },
    {
      "epoch": 8.620980002338907,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 36855
    },
    {
      "epoch": 8.621213893111916,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 36856
    },
    {
      "epoch": 8.621447783884925,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 36857
    },
    {
      "epoch": 8.621681674657935,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 36858
    },
    {
      "epoch": 8.621915565430944,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 36859
    },
    {
      "epoch": 8.622149456203953,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 36860
    },
    {
      "epoch": 8.622383346976962,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 36861
    },
    {
      "epoch": 8.62261723774997,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8711,
      "step": 36862
    },
    {
      "epoch": 8.62285112852298,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 36863
    },
    {
      "epoch": 8.623085019295988,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6981,
      "step": 36864
    },
    {
      "epoch": 8.623318910068997,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 36865
    },
    {
      "epoch": 8.623552800842006,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 36866
    },
    {
      "epoch": 8.623786691615015,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.4752,
      "step": 36867
    },
    {
      "epoch": 8.624020582388026,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 36868
    },
    {
      "epoch": 8.624254473161034,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7462,
      "step": 36869
    },
    {
      "epoch": 8.624488363934043,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.3205,
      "step": 36870
    },
    {
      "epoch": 8.624722254707052,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 36871
    },
    {
      "epoch": 8.624956145480061,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7025,
      "step": 36872
    },
    {
      "epoch": 8.62519003625307,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8476,
      "step": 36873
    },
    {
      "epoch": 8.625423927026079,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 36874
    },
    {
      "epoch": 8.625657817799087,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 36875
    },
    {
      "epoch": 8.625891708572096,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 36876
    },
    {
      "epoch": 8.626125599345105,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6085,
      "step": 36877
    },
    {
      "epoch": 8.626359490118116,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 36878
    },
    {
      "epoch": 8.626593380891125,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 36879
    },
    {
      "epoch": 8.626827271664133,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 36880
    },
    {
      "epoch": 8.627061162437142,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7146,
      "step": 36881
    },
    {
      "epoch": 8.627295053210151,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 36882
    },
    {
      "epoch": 8.62752894398316,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 36883
    },
    {
      "epoch": 8.627762834756169,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6106,
      "step": 36884
    },
    {
      "epoch": 8.627996725529178,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.684,
      "step": 36885
    },
    {
      "epoch": 8.628230616302186,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 36886
    },
    {
      "epoch": 8.628464507075195,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7022,
      "step": 36887
    },
    {
      "epoch": 8.628698397848204,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6344,
      "step": 36888
    },
    {
      "epoch": 8.628932288621215,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 36889
    },
    {
      "epoch": 8.629166179394224,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 36890
    },
    {
      "epoch": 8.629400070167232,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 36891
    },
    {
      "epoch": 8.629633960940241,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6236,
      "step": 36892
    },
    {
      "epoch": 8.62986785171325,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 36893
    },
    {
      "epoch": 8.630101742486259,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.46,
      "step": 36894
    },
    {
      "epoch": 8.630335633259268,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5638,
      "step": 36895
    },
    {
      "epoch": 8.630569524032277,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4426,
      "step": 36896
    },
    {
      "epoch": 8.630803414805285,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6343,
      "step": 36897
    },
    {
      "epoch": 8.631037305578294,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 36898
    },
    {
      "epoch": 8.631271196351303,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 36899
    },
    {
      "epoch": 8.631505087124314,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 36900
    },
    {
      "epoch": 8.631505087124314,
      "eval_runtime": 4.6418,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 36900
    },
    {
      "epoch": 8.631738977897323,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6714,
      "step": 36901
    },
    {
      "epoch": 8.631972868670331,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6358,
      "step": 36902
    },
    {
      "epoch": 8.63220675944334,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 36903
    },
    {
      "epoch": 8.632440650216349,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6431,
      "step": 36904
    },
    {
      "epoch": 8.632674540989358,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 36905
    },
    {
      "epoch": 8.632908431762367,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 36906
    },
    {
      "epoch": 8.633142322535376,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6872,
      "step": 36907
    },
    {
      "epoch": 8.633376213308384,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 36908
    },
    {
      "epoch": 8.633610104081393,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 36909
    },
    {
      "epoch": 8.633843994854402,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 36910
    },
    {
      "epoch": 8.634077885627413,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5271,
      "step": 36911
    },
    {
      "epoch": 8.634311776400422,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 36912
    },
    {
      "epoch": 8.63454566717343,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 36913
    },
    {
      "epoch": 8.63477955794644,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 36914
    },
    {
      "epoch": 8.635013448719448,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7191,
      "step": 36915
    },
    {
      "epoch": 8.635247339492457,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 36916
    },
    {
      "epoch": 8.635481230265466,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 36917
    },
    {
      "epoch": 8.635715121038475,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 36918
    },
    {
      "epoch": 8.635949011811483,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8676,
      "step": 36919
    },
    {
      "epoch": 8.636182902584492,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.3868,
      "step": 36920
    },
    {
      "epoch": 8.636416793357501,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 36921
    },
    {
      "epoch": 8.636650684130512,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 36922
    },
    {
      "epoch": 8.63688457490352,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6252,
      "step": 36923
    },
    {
      "epoch": 8.63711846567653,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 36924
    },
    {
      "epoch": 8.637352356449538,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 36925
    },
    {
      "epoch": 8.637586247222547,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 36926
    },
    {
      "epoch": 8.637820137995556,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 36927
    },
    {
      "epoch": 8.638054028768565,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 36928
    },
    {
      "epoch": 8.638287919541574,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 36929
    },
    {
      "epoch": 8.638521810314582,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 36930
    },
    {
      "epoch": 8.638755701087591,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4666,
      "step": 36931
    },
    {
      "epoch": 8.638989591860602,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 36932
    },
    {
      "epoch": 8.63922348263361,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 36933
    },
    {
      "epoch": 8.63945737340662,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6505,
      "step": 36934
    },
    {
      "epoch": 8.639691264179628,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5826,
      "step": 36935
    },
    {
      "epoch": 8.639925154952637,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6689,
      "step": 36936
    },
    {
      "epoch": 8.640159045725646,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 36937
    },
    {
      "epoch": 8.640392936498655,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.548,
      "step": 36938
    },
    {
      "epoch": 8.640626827271664,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 36939
    },
    {
      "epoch": 8.640860718044673,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 36940
    },
    {
      "epoch": 8.641094608817681,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5645,
      "step": 36941
    },
    {
      "epoch": 8.641328499590692,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 36942
    },
    {
      "epoch": 8.6415623903637,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 36943
    },
    {
      "epoch": 8.64179628113671,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 36944
    },
    {
      "epoch": 8.642030171909719,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.3668,
      "step": 36945
    },
    {
      "epoch": 8.642264062682727,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 36946
    },
    {
      "epoch": 8.642497953455736,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 36947
    },
    {
      "epoch": 8.642731844228745,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 36948
    },
    {
      "epoch": 8.642965735001754,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 36949
    },
    {
      "epoch": 8.643199625774763,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 36950
    },
    {
      "epoch": 8.643433516547772,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7157,
      "step": 36951
    },
    {
      "epoch": 8.64366740732078,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 36952
    },
    {
      "epoch": 8.643901298093791,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6997,
      "step": 36953
    },
    {
      "epoch": 8.6441351888668,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 36954
    },
    {
      "epoch": 8.644369079639809,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 36955
    },
    {
      "epoch": 8.644602970412818,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 36956
    },
    {
      "epoch": 8.644836861185826,
      "grad_norm": 7.84375,
      "learning_rate": 3e-05,
      "loss": 2.2737,
      "step": 36957
    },
    {
      "epoch": 8.645070751958835,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 36958
    },
    {
      "epoch": 8.645304642731844,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 36959
    },
    {
      "epoch": 8.645538533504853,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 36960
    },
    {
      "epoch": 8.645772424277862,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5925,
      "step": 36961
    },
    {
      "epoch": 8.64600631505087,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5453,
      "step": 36962
    },
    {
      "epoch": 8.64624020582388,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 36963
    },
    {
      "epoch": 8.64647409659689,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 36964
    },
    {
      "epoch": 8.646707987369899,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 36965
    },
    {
      "epoch": 8.646941878142908,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5768,
      "step": 36966
    },
    {
      "epoch": 8.647175768915917,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6551,
      "step": 36967
    },
    {
      "epoch": 8.647409659688925,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 36968
    },
    {
      "epoch": 8.647643550461934,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 36969
    },
    {
      "epoch": 8.647877441234943,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 36970
    },
    {
      "epoch": 8.648111332007952,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 36971
    },
    {
      "epoch": 8.64834522278096,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7463,
      "step": 36972
    },
    {
      "epoch": 8.64857911355397,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 36973
    },
    {
      "epoch": 8.648813004326978,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 36974
    },
    {
      "epoch": 8.649046895099989,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 36975
    },
    {
      "epoch": 8.649280785872998,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 36976
    },
    {
      "epoch": 8.649514676646007,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.5678,
      "step": 36977
    },
    {
      "epoch": 8.649748567419016,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 36978
    },
    {
      "epoch": 8.649982458192024,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 36979
    },
    {
      "epoch": 8.650216348965033,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.7253,
      "step": 36980
    },
    {
      "epoch": 8.650450239738042,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6872,
      "step": 36981
    },
    {
      "epoch": 8.650684130511051,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6016,
      "step": 36982
    },
    {
      "epoch": 8.65091802128406,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8468,
      "step": 36983
    },
    {
      "epoch": 8.651151912057069,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 36984
    },
    {
      "epoch": 8.65138580283008,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 36985
    },
    {
      "epoch": 8.651619693603088,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.4613,
      "step": 36986
    },
    {
      "epoch": 8.651853584376097,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 36987
    },
    {
      "epoch": 8.652087475149106,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 36988
    },
    {
      "epoch": 8.652321365922115,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5684,
      "step": 36989
    },
    {
      "epoch": 8.652555256695123,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7698,
      "step": 36990
    },
    {
      "epoch": 8.652789147468132,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 36991
    },
    {
      "epoch": 8.653023038241141,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 36992
    },
    {
      "epoch": 8.65325692901415,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.654,
      "step": 36993
    },
    {
      "epoch": 8.653490819787159,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 36994
    },
    {
      "epoch": 8.65372471056017,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 36995
    },
    {
      "epoch": 8.653958601333178,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 36996
    },
    {
      "epoch": 8.654192492106187,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 36997
    },
    {
      "epoch": 8.654426382879196,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6941,
      "step": 36998
    },
    {
      "epoch": 8.654660273652205,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 36999
    },
    {
      "epoch": 8.654894164425214,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 37000
    },
    {
      "epoch": 8.654894164425214,
      "eval_runtime": 4.6506,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 37000
    },
    {
      "epoch": 8.655128055198222,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7931,
      "step": 37001
    },
    {
      "epoch": 8.655361945971231,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6718,
      "step": 37002
    },
    {
      "epoch": 8.65559583674424,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 37003
    },
    {
      "epoch": 8.655829727517249,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 37004
    },
    {
      "epoch": 8.656063618290258,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8629,
      "step": 37005
    },
    {
      "epoch": 8.656297509063268,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 37006
    },
    {
      "epoch": 8.656531399836277,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.3775,
      "step": 37007
    },
    {
      "epoch": 8.656765290609286,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 37008
    },
    {
      "epoch": 8.656999181382295,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6278,
      "step": 37009
    },
    {
      "epoch": 8.657233072155304,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6176,
      "step": 37010
    },
    {
      "epoch": 8.657466962928313,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7851,
      "step": 37011
    },
    {
      "epoch": 8.657700853701321,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7269,
      "step": 37012
    },
    {
      "epoch": 8.65793474447433,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 37013
    },
    {
      "epoch": 8.658168635247339,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 37014
    },
    {
      "epoch": 8.658402526020348,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 37015
    },
    {
      "epoch": 8.658636416793357,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 37016
    },
    {
      "epoch": 8.658870307566367,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 37017
    },
    {
      "epoch": 8.659104198339376,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5803,
      "step": 37018
    },
    {
      "epoch": 8.659338089112385,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 37019
    },
    {
      "epoch": 8.659571979885394,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 37020
    },
    {
      "epoch": 8.659805870658403,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 37021
    },
    {
      "epoch": 8.660039761431412,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4276,
      "step": 37022
    },
    {
      "epoch": 8.66027365220442,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6973,
      "step": 37023
    },
    {
      "epoch": 8.66050754297743,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.577,
      "step": 37024
    },
    {
      "epoch": 8.660741433750438,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 37025
    },
    {
      "epoch": 8.660975324523447,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 37026
    },
    {
      "epoch": 8.661209215296456,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 37027
    },
    {
      "epoch": 8.661443106069466,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 37028
    },
    {
      "epoch": 8.661676996842475,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 37029
    },
    {
      "epoch": 8.661910887615484,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 37030
    },
    {
      "epoch": 8.662144778388493,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7398,
      "step": 37031
    },
    {
      "epoch": 8.662378669161502,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 37032
    },
    {
      "epoch": 8.66261255993451,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5674,
      "step": 37033
    },
    {
      "epoch": 8.66284645070752,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.496,
      "step": 37034
    },
    {
      "epoch": 8.663080341480528,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7214,
      "step": 37035
    },
    {
      "epoch": 8.663314232253537,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 37036
    },
    {
      "epoch": 8.663548123026546,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 37037
    },
    {
      "epoch": 8.663782013799555,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5128,
      "step": 37038
    },
    {
      "epoch": 8.664015904572565,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 37039
    },
    {
      "epoch": 8.664249795345574,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 37040
    },
    {
      "epoch": 8.664483686118583,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7436,
      "step": 37041
    },
    {
      "epoch": 8.664717576891592,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 37042
    },
    {
      "epoch": 8.6649514676646,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 37043
    },
    {
      "epoch": 8.66518535843761,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 37044
    },
    {
      "epoch": 8.665419249210618,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6304,
      "step": 37045
    },
    {
      "epoch": 8.665653139983627,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 37046
    },
    {
      "epoch": 8.665887030756636,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 37047
    },
    {
      "epoch": 8.666120921529645,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 37048
    },
    {
      "epoch": 8.666354812302655,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7087,
      "step": 37049
    },
    {
      "epoch": 8.666588703075664,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 37050
    },
    {
      "epoch": 8.666822593848673,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6566,
      "step": 37051
    },
    {
      "epoch": 8.667056484621682,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.6998,
      "step": 37052
    },
    {
      "epoch": 8.66729037539469,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 37053
    },
    {
      "epoch": 8.6675242661677,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 37054
    },
    {
      "epoch": 8.667758156940709,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 37055
    },
    {
      "epoch": 8.667992047713717,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 37056
    },
    {
      "epoch": 8.668225938486726,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 37057
    },
    {
      "epoch": 8.668459829259735,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.514,
      "step": 37058
    },
    {
      "epoch": 8.668693720032746,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 37059
    },
    {
      "epoch": 8.668927610805754,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 37060
    },
    {
      "epoch": 8.669161501578763,
      "grad_norm": 9.375,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 37061
    },
    {
      "epoch": 8.669395392351772,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7582,
      "step": 37062
    },
    {
      "epoch": 8.669629283124781,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.617,
      "step": 37063
    },
    {
      "epoch": 8.66986317389779,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 37064
    },
    {
      "epoch": 8.670097064670799,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 2.2994,
      "step": 37065
    },
    {
      "epoch": 8.670330955443807,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6714,
      "step": 37066
    },
    {
      "epoch": 8.670564846216816,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 37067
    },
    {
      "epoch": 8.670798736989825,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 37068
    },
    {
      "epoch": 8.671032627762834,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 37069
    },
    {
      "epoch": 8.671266518535845,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6151,
      "step": 37070
    },
    {
      "epoch": 8.671500409308853,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7204,
      "step": 37071
    },
    {
      "epoch": 8.671734300081862,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7657,
      "step": 37072
    },
    {
      "epoch": 8.671968190854871,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 37073
    },
    {
      "epoch": 8.67220208162788,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 37074
    },
    {
      "epoch": 8.672435972400889,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6052,
      "step": 37075
    },
    {
      "epoch": 8.672669863173898,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.789,
      "step": 37076
    },
    {
      "epoch": 8.672903753946906,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8815,
      "step": 37077
    },
    {
      "epoch": 8.673137644719915,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7328,
      "step": 37078
    },
    {
      "epoch": 8.673371535492924,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.4603,
      "step": 37079
    },
    {
      "epoch": 8.673605426265933,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 37080
    },
    {
      "epoch": 8.673839317038944,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 37081
    },
    {
      "epoch": 8.674073207811952,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 37082
    },
    {
      "epoch": 8.674307098584961,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 37083
    },
    {
      "epoch": 8.67454098935797,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 37084
    },
    {
      "epoch": 8.674774880130979,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 37085
    },
    {
      "epoch": 8.675008770903988,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 37086
    },
    {
      "epoch": 8.675242661676997,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 37087
    },
    {
      "epoch": 8.675476552450005,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5702,
      "step": 37088
    },
    {
      "epoch": 8.675710443223014,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2844,
      "step": 37089
    },
    {
      "epoch": 8.675944333996023,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7115,
      "step": 37090
    },
    {
      "epoch": 8.676178224769032,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 37091
    },
    {
      "epoch": 8.676412115542043,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.5967,
      "step": 37092
    },
    {
      "epoch": 8.676646006315051,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6174,
      "step": 37093
    },
    {
      "epoch": 8.67687989708806,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5281,
      "step": 37094
    },
    {
      "epoch": 8.67711378786107,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 37095
    },
    {
      "epoch": 8.677347678634078,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7094,
      "step": 37096
    },
    {
      "epoch": 8.677581569407087,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 37097
    },
    {
      "epoch": 8.677815460180096,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 37098
    },
    {
      "epoch": 8.678049350953104,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6738,
      "step": 37099
    },
    {
      "epoch": 8.678283241726113,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 37100
    },
    {
      "epoch": 8.678283241726113,
      "eval_runtime": 4.6477,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 37100
    },
    {
      "epoch": 8.678517132499122,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 37101
    },
    {
      "epoch": 8.678751023272131,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7182,
      "step": 37102
    },
    {
      "epoch": 8.678984914045142,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 37103
    },
    {
      "epoch": 8.67921880481815,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5873,
      "step": 37104
    },
    {
      "epoch": 8.67945269559116,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6731,
      "step": 37105
    },
    {
      "epoch": 8.679686586364168,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6088,
      "step": 37106
    },
    {
      "epoch": 8.679920477137177,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 37107
    },
    {
      "epoch": 8.680154367910186,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 37108
    },
    {
      "epoch": 8.680388258683195,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7998,
      "step": 37109
    },
    {
      "epoch": 8.680622149456203,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 37110
    },
    {
      "epoch": 8.680856040229212,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6747,
      "step": 37111
    },
    {
      "epoch": 8.681089931002221,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.3347,
      "step": 37112
    },
    {
      "epoch": 8.681323821775232,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 37113
    },
    {
      "epoch": 8.68155771254824,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 37114
    },
    {
      "epoch": 8.68179160332125,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 37115
    },
    {
      "epoch": 8.682025494094258,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.6725,
      "step": 37116
    },
    {
      "epoch": 8.682259384867267,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6174,
      "step": 37117
    },
    {
      "epoch": 8.682493275640276,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 37118
    },
    {
      "epoch": 8.682727166413285,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6066,
      "step": 37119
    },
    {
      "epoch": 8.682961057186294,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7609,
      "step": 37120
    },
    {
      "epoch": 8.683194947959302,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 37121
    },
    {
      "epoch": 8.683428838732311,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 37122
    },
    {
      "epoch": 8.683662729505322,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 37123
    },
    {
      "epoch": 8.68389662027833,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 37124
    },
    {
      "epoch": 8.68413051105134,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 37125
    },
    {
      "epoch": 8.684364401824348,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.4658,
      "step": 37126
    },
    {
      "epoch": 8.684598292597357,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 37127
    },
    {
      "epoch": 8.684832183370366,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 37128
    },
    {
      "epoch": 8.685066074143375,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 37129
    },
    {
      "epoch": 8.685299964916384,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 37130
    },
    {
      "epoch": 8.685533855689393,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6492,
      "step": 37131
    },
    {
      "epoch": 8.685767746462401,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4854,
      "step": 37132
    },
    {
      "epoch": 8.68600163723541,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 37133
    },
    {
      "epoch": 8.686235528008421,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8223,
      "step": 37134
    },
    {
      "epoch": 8.68646941878143,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6257,
      "step": 37135
    },
    {
      "epoch": 8.686703309554439,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 37136
    },
    {
      "epoch": 8.686937200327447,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6525,
      "step": 37137
    },
    {
      "epoch": 8.687171091100456,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 37138
    },
    {
      "epoch": 8.687404981873465,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 37139
    },
    {
      "epoch": 8.687638872646474,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5983,
      "step": 37140
    },
    {
      "epoch": 8.687872763419483,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 37141
    },
    {
      "epoch": 8.688106654192492,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 37142
    },
    {
      "epoch": 8.6883405449655,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 37143
    },
    {
      "epoch": 8.68857443573851,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6677,
      "step": 37144
    },
    {
      "epoch": 8.68880832651152,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 37145
    },
    {
      "epoch": 8.689042217284529,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 37146
    },
    {
      "epoch": 8.689276108057538,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.768,
      "step": 37147
    },
    {
      "epoch": 8.689509998830546,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 37148
    },
    {
      "epoch": 8.689743889603555,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7127,
      "step": 37149
    },
    {
      "epoch": 8.689977780376564,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6943,
      "step": 37150
    },
    {
      "epoch": 8.690211671149573,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.4771,
      "step": 37151
    },
    {
      "epoch": 8.690445561922582,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 37152
    },
    {
      "epoch": 8.69067945269559,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 1.5482,
      "step": 37153
    },
    {
      "epoch": 8.6909133434686,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7045,
      "step": 37154
    },
    {
      "epoch": 8.691147234241608,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5646,
      "step": 37155
    },
    {
      "epoch": 8.691381125014619,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 37156
    },
    {
      "epoch": 8.691615015787628,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.5591,
      "step": 37157
    },
    {
      "epoch": 8.691848906560637,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 37158
    },
    {
      "epoch": 8.692082797333645,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5324,
      "step": 37159
    },
    {
      "epoch": 8.692316688106654,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 37160
    },
    {
      "epoch": 8.692550578879663,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 37161
    },
    {
      "epoch": 8.692784469652672,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 37162
    },
    {
      "epoch": 8.69301836042568,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 37163
    },
    {
      "epoch": 8.69325225119869,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 37164
    },
    {
      "epoch": 8.693486141971698,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 37165
    },
    {
      "epoch": 8.693720032744707,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5311,
      "step": 37166
    },
    {
      "epoch": 8.693953923517718,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 37167
    },
    {
      "epoch": 8.694187814290727,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 37168
    },
    {
      "epoch": 8.694421705063736,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.5984,
      "step": 37169
    },
    {
      "epoch": 8.694655595836744,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 37170
    },
    {
      "epoch": 8.694889486609753,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 37171
    },
    {
      "epoch": 8.695123377382762,
      "grad_norm": 9.4375,
      "learning_rate": 3e-05,
      "loss": 1.741,
      "step": 37172
    },
    {
      "epoch": 8.695357268155771,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 37173
    },
    {
      "epoch": 8.69559115892878,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 37174
    },
    {
      "epoch": 8.695825049701789,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6973,
      "step": 37175
    },
    {
      "epoch": 8.696058940474797,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 37176
    },
    {
      "epoch": 8.696292831247808,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4255,
      "step": 37177
    },
    {
      "epoch": 8.696526722020817,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 37178
    },
    {
      "epoch": 8.696760612793826,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 37179
    },
    {
      "epoch": 8.696994503566835,
      "grad_norm": 7.375,
      "learning_rate": 3e-05,
      "loss": 1.6691,
      "step": 37180
    },
    {
      "epoch": 8.697228394339843,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 37181
    },
    {
      "epoch": 8.697462285112852,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 37182
    },
    {
      "epoch": 8.697696175885861,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 37183
    },
    {
      "epoch": 8.69793006665887,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6909,
      "step": 37184
    },
    {
      "epoch": 8.698163957431879,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 37185
    },
    {
      "epoch": 8.698397848204888,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.4204,
      "step": 37186
    },
    {
      "epoch": 8.698631738977898,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5944,
      "step": 37187
    },
    {
      "epoch": 8.698865629750907,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6416,
      "step": 37188
    },
    {
      "epoch": 8.699099520523916,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 37189
    },
    {
      "epoch": 8.699333411296925,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 37190
    },
    {
      "epoch": 8.699567302069934,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6824,
      "step": 37191
    },
    {
      "epoch": 8.699801192842942,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 37192
    },
    {
      "epoch": 8.700035083615951,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 37193
    },
    {
      "epoch": 8.70026897438896,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 37194
    },
    {
      "epoch": 8.700502865161969,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.66,
      "step": 37195
    },
    {
      "epoch": 8.700736755934978,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 37196
    },
    {
      "epoch": 8.700970646707987,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 37197
    },
    {
      "epoch": 8.701204537480997,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 37198
    },
    {
      "epoch": 8.701438428254006,
      "grad_norm": 14.75,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 37199
    },
    {
      "epoch": 8.701672319027015,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 37200
    },
    {
      "epoch": 8.701672319027015,
      "eval_runtime": 4.6254,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 37200
    },
    {
      "epoch": 8.701906209800024,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 37201
    },
    {
      "epoch": 8.702140100573033,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6613,
      "step": 37202
    },
    {
      "epoch": 8.702373991346041,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 37203
    },
    {
      "epoch": 8.70260788211905,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 37204
    },
    {
      "epoch": 8.702841772892059,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 37205
    },
    {
      "epoch": 8.703075663665068,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 37206
    },
    {
      "epoch": 8.703309554438077,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 37207
    },
    {
      "epoch": 8.703543445211086,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4751,
      "step": 37208
    },
    {
      "epoch": 8.703777335984096,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 37209
    },
    {
      "epoch": 8.704011226757105,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 37210
    },
    {
      "epoch": 8.704245117530114,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 37211
    },
    {
      "epoch": 8.704479008303123,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4868,
      "step": 37212
    },
    {
      "epoch": 8.704712899076132,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.5602,
      "step": 37213
    },
    {
      "epoch": 8.70494678984914,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 37214
    },
    {
      "epoch": 8.70518068062215,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.661,
      "step": 37215
    },
    {
      "epoch": 8.705414571395158,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8002,
      "step": 37216
    },
    {
      "epoch": 8.705648462168167,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5183,
      "step": 37217
    },
    {
      "epoch": 8.705882352941176,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 37218
    },
    {
      "epoch": 8.706116243714185,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6276,
      "step": 37219
    },
    {
      "epoch": 8.706350134487195,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 37220
    },
    {
      "epoch": 8.706584025260204,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 37221
    },
    {
      "epoch": 8.706817916033213,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 37222
    },
    {
      "epoch": 8.707051806806222,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 37223
    },
    {
      "epoch": 8.70728569757923,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.673,
      "step": 37224
    },
    {
      "epoch": 8.70751958835224,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.653,
      "step": 37225
    },
    {
      "epoch": 8.707753479125248,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 37226
    },
    {
      "epoch": 8.707987369898257,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 37227
    },
    {
      "epoch": 8.708221260671266,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 37228
    },
    {
      "epoch": 8.708455151444275,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7419,
      "step": 37229
    },
    {
      "epoch": 8.708689042217285,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5748,
      "step": 37230
    },
    {
      "epoch": 8.708922932990294,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6361,
      "step": 37231
    },
    {
      "epoch": 8.709156823763303,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.6682,
      "step": 37232
    },
    {
      "epoch": 8.709390714536312,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 37233
    },
    {
      "epoch": 8.70962460530932,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7818,
      "step": 37234
    },
    {
      "epoch": 8.70985849608233,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 37235
    },
    {
      "epoch": 8.710092386855338,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 37236
    },
    {
      "epoch": 8.710326277628347,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 37237
    },
    {
      "epoch": 8.710560168401356,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 37238
    },
    {
      "epoch": 8.710794059174365,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 37239
    },
    {
      "epoch": 8.711027949947376,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8677,
      "step": 37240
    },
    {
      "epoch": 8.711261840720384,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5363,
      "step": 37241
    },
    {
      "epoch": 8.711495731493393,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 37242
    },
    {
      "epoch": 8.711729622266402,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.705,
      "step": 37243
    },
    {
      "epoch": 8.71196351303941,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6987,
      "step": 37244
    },
    {
      "epoch": 8.71219740381242,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 37245
    },
    {
      "epoch": 8.712431294585429,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 37246
    },
    {
      "epoch": 8.712665185358437,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 37247
    },
    {
      "epoch": 8.712899076131446,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 37248
    },
    {
      "epoch": 8.713132966904455,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 37249
    },
    {
      "epoch": 8.713366857677464,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 37250
    },
    {
      "epoch": 8.713600748450474,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 37251
    },
    {
      "epoch": 8.713834639223483,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 37252
    },
    {
      "epoch": 8.714068529996492,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 37253
    },
    {
      "epoch": 8.714302420769501,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5448,
      "step": 37254
    },
    {
      "epoch": 8.71453631154251,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 37255
    },
    {
      "epoch": 8.714770202315519,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 37256
    },
    {
      "epoch": 8.715004093088528,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 37257
    },
    {
      "epoch": 8.715237983861536,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 37258
    },
    {
      "epoch": 8.715471874634545,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 37259
    },
    {
      "epoch": 8.715705765407554,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 37260
    },
    {
      "epoch": 8.715939656180563,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.6054,
      "step": 37261
    },
    {
      "epoch": 8.716173546953573,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.348,
      "step": 37262
    },
    {
      "epoch": 8.716407437726582,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 37263
    },
    {
      "epoch": 8.716641328499591,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7242,
      "step": 37264
    },
    {
      "epoch": 8.7168752192726,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 37265
    },
    {
      "epoch": 8.717109110045609,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.61,
      "step": 37266
    },
    {
      "epoch": 8.717343000818618,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6985,
      "step": 37267
    },
    {
      "epoch": 8.717576891591627,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8251,
      "step": 37268
    },
    {
      "epoch": 8.717810782364635,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6885,
      "step": 37269
    },
    {
      "epoch": 8.718044673137644,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 37270
    },
    {
      "epoch": 8.718278563910653,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5993,
      "step": 37271
    },
    {
      "epoch": 8.718512454683662,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 37272
    },
    {
      "epoch": 8.718746345456672,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 37273
    },
    {
      "epoch": 8.718980236229681,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 37274
    },
    {
      "epoch": 8.71921412700269,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 37275
    },
    {
      "epoch": 8.719448017775699,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 37276
    },
    {
      "epoch": 8.719681908548708,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 37277
    },
    {
      "epoch": 8.719915799321717,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.749,
      "step": 37278
    },
    {
      "epoch": 8.720149690094726,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6248,
      "step": 37279
    },
    {
      "epoch": 8.720383580867734,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 37280
    },
    {
      "epoch": 8.720617471640743,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 37281
    },
    {
      "epoch": 8.720851362413752,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7155,
      "step": 37282
    },
    {
      "epoch": 8.72108525318676,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4891,
      "step": 37283
    },
    {
      "epoch": 8.721319143959771,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6531,
      "step": 37284
    },
    {
      "epoch": 8.72155303473278,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 37285
    },
    {
      "epoch": 8.72178692550579,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 37286
    },
    {
      "epoch": 8.722020816278798,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6084,
      "step": 37287
    },
    {
      "epoch": 8.722254707051807,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 37288
    },
    {
      "epoch": 8.722488597824816,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 37289
    },
    {
      "epoch": 8.722722488597825,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 37290
    },
    {
      "epoch": 8.722956379370833,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 37291
    },
    {
      "epoch": 8.723190270143842,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 37292
    },
    {
      "epoch": 8.723424160916851,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 37293
    },
    {
      "epoch": 8.723658051689862,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 37294
    },
    {
      "epoch": 8.72389194246287,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 37295
    },
    {
      "epoch": 8.72412583323588,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.661,
      "step": 37296
    },
    {
      "epoch": 8.724359724008888,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 37297
    },
    {
      "epoch": 8.724593614781897,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6215,
      "step": 37298
    },
    {
      "epoch": 8.724827505554906,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 37299
    },
    {
      "epoch": 8.725061396327915,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 37300
    },
    {
      "epoch": 8.725061396327915,
      "eval_runtime": 4.6659,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 37300
    },
    {
      "epoch": 8.725295287100924,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5102,
      "step": 37301
    },
    {
      "epoch": 8.725529177873932,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 37302
    },
    {
      "epoch": 8.725763068646941,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 37303
    },
    {
      "epoch": 8.725996959419952,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.6197,
      "step": 37304
    },
    {
      "epoch": 8.72623085019296,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.5384,
      "step": 37305
    },
    {
      "epoch": 8.72646474096597,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8601,
      "step": 37306
    },
    {
      "epoch": 8.726698631738978,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 37307
    },
    {
      "epoch": 8.726932522511987,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.642,
      "step": 37308
    },
    {
      "epoch": 8.727166413284996,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 37309
    },
    {
      "epoch": 8.727400304058005,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 37310
    },
    {
      "epoch": 8.727634194831014,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 37311
    },
    {
      "epoch": 8.727868085604022,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 37312
    },
    {
      "epoch": 8.728101976377031,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.754,
      "step": 37313
    },
    {
      "epoch": 8.72833586715004,
      "grad_norm": 2.515625,
      "learning_rate": 3e-05,
      "loss": 1.6602,
      "step": 37314
    },
    {
      "epoch": 8.72856975792305,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 37315
    },
    {
      "epoch": 8.72880364869606,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6436,
      "step": 37316
    },
    {
      "epoch": 8.729037539469068,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 37317
    },
    {
      "epoch": 8.729271430242077,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 37318
    },
    {
      "epoch": 8.729505321015086,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 37319
    },
    {
      "epoch": 8.729739211788095,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 37320
    },
    {
      "epoch": 8.729973102561104,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 37321
    },
    {
      "epoch": 8.730206993334113,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 37322
    },
    {
      "epoch": 8.730440884107121,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.3834,
      "step": 37323
    },
    {
      "epoch": 8.73067477488013,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 37324
    },
    {
      "epoch": 8.73090866565314,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6949,
      "step": 37325
    },
    {
      "epoch": 8.73114255642615,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 37326
    },
    {
      "epoch": 8.731376447199159,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 37327
    },
    {
      "epoch": 8.731610337972167,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 37328
    },
    {
      "epoch": 8.731844228745176,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 37329
    },
    {
      "epoch": 8.732078119518185,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 37330
    },
    {
      "epoch": 8.732312010291194,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6555,
      "step": 37331
    },
    {
      "epoch": 8.732545901064203,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 37332
    },
    {
      "epoch": 8.732779791837212,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8195,
      "step": 37333
    },
    {
      "epoch": 8.73301368261022,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.3823,
      "step": 37334
    },
    {
      "epoch": 8.73324757338323,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 37335
    },
    {
      "epoch": 8.733481464156238,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7823,
      "step": 37336
    },
    {
      "epoch": 8.733715354929249,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6216,
      "step": 37337
    },
    {
      "epoch": 8.733949245702258,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 37338
    },
    {
      "epoch": 8.734183136475266,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 37339
    },
    {
      "epoch": 8.734417027248275,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7981,
      "step": 37340
    },
    {
      "epoch": 8.734650918021284,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7728,
      "step": 37341
    },
    {
      "epoch": 8.734884808794293,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.5403,
      "step": 37342
    },
    {
      "epoch": 8.735118699567302,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5015,
      "step": 37343
    },
    {
      "epoch": 8.73535259034031,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 37344
    },
    {
      "epoch": 8.73558648111332,
      "grad_norm": 8.0,
      "learning_rate": 3e-05,
      "loss": 2.6962,
      "step": 37345
    },
    {
      "epoch": 8.735820371886328,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 37346
    },
    {
      "epoch": 8.736054262659337,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 37347
    },
    {
      "epoch": 8.736288153432348,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 37348
    },
    {
      "epoch": 8.736522044205357,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 37349
    },
    {
      "epoch": 8.736755934978365,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 37350
    },
    {
      "epoch": 8.736989825751374,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 37351
    },
    {
      "epoch": 8.737223716524383,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 37352
    },
    {
      "epoch": 8.737457607297392,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7616,
      "step": 37353
    },
    {
      "epoch": 8.7376914980704,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7628,
      "step": 37354
    },
    {
      "epoch": 8.73792538884341,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 37355
    },
    {
      "epoch": 8.738159279616418,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3273,
      "step": 37356
    },
    {
      "epoch": 8.738393170389427,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 37357
    },
    {
      "epoch": 8.738627061162438,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.5178,
      "step": 37358
    },
    {
      "epoch": 8.738860951935447,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 37359
    },
    {
      "epoch": 8.739094842708456,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6975,
      "step": 37360
    },
    {
      "epoch": 8.739328733481464,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4981,
      "step": 37361
    },
    {
      "epoch": 8.739562624254473,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 37362
    },
    {
      "epoch": 8.739796515027482,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 37363
    },
    {
      "epoch": 8.740030405800491,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6247,
      "step": 37364
    },
    {
      "epoch": 8.7402642965735,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.606,
      "step": 37365
    },
    {
      "epoch": 8.740498187346509,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 37366
    },
    {
      "epoch": 8.740732078119517,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 37367
    },
    {
      "epoch": 8.740965968892528,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5266,
      "step": 37368
    },
    {
      "epoch": 8.741199859665537,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 37369
    },
    {
      "epoch": 8.741433750438546,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 37370
    },
    {
      "epoch": 8.741667641211555,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.559,
      "step": 37371
    },
    {
      "epoch": 8.741901531984563,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 37372
    },
    {
      "epoch": 8.742135422757572,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.6206,
      "step": 37373
    },
    {
      "epoch": 8.742369313530581,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 37374
    },
    {
      "epoch": 8.74260320430359,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 37375
    },
    {
      "epoch": 8.742837095076599,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5762,
      "step": 37376
    },
    {
      "epoch": 8.743070985849608,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 37377
    },
    {
      "epoch": 8.743304876622616,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 37378
    },
    {
      "epoch": 8.743538767395627,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5656,
      "step": 37379
    },
    {
      "epoch": 8.743772658168636,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 37380
    },
    {
      "epoch": 8.744006548941645,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.4339,
      "step": 37381
    },
    {
      "epoch": 8.744240439714654,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 37382
    },
    {
      "epoch": 8.744474330487662,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.4571,
      "step": 37383
    },
    {
      "epoch": 8.744708221260671,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 37384
    },
    {
      "epoch": 8.74494211203368,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 37385
    },
    {
      "epoch": 8.745176002806689,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.4878,
      "step": 37386
    },
    {
      "epoch": 8.745409893579698,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 37387
    },
    {
      "epoch": 8.745643784352707,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 37388
    },
    {
      "epoch": 8.745877675125715,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5829,
      "step": 37389
    },
    {
      "epoch": 8.746111565898726,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 37390
    },
    {
      "epoch": 8.746345456671735,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 37391
    },
    {
      "epoch": 8.746579347444744,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 37392
    },
    {
      "epoch": 8.746813238217753,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.3216,
      "step": 37393
    },
    {
      "epoch": 8.747047128990761,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8693,
      "step": 37394
    },
    {
      "epoch": 8.74728101976377,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6034,
      "step": 37395
    },
    {
      "epoch": 8.747514910536779,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 37396
    },
    {
      "epoch": 8.747748801309788,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 37397
    },
    {
      "epoch": 8.747982692082797,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 37398
    },
    {
      "epoch": 8.748216582855806,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 37399
    },
    {
      "epoch": 8.748450473628814,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 37400
    },
    {
      "epoch": 8.748450473628814,
      "eval_runtime": 4.6119,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 37400
    },
    {
      "epoch": 8.748684364401825,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 37401
    },
    {
      "epoch": 8.748918255174834,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 37402
    },
    {
      "epoch": 8.749152145947843,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 37403
    },
    {
      "epoch": 8.749386036720852,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 37404
    },
    {
      "epoch": 8.74961992749386,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 37405
    },
    {
      "epoch": 8.74985381826687,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 37406
    },
    {
      "epoch": 8.750087709039878,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 37407
    },
    {
      "epoch": 8.750321599812887,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 37408
    },
    {
      "epoch": 8.750555490585896,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 37409
    },
    {
      "epoch": 8.750789381358905,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 37410
    },
    {
      "epoch": 8.751023272131915,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 37411
    },
    {
      "epoch": 8.751257162904924,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6803,
      "step": 37412
    },
    {
      "epoch": 8.751491053677933,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 37413
    },
    {
      "epoch": 8.751724944450942,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 37414
    },
    {
      "epoch": 8.75195883522395,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 37415
    },
    {
      "epoch": 8.75219272599696,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 37416
    },
    {
      "epoch": 8.752426616769968,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9381,
      "step": 37417
    },
    {
      "epoch": 8.752660507542977,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 37418
    },
    {
      "epoch": 8.752894398315986,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 37419
    },
    {
      "epoch": 8.753128289088995,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 37420
    },
    {
      "epoch": 8.753362179862005,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 37421
    },
    {
      "epoch": 8.753596070635014,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6359,
      "step": 37422
    },
    {
      "epoch": 8.753829961408023,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6046,
      "step": 37423
    },
    {
      "epoch": 8.754063852181032,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 37424
    },
    {
      "epoch": 8.75429774295404,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 37425
    },
    {
      "epoch": 8.75453163372705,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 37426
    },
    {
      "epoch": 8.754765524500058,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.6854,
      "step": 37427
    },
    {
      "epoch": 8.754999415273067,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5984,
      "step": 37428
    },
    {
      "epoch": 8.755233306046076,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7238,
      "step": 37429
    },
    {
      "epoch": 8.755467196819085,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7077,
      "step": 37430
    },
    {
      "epoch": 8.755701087592094,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 37431
    },
    {
      "epoch": 8.755934978365104,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 37432
    },
    {
      "epoch": 8.756168869138113,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 37433
    },
    {
      "epoch": 8.756402759911122,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 37434
    },
    {
      "epoch": 8.75663665068413,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 37435
    },
    {
      "epoch": 8.75687054145714,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7604,
      "step": 37436
    },
    {
      "epoch": 8.757104432230149,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 37437
    },
    {
      "epoch": 8.757338323003157,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 37438
    },
    {
      "epoch": 8.757572213776166,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 37439
    },
    {
      "epoch": 8.757806104549175,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6446,
      "step": 37440
    },
    {
      "epoch": 8.758039995322184,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 37441
    },
    {
      "epoch": 8.758273886095193,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 37442
    },
    {
      "epoch": 8.758507776868203,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6604,
      "step": 37443
    },
    {
      "epoch": 8.758741667641212,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 37444
    },
    {
      "epoch": 8.758975558414221,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6971,
      "step": 37445
    },
    {
      "epoch": 8.75920944918723,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 37446
    },
    {
      "epoch": 8.759443339960239,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7239,
      "step": 37447
    },
    {
      "epoch": 8.759677230733248,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 37448
    },
    {
      "epoch": 8.759911121506256,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 37449
    },
    {
      "epoch": 8.760145012279265,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1115,
      "step": 37450
    },
    {
      "epoch": 8.760378903052274,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 37451
    },
    {
      "epoch": 8.760612793825283,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 37452
    },
    {
      "epoch": 8.760846684598292,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 37453
    },
    {
      "epoch": 8.761080575371302,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7558,
      "step": 37454
    },
    {
      "epoch": 8.761314466144311,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 37455
    },
    {
      "epoch": 8.76154835691732,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.819,
      "step": 37456
    },
    {
      "epoch": 8.761782247690329,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 37457
    },
    {
      "epoch": 8.762016138463338,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8603,
      "step": 37458
    },
    {
      "epoch": 8.762250029236347,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.625,
      "step": 37459
    },
    {
      "epoch": 8.762483920009355,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.696,
      "step": 37460
    },
    {
      "epoch": 8.762717810782364,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.784,
      "step": 37461
    },
    {
      "epoch": 8.762951701555373,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 37462
    },
    {
      "epoch": 8.763185592328382,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 37463
    },
    {
      "epoch": 8.76341948310139,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5636,
      "step": 37464
    },
    {
      "epoch": 8.763653373874401,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 37465
    },
    {
      "epoch": 8.76388726464741,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6374,
      "step": 37466
    },
    {
      "epoch": 8.764121155420419,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 37467
    },
    {
      "epoch": 8.764355046193428,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 37468
    },
    {
      "epoch": 8.764588936966437,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 37469
    },
    {
      "epoch": 8.764822827739446,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 37470
    },
    {
      "epoch": 8.765056718512454,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8684,
      "step": 37471
    },
    {
      "epoch": 8.765290609285463,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7328,
      "step": 37472
    },
    {
      "epoch": 8.765524500058472,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 37473
    },
    {
      "epoch": 8.765758390831481,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8064,
      "step": 37474
    },
    {
      "epoch": 8.765992281604492,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 37475
    },
    {
      "epoch": 8.7662261723775,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 37476
    },
    {
      "epoch": 8.76646006315051,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 37477
    },
    {
      "epoch": 8.766693953923518,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 37478
    },
    {
      "epoch": 8.766927844696527,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5817,
      "step": 37479
    },
    {
      "epoch": 8.767161735469536,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 37480
    },
    {
      "epoch": 8.767395626242545,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 37481
    },
    {
      "epoch": 8.767629517015553,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 37482
    },
    {
      "epoch": 8.767863407788562,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5915,
      "step": 37483
    },
    {
      "epoch": 8.768097298561571,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 37484
    },
    {
      "epoch": 8.768331189334582,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.813,
      "step": 37485
    },
    {
      "epoch": 8.76856508010759,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 37486
    },
    {
      "epoch": 8.7687989708806,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 37487
    },
    {
      "epoch": 8.769032861653608,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 37488
    },
    {
      "epoch": 8.769266752426617,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 37489
    },
    {
      "epoch": 8.769500643199626,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6951,
      "step": 37490
    },
    {
      "epoch": 8.769734533972635,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 37491
    },
    {
      "epoch": 8.769968424745644,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 37492
    },
    {
      "epoch": 8.770202315518652,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.6492,
      "step": 37493
    },
    {
      "epoch": 8.770436206291661,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6147,
      "step": 37494
    },
    {
      "epoch": 8.77067009706467,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.3435,
      "step": 37495
    },
    {
      "epoch": 8.77090398783768,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.813,
      "step": 37496
    },
    {
      "epoch": 8.77113787861069,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.666,
      "step": 37497
    },
    {
      "epoch": 8.771371769383698,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 37498
    },
    {
      "epoch": 8.771605660156707,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 37499
    },
    {
      "epoch": 8.771839550929716,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 37500
    },
    {
      "epoch": 8.771839550929716,
      "eval_runtime": 4.6248,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 37500
    },
    {
      "epoch": 8.772073441702725,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.619,
      "step": 37501
    },
    {
      "epoch": 8.772307332475734,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 37502
    },
    {
      "epoch": 8.772541223248743,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7307,
      "step": 37503
    },
    {
      "epoch": 8.772775114021751,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 37504
    },
    {
      "epoch": 8.77300900479476,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6789,
      "step": 37505
    },
    {
      "epoch": 8.773242895567769,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4657,
      "step": 37506
    },
    {
      "epoch": 8.77347678634078,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6245,
      "step": 37507
    },
    {
      "epoch": 8.773710677113788,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.4162,
      "step": 37508
    },
    {
      "epoch": 8.773944567886797,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 37509
    },
    {
      "epoch": 8.774178458659806,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 37510
    },
    {
      "epoch": 8.774412349432815,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 37511
    },
    {
      "epoch": 8.774646240205824,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7819,
      "step": 37512
    },
    {
      "epoch": 8.774880130978833,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6377,
      "step": 37513
    },
    {
      "epoch": 8.775114021751842,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 37514
    },
    {
      "epoch": 8.77534791252485,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7121,
      "step": 37515
    },
    {
      "epoch": 8.77558180329786,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5193,
      "step": 37516
    },
    {
      "epoch": 8.775815694070868,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6237,
      "step": 37517
    },
    {
      "epoch": 8.776049584843879,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5711,
      "step": 37518
    },
    {
      "epoch": 8.776283475616887,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7092,
      "step": 37519
    },
    {
      "epoch": 8.776517366389896,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 37520
    },
    {
      "epoch": 8.776751257162905,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 37521
    },
    {
      "epoch": 8.776985147935914,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 37522
    },
    {
      "epoch": 8.777219038708923,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 37523
    },
    {
      "epoch": 8.777452929481932,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 37524
    },
    {
      "epoch": 8.77768682025494,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 37525
    },
    {
      "epoch": 8.77792071102795,
      "grad_norm": 7.28125,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 37526
    },
    {
      "epoch": 8.778154601800958,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8401,
      "step": 37527
    },
    {
      "epoch": 8.778388492573967,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 37528
    },
    {
      "epoch": 8.778622383346978,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 37529
    },
    {
      "epoch": 8.778856274119986,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 37530
    },
    {
      "epoch": 8.779090164892995,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 37531
    },
    {
      "epoch": 8.779324055666004,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 37532
    },
    {
      "epoch": 8.779557946439013,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6678,
      "step": 37533
    },
    {
      "epoch": 8.779791837212022,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5772,
      "step": 37534
    },
    {
      "epoch": 8.78002572798503,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7355,
      "step": 37535
    },
    {
      "epoch": 8.78025961875804,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 37536
    },
    {
      "epoch": 8.780493509531048,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.3733,
      "step": 37537
    },
    {
      "epoch": 8.780727400304057,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 37538
    },
    {
      "epoch": 8.780961291077068,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6295,
      "step": 37539
    },
    {
      "epoch": 8.781195181850077,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 37540
    },
    {
      "epoch": 8.781429072623085,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8201,
      "step": 37541
    },
    {
      "epoch": 8.781662963396094,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7596,
      "step": 37542
    },
    {
      "epoch": 8.781896854169103,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7802,
      "step": 37543
    },
    {
      "epoch": 8.782130744942112,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 37544
    },
    {
      "epoch": 8.78236463571512,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7033,
      "step": 37545
    },
    {
      "epoch": 8.78259852648813,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9096,
      "step": 37546
    },
    {
      "epoch": 8.782832417261139,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 37547
    },
    {
      "epoch": 8.783066308034147,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 37548
    },
    {
      "epoch": 8.783300198807158,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 37549
    },
    {
      "epoch": 8.783534089580167,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 37550
    },
    {
      "epoch": 8.783767980353176,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6002,
      "step": 37551
    },
    {
      "epoch": 8.784001871126184,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 37552
    },
    {
      "epoch": 8.784235761899193,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 37553
    },
    {
      "epoch": 8.784469652672202,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 37554
    },
    {
      "epoch": 8.784703543445211,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 37555
    },
    {
      "epoch": 8.78493743421822,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5048,
      "step": 37556
    },
    {
      "epoch": 8.785171324991229,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 37557
    },
    {
      "epoch": 8.785405215764237,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 37558
    },
    {
      "epoch": 8.785639106537246,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 37559
    },
    {
      "epoch": 8.785872997310257,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7511,
      "step": 37560
    },
    {
      "epoch": 8.786106888083266,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7427,
      "step": 37561
    },
    {
      "epoch": 8.786340778856275,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 37562
    },
    {
      "epoch": 8.786574669629283,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4951,
      "step": 37563
    },
    {
      "epoch": 8.786808560402292,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 37564
    },
    {
      "epoch": 8.787042451175301,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7123,
      "step": 37565
    },
    {
      "epoch": 8.78727634194831,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6614,
      "step": 37566
    },
    {
      "epoch": 8.787510232721319,
      "grad_norm": 7.46875,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 37567
    },
    {
      "epoch": 8.787744123494328,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7067,
      "step": 37568
    },
    {
      "epoch": 8.787978014267336,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 37569
    },
    {
      "epoch": 8.788211905040345,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5803,
      "step": 37570
    },
    {
      "epoch": 8.788445795813356,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 37571
    },
    {
      "epoch": 8.788679686586365,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 37572
    },
    {
      "epoch": 8.788913577359374,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 37573
    },
    {
      "epoch": 8.789147468132382,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7725,
      "step": 37574
    },
    {
      "epoch": 8.789381358905391,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 37575
    },
    {
      "epoch": 8.7896152496784,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 37576
    },
    {
      "epoch": 8.789849140451409,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 37577
    },
    {
      "epoch": 8.790083031224418,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.7063,
      "step": 37578
    },
    {
      "epoch": 8.790316921997427,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6333,
      "step": 37579
    },
    {
      "epoch": 8.790550812770435,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5397,
      "step": 37580
    },
    {
      "epoch": 8.790784703543444,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6158,
      "step": 37581
    },
    {
      "epoch": 8.791018594316455,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 37582
    },
    {
      "epoch": 8.791252485089464,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 37583
    },
    {
      "epoch": 8.791486375862473,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 37584
    },
    {
      "epoch": 8.791720266635481,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 37585
    },
    {
      "epoch": 8.79195415740849,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 37586
    },
    {
      "epoch": 8.7921880481815,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5323,
      "step": 37587
    },
    {
      "epoch": 8.792421938954508,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 37588
    },
    {
      "epoch": 8.792655829727517,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5424,
      "step": 37589
    },
    {
      "epoch": 8.792889720500526,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 37590
    },
    {
      "epoch": 8.793123611273534,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6692,
      "step": 37591
    },
    {
      "epoch": 8.793357502046545,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 37592
    },
    {
      "epoch": 8.793591392819554,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 37593
    },
    {
      "epoch": 8.793825283592563,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 37594
    },
    {
      "epoch": 8.794059174365572,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 37595
    },
    {
      "epoch": 8.79429306513858,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6622,
      "step": 37596
    },
    {
      "epoch": 8.79452695591159,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8124,
      "step": 37597
    },
    {
      "epoch": 8.794760846684598,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 37598
    },
    {
      "epoch": 8.794994737457607,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 37599
    },
    {
      "epoch": 8.795228628230616,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 37600
    },
    {
      "epoch": 8.795228628230616,
      "eval_runtime": 4.5968,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 37600
    },
    {
      "epoch": 8.795462519003625,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6498,
      "step": 37601
    },
    {
      "epoch": 8.795696409776635,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 37602
    },
    {
      "epoch": 8.795930300549644,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 37603
    },
    {
      "epoch": 8.796164191322653,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4336,
      "step": 37604
    },
    {
      "epoch": 8.796398082095662,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 37605
    },
    {
      "epoch": 8.79663197286867,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.73,
      "step": 37606
    },
    {
      "epoch": 8.79686586364168,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 37607
    },
    {
      "epoch": 8.797099754414688,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 37608
    },
    {
      "epoch": 8.797333645187697,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 37609
    },
    {
      "epoch": 8.797567535960706,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7016,
      "step": 37610
    },
    {
      "epoch": 8.797801426733715,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9001,
      "step": 37611
    },
    {
      "epoch": 8.798035317506724,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.3925,
      "step": 37612
    },
    {
      "epoch": 8.798269208279734,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 37613
    },
    {
      "epoch": 8.798503099052743,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4026,
      "step": 37614
    },
    {
      "epoch": 8.798736989825752,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6959,
      "step": 37615
    },
    {
      "epoch": 8.79897088059876,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 37616
    },
    {
      "epoch": 8.79920477137177,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 37617
    },
    {
      "epoch": 8.799438662144778,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7337,
      "step": 37618
    },
    {
      "epoch": 8.799672552917787,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 37619
    },
    {
      "epoch": 8.799906443690796,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6148,
      "step": 37620
    },
    {
      "epoch": 8.800140334463805,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 37621
    },
    {
      "epoch": 8.800374225236814,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.449,
      "step": 37622
    },
    {
      "epoch": 8.800608116009823,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.276,
      "step": 37623
    },
    {
      "epoch": 8.800842006782833,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6044,
      "step": 37624
    },
    {
      "epoch": 8.801075897555842,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 37625
    },
    {
      "epoch": 8.801309788328851,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 37626
    },
    {
      "epoch": 8.80154367910186,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 37627
    },
    {
      "epoch": 8.801777569874869,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 37628
    },
    {
      "epoch": 8.802011460647877,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 37629
    },
    {
      "epoch": 8.802245351420886,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 37630
    },
    {
      "epoch": 8.802479242193895,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 37631
    },
    {
      "epoch": 8.802713132966904,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4374,
      "step": 37632
    },
    {
      "epoch": 8.802947023739913,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 37633
    },
    {
      "epoch": 8.803180914512922,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 37634
    },
    {
      "epoch": 8.803414805285932,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 37635
    },
    {
      "epoch": 8.803648696058941,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6739,
      "step": 37636
    },
    {
      "epoch": 8.80388258683195,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.696,
      "step": 37637
    },
    {
      "epoch": 8.804116477604959,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 37638
    },
    {
      "epoch": 8.804350368377968,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 37639
    },
    {
      "epoch": 8.804584259150976,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 37640
    },
    {
      "epoch": 8.804818149923985,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 37641
    },
    {
      "epoch": 8.805052040696994,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8028,
      "step": 37642
    },
    {
      "epoch": 8.805285931470003,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5256,
      "step": 37643
    },
    {
      "epoch": 8.805519822243012,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 37644
    },
    {
      "epoch": 8.80575371301602,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 37645
    },
    {
      "epoch": 8.805987603789031,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6955,
      "step": 37646
    },
    {
      "epoch": 8.80622149456204,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 37647
    },
    {
      "epoch": 8.806455385335049,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6502,
      "step": 37648
    },
    {
      "epoch": 8.806689276108058,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6326,
      "step": 37649
    },
    {
      "epoch": 8.806923166881067,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 37650
    },
    {
      "epoch": 8.807157057654075,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7851,
      "step": 37651
    },
    {
      "epoch": 8.807390948427084,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 37652
    },
    {
      "epoch": 8.807624839200093,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5115,
      "step": 37653
    },
    {
      "epoch": 8.807858729973102,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 37654
    },
    {
      "epoch": 8.80809262074611,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 37655
    },
    {
      "epoch": 8.808326511519121,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 37656
    },
    {
      "epoch": 8.80856040229213,
      "grad_norm": 7.59375,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 37657
    },
    {
      "epoch": 8.808794293065139,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 37658
    },
    {
      "epoch": 8.809028183838148,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 37659
    },
    {
      "epoch": 8.809262074611157,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 37660
    },
    {
      "epoch": 8.809495965384166,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6023,
      "step": 37661
    },
    {
      "epoch": 8.809729856157174,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 37662
    },
    {
      "epoch": 8.809963746930183,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 37663
    },
    {
      "epoch": 8.810197637703192,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6825,
      "step": 37664
    },
    {
      "epoch": 8.810431528476201,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 37665
    },
    {
      "epoch": 8.810665419249212,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.772,
      "step": 37666
    },
    {
      "epoch": 8.81089931002222,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 37667
    },
    {
      "epoch": 8.81113320079523,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 37668
    },
    {
      "epoch": 8.811367091568238,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 37669
    },
    {
      "epoch": 8.811600982341247,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 37670
    },
    {
      "epoch": 8.811834873114256,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 37671
    },
    {
      "epoch": 8.812068763887265,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 37672
    },
    {
      "epoch": 8.812302654660273,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8337,
      "step": 37673
    },
    {
      "epoch": 8.812536545433282,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 37674
    },
    {
      "epoch": 8.812770436206291,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7771,
      "step": 37675
    },
    {
      "epoch": 8.8130043269793,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 37676
    },
    {
      "epoch": 8.81323821775231,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7187,
      "step": 37677
    },
    {
      "epoch": 8.81347210852532,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6515,
      "step": 37678
    },
    {
      "epoch": 8.813705999298328,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 37679
    },
    {
      "epoch": 8.813939890071337,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 37680
    },
    {
      "epoch": 8.814173780844346,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6237,
      "step": 37681
    },
    {
      "epoch": 8.814407671617355,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 37682
    },
    {
      "epoch": 8.814641562390364,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6131,
      "step": 37683
    },
    {
      "epoch": 8.814875453163372,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 37684
    },
    {
      "epoch": 8.815109343936381,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4481,
      "step": 37685
    },
    {
      "epoch": 8.81534323470939,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 37686
    },
    {
      "epoch": 8.815577125482399,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6744,
      "step": 37687
    },
    {
      "epoch": 8.81581101625541,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 37688
    },
    {
      "epoch": 8.816044907028418,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.4732,
      "step": 37689
    },
    {
      "epoch": 8.816278797801427,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.5695,
      "step": 37690
    },
    {
      "epoch": 8.816512688574436,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 37691
    },
    {
      "epoch": 8.816746579347445,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 37692
    },
    {
      "epoch": 8.816980470120454,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 37693
    },
    {
      "epoch": 8.817214360893463,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3275,
      "step": 37694
    },
    {
      "epoch": 8.817448251666471,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5195,
      "step": 37695
    },
    {
      "epoch": 8.81768214243948,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5165,
      "step": 37696
    },
    {
      "epoch": 8.817916033212489,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8945,
      "step": 37697
    },
    {
      "epoch": 8.818149923985498,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 37698
    },
    {
      "epoch": 8.818383814758509,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 37699
    },
    {
      "epoch": 8.818617705531517,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.5474,
      "step": 37700
    },
    {
      "epoch": 8.818617705531517,
      "eval_runtime": 4.6513,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 37700
    },
    {
      "epoch": 8.818851596304526,
      "grad_norm": 10.75,
      "learning_rate": 3e-05,
      "loss": 1.6903,
      "step": 37701
    },
    {
      "epoch": 8.819085487077535,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.3155,
      "step": 37702
    },
    {
      "epoch": 8.819319377850544,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 37703
    },
    {
      "epoch": 8.819553268623553,
      "grad_norm": 2.609375,
      "learning_rate": 3e-05,
      "loss": 1.6012,
      "step": 37704
    },
    {
      "epoch": 8.819787159396562,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 37705
    },
    {
      "epoch": 8.82002105016957,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.76,
      "step": 37706
    },
    {
      "epoch": 8.82025494094258,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.6324,
      "step": 37707
    },
    {
      "epoch": 8.820488831715588,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 37708
    },
    {
      "epoch": 8.820722722488597,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 37709
    },
    {
      "epoch": 8.820956613261608,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 37710
    },
    {
      "epoch": 8.821190504034616,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6371,
      "step": 37711
    },
    {
      "epoch": 8.821424394807625,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7736,
      "step": 37712
    },
    {
      "epoch": 8.821658285580634,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 37713
    },
    {
      "epoch": 8.821892176353643,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 37714
    },
    {
      "epoch": 8.822126067126652,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 37715
    },
    {
      "epoch": 8.82235995789966,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6306,
      "step": 37716
    },
    {
      "epoch": 8.82259384867267,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 37717
    },
    {
      "epoch": 8.822827739445678,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 37718
    },
    {
      "epoch": 8.823061630218687,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 37719
    },
    {
      "epoch": 8.823295520991698,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.647,
      "step": 37720
    },
    {
      "epoch": 8.823529411764707,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 37721
    },
    {
      "epoch": 8.823763302537715,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7545,
      "step": 37722
    },
    {
      "epoch": 8.823997193310724,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 37723
    },
    {
      "epoch": 8.824231084083733,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 37724
    },
    {
      "epoch": 8.824464974856742,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.6317,
      "step": 37725
    },
    {
      "epoch": 8.82469886562975,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5857,
      "step": 37726
    },
    {
      "epoch": 8.82493275640276,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 37727
    },
    {
      "epoch": 8.825166647175768,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 37728
    },
    {
      "epoch": 8.825400537948777,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7256,
      "step": 37729
    },
    {
      "epoch": 8.825634428721788,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 37730
    },
    {
      "epoch": 8.825868319494797,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 37731
    },
    {
      "epoch": 8.826102210267806,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.4338,
      "step": 37732
    },
    {
      "epoch": 8.826336101040814,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.569,
      "step": 37733
    },
    {
      "epoch": 8.826569991813823,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 37734
    },
    {
      "epoch": 8.826803882586832,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 37735
    },
    {
      "epoch": 8.82703777335984,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.772,
      "step": 37736
    },
    {
      "epoch": 8.82727166413285,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.5938,
      "step": 37737
    },
    {
      "epoch": 8.827505554905859,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 37738
    },
    {
      "epoch": 8.827739445678867,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7586,
      "step": 37739
    },
    {
      "epoch": 8.827973336451876,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 37740
    },
    {
      "epoch": 8.828207227224887,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 37741
    },
    {
      "epoch": 8.828441117997896,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 37742
    },
    {
      "epoch": 8.828675008770904,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7318,
      "step": 37743
    },
    {
      "epoch": 8.828908899543913,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 37744
    },
    {
      "epoch": 8.829142790316922,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5175,
      "step": 37745
    },
    {
      "epoch": 8.829376681089931,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 37746
    },
    {
      "epoch": 8.82961057186294,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 37747
    },
    {
      "epoch": 8.829844462635949,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 37748
    },
    {
      "epoch": 8.830078353408958,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 37749
    },
    {
      "epoch": 8.830312244181966,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.3939,
      "step": 37750
    },
    {
      "epoch": 8.830546134954975,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.495,
      "step": 37751
    },
    {
      "epoch": 8.830780025727986,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6912,
      "step": 37752
    },
    {
      "epoch": 8.831013916500995,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5572,
      "step": 37753
    },
    {
      "epoch": 8.831247807274003,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5638,
      "step": 37754
    },
    {
      "epoch": 8.831481698047012,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6184,
      "step": 37755
    },
    {
      "epoch": 8.831715588820021,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6784,
      "step": 37756
    },
    {
      "epoch": 8.83194947959303,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 37757
    },
    {
      "epoch": 8.832183370366039,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8517,
      "step": 37758
    },
    {
      "epoch": 8.832417261139048,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 37759
    },
    {
      "epoch": 8.832651151912057,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 37760
    },
    {
      "epoch": 8.832885042685065,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 37761
    },
    {
      "epoch": 8.833118933458074,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 37762
    },
    {
      "epoch": 8.833352824231085,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 37763
    },
    {
      "epoch": 8.833586715004094,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6686,
      "step": 37764
    },
    {
      "epoch": 8.833820605777102,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 37765
    },
    {
      "epoch": 8.834054496550111,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6234,
      "step": 37766
    },
    {
      "epoch": 8.83428838732312,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 37767
    },
    {
      "epoch": 8.834522278096129,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7254,
      "step": 37768
    },
    {
      "epoch": 8.834756168869138,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5734,
      "step": 37769
    },
    {
      "epoch": 8.834990059642147,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 37770
    },
    {
      "epoch": 8.835223950415156,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6491,
      "step": 37771
    },
    {
      "epoch": 8.835457841188164,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 37772
    },
    {
      "epoch": 8.835691731961173,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4844,
      "step": 37773
    },
    {
      "epoch": 8.835925622734184,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4568,
      "step": 37774
    },
    {
      "epoch": 8.836159513507193,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 37775
    },
    {
      "epoch": 8.836393404280201,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 37776
    },
    {
      "epoch": 8.83662729505321,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 37777
    },
    {
      "epoch": 8.83686118582622,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5535,
      "step": 37778
    },
    {
      "epoch": 8.837095076599228,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7527,
      "step": 37779
    },
    {
      "epoch": 8.837328967372237,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7119,
      "step": 37780
    },
    {
      "epoch": 8.837562858145246,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5724,
      "step": 37781
    },
    {
      "epoch": 8.837796748918255,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 37782
    },
    {
      "epoch": 8.838030639691265,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7335,
      "step": 37783
    },
    {
      "epoch": 8.838264530464274,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 37784
    },
    {
      "epoch": 8.838498421237283,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 37785
    },
    {
      "epoch": 8.838732312010292,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 37786
    },
    {
      "epoch": 8.8389662027833,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 37787
    },
    {
      "epoch": 8.83920009355631,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 37788
    },
    {
      "epoch": 8.839433984329318,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 37789
    },
    {
      "epoch": 8.839667875102327,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7981,
      "step": 37790
    },
    {
      "epoch": 8.839901765875336,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7663,
      "step": 37791
    },
    {
      "epoch": 8.840135656648345,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.703,
      "step": 37792
    },
    {
      "epoch": 8.840369547421354,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 37793
    },
    {
      "epoch": 8.840603438194364,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.2546,
      "step": 37794
    },
    {
      "epoch": 8.840837328967373,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7281,
      "step": 37795
    },
    {
      "epoch": 8.841071219740382,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6736,
      "step": 37796
    },
    {
      "epoch": 8.84130511051339,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 37797
    },
    {
      "epoch": 8.8415390012864,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.62,
      "step": 37798
    },
    {
      "epoch": 8.841772892059408,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 37799
    },
    {
      "epoch": 8.842006782832417,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 37800
    },
    {
      "epoch": 8.842006782832417,
      "eval_runtime": 4.654,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 37800
    },
    {
      "epoch": 8.842240673605426,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6743,
      "step": 37801
    },
    {
      "epoch": 8.842474564378435,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 37802
    },
    {
      "epoch": 8.842708455151444,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7116,
      "step": 37803
    },
    {
      "epoch": 8.842942345924452,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 37804
    },
    {
      "epoch": 8.843176236697463,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5913,
      "step": 37805
    },
    {
      "epoch": 8.843410127470472,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.518,
      "step": 37806
    },
    {
      "epoch": 8.84364401824348,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5801,
      "step": 37807
    },
    {
      "epoch": 8.84387790901649,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.546,
      "step": 37808
    },
    {
      "epoch": 8.844111799789498,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7625,
      "step": 37809
    },
    {
      "epoch": 8.844345690562507,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7814,
      "step": 37810
    },
    {
      "epoch": 8.844579581335516,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 37811
    },
    {
      "epoch": 8.844813472108525,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8169,
      "step": 37812
    },
    {
      "epoch": 8.845047362881534,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 37813
    },
    {
      "epoch": 8.845281253654543,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 37814
    },
    {
      "epoch": 8.845515144427551,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 37815
    },
    {
      "epoch": 8.845749035200562,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6844,
      "step": 37816
    },
    {
      "epoch": 8.845982925973571,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 37817
    },
    {
      "epoch": 8.84621681674658,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7547,
      "step": 37818
    },
    {
      "epoch": 8.846450707519589,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 37819
    },
    {
      "epoch": 8.846684598292597,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8617,
      "step": 37820
    },
    {
      "epoch": 8.846918489065606,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 37821
    },
    {
      "epoch": 8.847152379838615,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 37822
    },
    {
      "epoch": 8.847386270611624,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 37823
    },
    {
      "epoch": 8.847620161384633,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 37824
    },
    {
      "epoch": 8.847854052157642,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6876,
      "step": 37825
    },
    {
      "epoch": 8.84808794293065,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 37826
    },
    {
      "epoch": 8.848321833703661,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5706,
      "step": 37827
    },
    {
      "epoch": 8.84855572447667,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 37828
    },
    {
      "epoch": 8.848789615249679,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5566,
      "step": 37829
    },
    {
      "epoch": 8.849023506022688,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 37830
    },
    {
      "epoch": 8.849257396795696,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 37831
    },
    {
      "epoch": 8.849491287568705,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 37832
    },
    {
      "epoch": 8.849725178341714,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5483,
      "step": 37833
    },
    {
      "epoch": 8.849959069114723,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 37834
    },
    {
      "epoch": 8.850192959887732,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.623,
      "step": 37835
    },
    {
      "epoch": 8.85042685066074,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 37836
    },
    {
      "epoch": 8.850660741433751,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 37837
    },
    {
      "epoch": 8.85089463220676,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 37838
    },
    {
      "epoch": 8.851128522979769,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 37839
    },
    {
      "epoch": 8.851362413752778,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 37840
    },
    {
      "epoch": 8.851596304525787,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 37841
    },
    {
      "epoch": 8.851830195298795,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 37842
    },
    {
      "epoch": 8.852064086071804,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7729,
      "step": 37843
    },
    {
      "epoch": 8.852297976844813,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9087,
      "step": 37844
    },
    {
      "epoch": 8.852531867617822,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6658,
      "step": 37845
    },
    {
      "epoch": 8.85276575839083,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6391,
      "step": 37846
    },
    {
      "epoch": 8.852999649163841,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 37847
    },
    {
      "epoch": 8.85323353993685,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 37848
    },
    {
      "epoch": 8.853467430709859,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.4961,
      "step": 37849
    },
    {
      "epoch": 8.853701321482868,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 37850
    },
    {
      "epoch": 8.853935212255877,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5869,
      "step": 37851
    },
    {
      "epoch": 8.854169103028886,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 37852
    },
    {
      "epoch": 8.854402993801894,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 37853
    },
    {
      "epoch": 8.854636884574903,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6692,
      "step": 37854
    },
    {
      "epoch": 8.854870775347912,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 37855
    },
    {
      "epoch": 8.855104666120921,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 37856
    },
    {
      "epoch": 8.85533855689393,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 37857
    },
    {
      "epoch": 8.85557244766694,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 37858
    },
    {
      "epoch": 8.85580633843995,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5981,
      "step": 37859
    },
    {
      "epoch": 8.856040229212958,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 37860
    },
    {
      "epoch": 8.856274119985967,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 37861
    },
    {
      "epoch": 8.856508010758976,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6937,
      "step": 37862
    },
    {
      "epoch": 8.856741901531985,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 37863
    },
    {
      "epoch": 8.856975792304993,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.3239,
      "step": 37864
    },
    {
      "epoch": 8.857209683078002,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.798,
      "step": 37865
    },
    {
      "epoch": 8.857443573851011,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 37866
    },
    {
      "epoch": 8.85767746462402,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 37867
    },
    {
      "epoch": 8.857911355397029,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 37868
    },
    {
      "epoch": 8.85814524617004,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 37869
    },
    {
      "epoch": 8.858379136943048,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6239,
      "step": 37870
    },
    {
      "epoch": 8.858613027716057,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 37871
    },
    {
      "epoch": 8.858846918489066,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5588,
      "step": 37872
    },
    {
      "epoch": 8.859080809262075,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 37873
    },
    {
      "epoch": 8.859314700035084,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 37874
    },
    {
      "epoch": 8.859548590808092,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 37875
    },
    {
      "epoch": 8.859782481581101,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7269,
      "step": 37876
    },
    {
      "epoch": 8.86001637235411,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 37877
    },
    {
      "epoch": 8.860250263127119,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7622,
      "step": 37878
    },
    {
      "epoch": 8.860484153900128,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.4539,
      "step": 37879
    },
    {
      "epoch": 8.860718044673138,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.56,
      "step": 37880
    },
    {
      "epoch": 8.860951935446147,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6404,
      "step": 37881
    },
    {
      "epoch": 8.861185826219156,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.3934,
      "step": 37882
    },
    {
      "epoch": 8.861419716992165,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7777,
      "step": 37883
    },
    {
      "epoch": 8.861653607765174,
      "grad_norm": 8.4375,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 37884
    },
    {
      "epoch": 8.861887498538183,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 37885
    },
    {
      "epoch": 8.862121389311191,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.8487,
      "step": 37886
    },
    {
      "epoch": 8.8623552800842,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6128,
      "step": 37887
    },
    {
      "epoch": 8.862589170857209,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5822,
      "step": 37888
    },
    {
      "epoch": 8.862823061630218,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 37889
    },
    {
      "epoch": 8.863056952403227,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7781,
      "step": 37890
    },
    {
      "epoch": 8.863290843176237,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7918,
      "step": 37891
    },
    {
      "epoch": 8.863524733949246,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 37892
    },
    {
      "epoch": 8.863758624722255,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 37893
    },
    {
      "epoch": 8.863992515495264,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 37894
    },
    {
      "epoch": 8.864226406268273,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 37895
    },
    {
      "epoch": 8.864460297041282,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.55,
      "step": 37896
    },
    {
      "epoch": 8.86469418781429,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.5408,
      "step": 37897
    },
    {
      "epoch": 8.8649280785873,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6361,
      "step": 37898
    },
    {
      "epoch": 8.865161969360308,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 37899
    },
    {
      "epoch": 8.865395860133317,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 37900
    },
    {
      "epoch": 8.865395860133317,
      "eval_runtime": 4.6766,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 37900
    },
    {
      "epoch": 8.865629750906328,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6811,
      "step": 37901
    },
    {
      "epoch": 8.865863641679336,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7082,
      "step": 37902
    },
    {
      "epoch": 8.866097532452345,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 37903
    },
    {
      "epoch": 8.866331423225354,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 37904
    },
    {
      "epoch": 8.866565313998363,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 37905
    },
    {
      "epoch": 8.866799204771372,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 37906
    },
    {
      "epoch": 8.86703309554438,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 37907
    },
    {
      "epoch": 8.86726698631739,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 37908
    },
    {
      "epoch": 8.867500877090398,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6624,
      "step": 37909
    },
    {
      "epoch": 8.867734767863407,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 37910
    },
    {
      "epoch": 8.867968658636418,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 37911
    },
    {
      "epoch": 8.868202549409427,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 37912
    },
    {
      "epoch": 8.868436440182435,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6342,
      "step": 37913
    },
    {
      "epoch": 8.868670330955444,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 37914
    },
    {
      "epoch": 8.868904221728453,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 37915
    },
    {
      "epoch": 8.869138112501462,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 37916
    },
    {
      "epoch": 8.86937200327447,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 37917
    },
    {
      "epoch": 8.86960589404748,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6002,
      "step": 37918
    },
    {
      "epoch": 8.869839784820488,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.3666,
      "step": 37919
    },
    {
      "epoch": 8.870073675593497,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4532,
      "step": 37920
    },
    {
      "epoch": 8.870307566366506,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8466,
      "step": 37921
    },
    {
      "epoch": 8.870541457139517,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5249,
      "step": 37922
    },
    {
      "epoch": 8.870775347912526,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.3296,
      "step": 37923
    },
    {
      "epoch": 8.871009238685534,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.69,
      "step": 37924
    },
    {
      "epoch": 8.871243129458543,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8108,
      "step": 37925
    },
    {
      "epoch": 8.871477020231552,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6318,
      "step": 37926
    },
    {
      "epoch": 8.87171091100456,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 37927
    },
    {
      "epoch": 8.87194480177757,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 37928
    },
    {
      "epoch": 8.872178692550579,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5742,
      "step": 37929
    },
    {
      "epoch": 8.872412583323587,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 37930
    },
    {
      "epoch": 8.872646474096596,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 37931
    },
    {
      "epoch": 8.872880364869605,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 37932
    },
    {
      "epoch": 8.873114255642616,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7119,
      "step": 37933
    },
    {
      "epoch": 8.873348146415625,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 37934
    },
    {
      "epoch": 8.873582037188633,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 37935
    },
    {
      "epoch": 8.873815927961642,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 37936
    },
    {
      "epoch": 8.874049818734651,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6856,
      "step": 37937
    },
    {
      "epoch": 8.87428370950766,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 37938
    },
    {
      "epoch": 8.874517600280669,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 37939
    },
    {
      "epoch": 8.874751491053678,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7406,
      "step": 37940
    },
    {
      "epoch": 8.874985381826686,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 37941
    },
    {
      "epoch": 8.875219272599695,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6398,
      "step": 37942
    },
    {
      "epoch": 8.875453163372704,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 37943
    },
    {
      "epoch": 8.875687054145715,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 37944
    },
    {
      "epoch": 8.875920944918724,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 37945
    },
    {
      "epoch": 8.876154835691732,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 37946
    },
    {
      "epoch": 8.876388726464741,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 37947
    },
    {
      "epoch": 8.87662261723775,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 37948
    },
    {
      "epoch": 8.876856508010759,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 37949
    },
    {
      "epoch": 8.877090398783768,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6144,
      "step": 37950
    },
    {
      "epoch": 8.877324289556777,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.5933,
      "step": 37951
    },
    {
      "epoch": 8.877558180329785,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.628,
      "step": 37952
    },
    {
      "epoch": 8.877792071102794,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 37953
    },
    {
      "epoch": 8.878025961875803,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5185,
      "step": 37954
    },
    {
      "epoch": 8.878259852648814,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 37955
    },
    {
      "epoch": 8.878493743421823,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 37956
    },
    {
      "epoch": 8.878727634194831,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 37957
    },
    {
      "epoch": 8.87896152496784,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.476,
      "step": 37958
    },
    {
      "epoch": 8.879195415740849,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 37959
    },
    {
      "epoch": 8.879429306513858,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6349,
      "step": 37960
    },
    {
      "epoch": 8.879663197286867,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 37961
    },
    {
      "epoch": 8.879897088059876,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 37962
    },
    {
      "epoch": 8.880130978832884,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 37963
    },
    {
      "epoch": 8.880364869605893,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7858,
      "step": 37964
    },
    {
      "epoch": 8.880598760378904,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 37965
    },
    {
      "epoch": 8.880832651151913,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 37966
    },
    {
      "epoch": 8.881066541924922,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2961,
      "step": 37967
    },
    {
      "epoch": 8.88130043269793,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 37968
    },
    {
      "epoch": 8.88153432347094,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 37969
    },
    {
      "epoch": 8.881768214243948,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 37970
    },
    {
      "epoch": 8.882002105016957,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.3419,
      "step": 37971
    },
    {
      "epoch": 8.882235995789966,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7456,
      "step": 37972
    },
    {
      "epoch": 8.882469886562975,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 37973
    },
    {
      "epoch": 8.882703777335983,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6372,
      "step": 37974
    },
    {
      "epoch": 8.882937668108994,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6711,
      "step": 37975
    },
    {
      "epoch": 8.883171558882003,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8282,
      "step": 37976
    },
    {
      "epoch": 8.883405449655012,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 37977
    },
    {
      "epoch": 8.88363934042802,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 37978
    },
    {
      "epoch": 8.88387323120103,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 37979
    },
    {
      "epoch": 8.884107121974038,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8547,
      "step": 37980
    },
    {
      "epoch": 8.884341012747047,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 37981
    },
    {
      "epoch": 8.884574903520056,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4959,
      "step": 37982
    },
    {
      "epoch": 8.884808794293065,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 37983
    },
    {
      "epoch": 8.885042685066074,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.1233,
      "step": 37984
    },
    {
      "epoch": 8.885276575839082,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7705,
      "step": 37985
    },
    {
      "epoch": 8.885510466612093,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 37986
    },
    {
      "epoch": 8.885744357385102,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 37987
    },
    {
      "epoch": 8.88597824815811,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 37988
    },
    {
      "epoch": 8.88621213893112,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6365,
      "step": 37989
    },
    {
      "epoch": 8.886446029704128,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7228,
      "step": 37990
    },
    {
      "epoch": 8.886679920477137,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.3743,
      "step": 37991
    },
    {
      "epoch": 8.886913811250146,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5354,
      "step": 37992
    },
    {
      "epoch": 8.887147702023155,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4303,
      "step": 37993
    },
    {
      "epoch": 8.887381592796164,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.342,
      "step": 37994
    },
    {
      "epoch": 8.887615483569173,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 37995
    },
    {
      "epoch": 8.887849374342181,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 37996
    },
    {
      "epoch": 8.888083265115192,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 37997
    },
    {
      "epoch": 8.8883171558882,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 37998
    },
    {
      "epoch": 8.88855104666121,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 37999
    },
    {
      "epoch": 8.888784937434218,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 38000
    },
    {
      "epoch": 8.888784937434218,
      "eval_runtime": 4.6151,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 38000
    },
    {
      "epoch": 8.889018828207227,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.4407,
      "step": 38001
    },
    {
      "epoch": 8.889252718980236,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 38002
    },
    {
      "epoch": 8.889486609753245,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6242,
      "step": 38003
    },
    {
      "epoch": 8.889720500526254,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6726,
      "step": 38004
    },
    {
      "epoch": 8.889954391299263,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 38005
    },
    {
      "epoch": 8.890188282072272,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 38006
    },
    {
      "epoch": 8.89042217284528,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 38007
    },
    {
      "epoch": 8.890656063618291,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6582,
      "step": 38008
    },
    {
      "epoch": 8.8908899543913,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 38009
    },
    {
      "epoch": 8.891123845164309,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 38010
    },
    {
      "epoch": 8.891357735937317,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7292,
      "step": 38011
    },
    {
      "epoch": 8.891591626710326,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 38012
    },
    {
      "epoch": 8.891825517483335,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 38013
    },
    {
      "epoch": 8.892059408256344,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 38014
    },
    {
      "epoch": 8.892293299029353,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 38015
    },
    {
      "epoch": 8.892527189802362,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8115,
      "step": 38016
    },
    {
      "epoch": 8.89276108057537,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6678,
      "step": 38017
    },
    {
      "epoch": 8.892994971348381,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 38018
    },
    {
      "epoch": 8.89322886212139,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 38019
    },
    {
      "epoch": 8.893462752894399,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 38020
    },
    {
      "epoch": 8.893696643667408,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 38021
    },
    {
      "epoch": 8.893930534440416,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8322,
      "step": 38022
    },
    {
      "epoch": 8.894164425213425,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 38023
    },
    {
      "epoch": 8.894398315986434,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.3662,
      "step": 38024
    },
    {
      "epoch": 8.894632206759443,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7219,
      "step": 38025
    },
    {
      "epoch": 8.894866097532452,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 38026
    },
    {
      "epoch": 8.89509998830546,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5241,
      "step": 38027
    },
    {
      "epoch": 8.895333879078471,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7368,
      "step": 38028
    },
    {
      "epoch": 8.89556776985148,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 38029
    },
    {
      "epoch": 8.895801660624489,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6409,
      "step": 38030
    },
    {
      "epoch": 8.896035551397498,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 38031
    },
    {
      "epoch": 8.896269442170507,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7694,
      "step": 38032
    },
    {
      "epoch": 8.896503332943515,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6645,
      "step": 38033
    },
    {
      "epoch": 8.896737223716524,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 38034
    },
    {
      "epoch": 8.896971114489533,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 38035
    },
    {
      "epoch": 8.897205005262542,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 38036
    },
    {
      "epoch": 8.89743889603555,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 38037
    },
    {
      "epoch": 8.89767278680856,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6701,
      "step": 38038
    },
    {
      "epoch": 8.89790667758157,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 38039
    },
    {
      "epoch": 8.89814056835458,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 38040
    },
    {
      "epoch": 8.898374459127588,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 38041
    },
    {
      "epoch": 8.898608349900597,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4785,
      "step": 38042
    },
    {
      "epoch": 8.898842240673606,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 38043
    },
    {
      "epoch": 8.899076131446614,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 38044
    },
    {
      "epoch": 8.899310022219623,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 38045
    },
    {
      "epoch": 8.899543912992632,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 38046
    },
    {
      "epoch": 8.899777803765641,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 38047
    },
    {
      "epoch": 8.90001169453865,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7228,
      "step": 38048
    },
    {
      "epoch": 8.900245585311659,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 38049
    },
    {
      "epoch": 8.90047947608467,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 38050
    },
    {
      "epoch": 8.900713366857678,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 38051
    },
    {
      "epoch": 8.900947257630687,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 38052
    },
    {
      "epoch": 8.901181148403696,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 38053
    },
    {
      "epoch": 8.901415039176705,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 38054
    },
    {
      "epoch": 8.901648929949713,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 38055
    },
    {
      "epoch": 8.901882820722722,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 38056
    },
    {
      "epoch": 8.902116711495731,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 38057
    },
    {
      "epoch": 8.90235060226874,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 38058
    },
    {
      "epoch": 8.902584493041749,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7469,
      "step": 38059
    },
    {
      "epoch": 8.902818383814758,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.564,
      "step": 38060
    },
    {
      "epoch": 8.903052274587768,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 38061
    },
    {
      "epoch": 8.903286165360777,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 38062
    },
    {
      "epoch": 8.903520056133786,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6307,
      "step": 38063
    },
    {
      "epoch": 8.903753946906795,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 38064
    },
    {
      "epoch": 8.903987837679804,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5848,
      "step": 38065
    },
    {
      "epoch": 8.904221728452812,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 38066
    },
    {
      "epoch": 8.904455619225821,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6459,
      "step": 38067
    },
    {
      "epoch": 8.90468950999883,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9244,
      "step": 38068
    },
    {
      "epoch": 8.904923400771839,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.916,
      "step": 38069
    },
    {
      "epoch": 8.905157291544848,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 38070
    },
    {
      "epoch": 8.905391182317857,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 38071
    },
    {
      "epoch": 8.905625073090867,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 38072
    },
    {
      "epoch": 8.905858963863876,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6368,
      "step": 38073
    },
    {
      "epoch": 8.906092854636885,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 38074
    },
    {
      "epoch": 8.906326745409894,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 38075
    },
    {
      "epoch": 8.906560636182903,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 38076
    },
    {
      "epoch": 8.906794526955911,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 38077
    },
    {
      "epoch": 8.90702841772892,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4933,
      "step": 38078
    },
    {
      "epoch": 8.90726230850193,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 38079
    },
    {
      "epoch": 8.907496199274938,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 38080
    },
    {
      "epoch": 8.907730090047947,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 38081
    },
    {
      "epoch": 8.907963980820957,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5096,
      "step": 38082
    },
    {
      "epoch": 8.908197871593966,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 38083
    },
    {
      "epoch": 8.908431762366975,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6718,
      "step": 38084
    },
    {
      "epoch": 8.908665653139984,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7352,
      "step": 38085
    },
    {
      "epoch": 8.908899543912993,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 38086
    },
    {
      "epoch": 8.909133434686002,
      "grad_norm": 8.0,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 38087
    },
    {
      "epoch": 8.90936732545901,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 38088
    },
    {
      "epoch": 8.90960121623202,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6988,
      "step": 38089
    },
    {
      "epoch": 8.909835107005028,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 38090
    },
    {
      "epoch": 8.910068997778037,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 38091
    },
    {
      "epoch": 8.910302888551048,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 38092
    },
    {
      "epoch": 8.910536779324056,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 38093
    },
    {
      "epoch": 8.910770670097065,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.3829,
      "step": 38094
    },
    {
      "epoch": 8.911004560870074,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5731,
      "step": 38095
    },
    {
      "epoch": 8.911238451643083,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 38096
    },
    {
      "epoch": 8.911472342416092,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7475,
      "step": 38097
    },
    {
      "epoch": 8.9117062331891,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 38098
    },
    {
      "epoch": 8.91194012396211,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 38099
    },
    {
      "epoch": 8.912174014735118,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 38100
    },
    {
      "epoch": 8.912174014735118,
      "eval_runtime": 4.666,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 38100
    },
    {
      "epoch": 8.912407905508127,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7854,
      "step": 38101
    },
    {
      "epoch": 8.912641796281136,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6154,
      "step": 38102
    },
    {
      "epoch": 8.912875687054147,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 38103
    },
    {
      "epoch": 8.913109577827155,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 38104
    },
    {
      "epoch": 8.913343468600164,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 38105
    },
    {
      "epoch": 8.913577359373173,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 38106
    },
    {
      "epoch": 8.913811250146182,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 38107
    },
    {
      "epoch": 8.91404514091919,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 38108
    },
    {
      "epoch": 8.9142790316922,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 38109
    },
    {
      "epoch": 8.914512922465208,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6136,
      "step": 38110
    },
    {
      "epoch": 8.914746813238217,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7037,
      "step": 38111
    },
    {
      "epoch": 8.914980704011226,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6643,
      "step": 38112
    },
    {
      "epoch": 8.915214594784235,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 38113
    },
    {
      "epoch": 8.915448485557246,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 38114
    },
    {
      "epoch": 8.915682376330254,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 38115
    },
    {
      "epoch": 8.915916267103263,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 38116
    },
    {
      "epoch": 8.916150157876272,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 38117
    },
    {
      "epoch": 8.916384048649281,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6216,
      "step": 38118
    },
    {
      "epoch": 8.91661793942229,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 38119
    },
    {
      "epoch": 8.916851830195299,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8634,
      "step": 38120
    },
    {
      "epoch": 8.917085720968307,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 38121
    },
    {
      "epoch": 8.917319611741316,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7246,
      "step": 38122
    },
    {
      "epoch": 8.917553502514325,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5051,
      "step": 38123
    },
    {
      "epoch": 8.917787393287334,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 38124
    },
    {
      "epoch": 8.918021284060345,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 38125
    },
    {
      "epoch": 8.918255174833353,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 38126
    },
    {
      "epoch": 8.918489065606362,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 38127
    },
    {
      "epoch": 8.918722956379371,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 38128
    },
    {
      "epoch": 8.91895684715238,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 38129
    },
    {
      "epoch": 8.919190737925389,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7599,
      "step": 38130
    },
    {
      "epoch": 8.919424628698398,
      "grad_norm": 7.75,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 38131
    },
    {
      "epoch": 8.919658519471406,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 38132
    },
    {
      "epoch": 8.919892410244415,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5667,
      "step": 38133
    },
    {
      "epoch": 8.920126301017424,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4718,
      "step": 38134
    },
    {
      "epoch": 8.920360191790433,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5446,
      "step": 38135
    },
    {
      "epoch": 8.920594082563444,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6749,
      "step": 38136
    },
    {
      "epoch": 8.920827973336452,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6466,
      "step": 38137
    },
    {
      "epoch": 8.921061864109461,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 38138
    },
    {
      "epoch": 8.92129575488247,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 38139
    },
    {
      "epoch": 8.921529645655479,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6824,
      "step": 38140
    },
    {
      "epoch": 8.921763536428488,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5397,
      "step": 38141
    },
    {
      "epoch": 8.921997427201497,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 38142
    },
    {
      "epoch": 8.922231317974505,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4038,
      "step": 38143
    },
    {
      "epoch": 8.922465208747514,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 38144
    },
    {
      "epoch": 8.922699099520523,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7411,
      "step": 38145
    },
    {
      "epoch": 8.922932990293534,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 38146
    },
    {
      "epoch": 8.923166881066543,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7018,
      "step": 38147
    },
    {
      "epoch": 8.923400771839551,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4666,
      "step": 38148
    },
    {
      "epoch": 8.92363466261256,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6065,
      "step": 38149
    },
    {
      "epoch": 8.923868553385569,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 38150
    },
    {
      "epoch": 8.924102444158578,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 38151
    },
    {
      "epoch": 8.924336334931587,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 38152
    },
    {
      "epoch": 8.924570225704596,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6158,
      "step": 38153
    },
    {
      "epoch": 8.924804116477604,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7725,
      "step": 38154
    },
    {
      "epoch": 8.925038007250613,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6351,
      "step": 38155
    },
    {
      "epoch": 8.925271898023624,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 38156
    },
    {
      "epoch": 8.925505788796633,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 38157
    },
    {
      "epoch": 8.925739679569642,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.674,
      "step": 38158
    },
    {
      "epoch": 8.92597357034265,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6047,
      "step": 38159
    },
    {
      "epoch": 8.92620746111566,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 38160
    },
    {
      "epoch": 8.926441351888668,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.3576,
      "step": 38161
    },
    {
      "epoch": 8.926675242661677,
      "grad_norm": 13.5625,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 38162
    },
    {
      "epoch": 8.926909133434686,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.3341,
      "step": 38163
    },
    {
      "epoch": 8.927143024207695,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 38164
    },
    {
      "epoch": 8.927376914980703,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6108,
      "step": 38165
    },
    {
      "epoch": 8.927610805753712,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7388,
      "step": 38166
    },
    {
      "epoch": 8.927844696526723,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 38167
    },
    {
      "epoch": 8.928078587299732,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 38168
    },
    {
      "epoch": 8.92831247807274,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9937,
      "step": 38169
    },
    {
      "epoch": 8.92854636884575,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 38170
    },
    {
      "epoch": 8.928780259618758,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.5949,
      "step": 38171
    },
    {
      "epoch": 8.929014150391767,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 38172
    },
    {
      "epoch": 8.929248041164776,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 38173
    },
    {
      "epoch": 8.929481931937785,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.5769,
      "step": 38174
    },
    {
      "epoch": 8.929715822710794,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.779,
      "step": 38175
    },
    {
      "epoch": 8.929949713483802,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6382,
      "step": 38176
    },
    {
      "epoch": 8.930183604256811,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.613,
      "step": 38177
    },
    {
      "epoch": 8.930417495029822,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6695,
      "step": 38178
    },
    {
      "epoch": 8.93065138580283,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 38179
    },
    {
      "epoch": 8.93088527657584,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.65,
      "step": 38180
    },
    {
      "epoch": 8.931119167348848,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 38181
    },
    {
      "epoch": 8.931353058121857,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 38182
    },
    {
      "epoch": 8.931586948894866,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 38183
    },
    {
      "epoch": 8.931820839667875,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5643,
      "step": 38184
    },
    {
      "epoch": 8.932054730440884,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 38185
    },
    {
      "epoch": 8.932288621213893,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 38186
    },
    {
      "epoch": 8.932522511986901,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 38187
    },
    {
      "epoch": 8.93275640275991,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7396,
      "step": 38188
    },
    {
      "epoch": 8.93299029353292,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 38189
    },
    {
      "epoch": 8.93322418430593,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6481,
      "step": 38190
    },
    {
      "epoch": 8.933458075078939,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 38191
    },
    {
      "epoch": 8.933691965851947,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 38192
    },
    {
      "epoch": 8.933925856624956,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 38193
    },
    {
      "epoch": 8.934159747397965,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 38194
    },
    {
      "epoch": 8.934393638170974,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 38195
    },
    {
      "epoch": 8.934627528943983,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8462,
      "step": 38196
    },
    {
      "epoch": 8.934861419716992,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 38197
    },
    {
      "epoch": 8.93509531049,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 38198
    },
    {
      "epoch": 8.935329201263011,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6294,
      "step": 38199
    },
    {
      "epoch": 8.93556309203602,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 38200
    },
    {
      "epoch": 8.93556309203602,
      "eval_runtime": 4.7425,
      "eval_samples_per_second": 0.211,
      "eval_steps_per_second": 0.211,
      "step": 38200
    },
    {
      "epoch": 8.935796982809029,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.4497,
      "step": 38201
    },
    {
      "epoch": 8.936030873582038,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6474,
      "step": 38202
    },
    {
      "epoch": 8.936264764355046,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 38203
    },
    {
      "epoch": 8.936498655128055,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6248,
      "step": 38204
    },
    {
      "epoch": 8.936732545901064,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 38205
    },
    {
      "epoch": 8.936966436674073,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 38206
    },
    {
      "epoch": 8.937200327447082,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1401,
      "step": 38207
    },
    {
      "epoch": 8.93743421822009,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 38208
    },
    {
      "epoch": 8.937668108993101,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 38209
    },
    {
      "epoch": 8.93790199976611,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6837,
      "step": 38210
    },
    {
      "epoch": 8.938135890539119,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5804,
      "step": 38211
    },
    {
      "epoch": 8.938369781312128,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 38212
    },
    {
      "epoch": 8.938603672085137,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6581,
      "step": 38213
    },
    {
      "epoch": 8.938837562858145,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5731,
      "step": 38214
    },
    {
      "epoch": 8.939071453631154,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 38215
    },
    {
      "epoch": 8.939305344404163,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.4369,
      "step": 38216
    },
    {
      "epoch": 8.939539235177172,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7664,
      "step": 38217
    },
    {
      "epoch": 8.93977312595018,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5028,
      "step": 38218
    },
    {
      "epoch": 8.94000701672319,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 38219
    },
    {
      "epoch": 8.9402409074962,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 38220
    },
    {
      "epoch": 8.940474798269209,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 38221
    },
    {
      "epoch": 8.940708689042218,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7933,
      "step": 38222
    },
    {
      "epoch": 8.940942579815227,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 38223
    },
    {
      "epoch": 8.941176470588236,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 38224
    },
    {
      "epoch": 8.941410361361244,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 38225
    },
    {
      "epoch": 8.941644252134253,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 38226
    },
    {
      "epoch": 8.941878142907262,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 38227
    },
    {
      "epoch": 8.94211203368027,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 38228
    },
    {
      "epoch": 8.94234592445328,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 38229
    },
    {
      "epoch": 8.942579815226289,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 38230
    },
    {
      "epoch": 8.9428137059993,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 38231
    },
    {
      "epoch": 8.943047596772308,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6999,
      "step": 38232
    },
    {
      "epoch": 8.943281487545317,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 38233
    },
    {
      "epoch": 8.943515378318326,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 38234
    },
    {
      "epoch": 8.943749269091334,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.573,
      "step": 38235
    },
    {
      "epoch": 8.943983159864343,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 38236
    },
    {
      "epoch": 8.944217050637352,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 38237
    },
    {
      "epoch": 8.944450941410361,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.5975,
      "step": 38238
    },
    {
      "epoch": 8.94468483218337,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 38239
    },
    {
      "epoch": 8.944918722956379,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 38240
    },
    {
      "epoch": 8.945152613729388,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 38241
    },
    {
      "epoch": 8.945386504502398,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 38242
    },
    {
      "epoch": 8.945620395275407,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7048,
      "step": 38243
    },
    {
      "epoch": 8.945854286048416,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7872,
      "step": 38244
    },
    {
      "epoch": 8.946088176821425,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 38245
    },
    {
      "epoch": 8.946322067594433,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 38246
    },
    {
      "epoch": 8.946555958367442,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7807,
      "step": 38247
    },
    {
      "epoch": 8.946789849140451,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 38248
    },
    {
      "epoch": 8.94702373991346,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5843,
      "step": 38249
    },
    {
      "epoch": 8.947257630686469,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6302,
      "step": 38250
    },
    {
      "epoch": 8.947491521459478,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 38251
    },
    {
      "epoch": 8.947725412232487,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 38252
    },
    {
      "epoch": 8.947959303005497,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7498,
      "step": 38253
    },
    {
      "epoch": 8.948193193778506,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 38254
    },
    {
      "epoch": 8.948427084551515,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 38255
    },
    {
      "epoch": 8.948660975324524,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.565,
      "step": 38256
    },
    {
      "epoch": 8.948894866097532,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6751,
      "step": 38257
    },
    {
      "epoch": 8.949128756870541,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7447,
      "step": 38258
    },
    {
      "epoch": 8.94936264764355,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 38259
    },
    {
      "epoch": 8.949596538416559,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7025,
      "step": 38260
    },
    {
      "epoch": 8.949830429189568,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 38261
    },
    {
      "epoch": 8.950064319962577,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 38262
    },
    {
      "epoch": 8.950298210735587,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5431,
      "step": 38263
    },
    {
      "epoch": 8.950532101508596,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 38264
    },
    {
      "epoch": 8.950765992281605,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 38265
    },
    {
      "epoch": 8.950999883054614,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.3998,
      "step": 38266
    },
    {
      "epoch": 8.951233773827623,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6695,
      "step": 38267
    },
    {
      "epoch": 8.951467664600631,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5832,
      "step": 38268
    },
    {
      "epoch": 8.95170155537364,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7552,
      "step": 38269
    },
    {
      "epoch": 8.95193544614665,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 38270
    },
    {
      "epoch": 8.952169336919658,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 38271
    },
    {
      "epoch": 8.952403227692667,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 38272
    },
    {
      "epoch": 8.952637118465677,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 38273
    },
    {
      "epoch": 8.952871009238686,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 38274
    },
    {
      "epoch": 8.953104900011695,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8837,
      "step": 38275
    },
    {
      "epoch": 8.953338790784704,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.623,
      "step": 38276
    },
    {
      "epoch": 8.953572681557713,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4042,
      "step": 38277
    },
    {
      "epoch": 8.953806572330722,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8219,
      "step": 38278
    },
    {
      "epoch": 8.95404046310373,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 38279
    },
    {
      "epoch": 8.95427435387674,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 38280
    },
    {
      "epoch": 8.954508244649748,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 38281
    },
    {
      "epoch": 8.954742135422757,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 38282
    },
    {
      "epoch": 8.954976026195766,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 38283
    },
    {
      "epoch": 8.955209916968776,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.526,
      "step": 38284
    },
    {
      "epoch": 8.955443807741785,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 38285
    },
    {
      "epoch": 8.955677698514794,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 38286
    },
    {
      "epoch": 8.955911589287803,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.857,
      "step": 38287
    },
    {
      "epoch": 8.956145480060812,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6835,
      "step": 38288
    },
    {
      "epoch": 8.95637937083382,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.373,
      "step": 38289
    },
    {
      "epoch": 8.95661326160683,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 38290
    },
    {
      "epoch": 8.956847152379838,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7339,
      "step": 38291
    },
    {
      "epoch": 8.957081043152847,
      "grad_norm": 7.90625,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 38292
    },
    {
      "epoch": 8.957314933925856,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 38293
    },
    {
      "epoch": 8.957548824698865,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 38294
    },
    {
      "epoch": 8.957782715471875,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 38295
    },
    {
      "epoch": 8.958016606244884,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 38296
    },
    {
      "epoch": 8.958250497017893,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 38297
    },
    {
      "epoch": 8.958484387790902,
      "grad_norm": 8.5625,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 38298
    },
    {
      "epoch": 8.95871827856391,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.714,
      "step": 38299
    },
    {
      "epoch": 8.95895216933692,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7614,
      "step": 38300
    },
    {
      "epoch": 8.95895216933692,
      "eval_runtime": 4.6429,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 38300
    },
    {
      "epoch": 8.959186060109928,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 38301
    },
    {
      "epoch": 8.959419950882937,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 38302
    },
    {
      "epoch": 8.959653841655946,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 38303
    },
    {
      "epoch": 8.959887732428955,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8349,
      "step": 38304
    },
    {
      "epoch": 8.960121623201964,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 38305
    },
    {
      "epoch": 8.960355513974974,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 38306
    },
    {
      "epoch": 8.960589404747983,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 38307
    },
    {
      "epoch": 8.960823295520992,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 38308
    },
    {
      "epoch": 8.961057186294001,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2395,
      "step": 38309
    },
    {
      "epoch": 8.96129107706701,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 38310
    },
    {
      "epoch": 8.961524967840019,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6682,
      "step": 38311
    },
    {
      "epoch": 8.961758858613027,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4786,
      "step": 38312
    },
    {
      "epoch": 8.961992749386036,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6289,
      "step": 38313
    },
    {
      "epoch": 8.962226640159045,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5756,
      "step": 38314
    },
    {
      "epoch": 8.962460530932054,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6553,
      "step": 38315
    },
    {
      "epoch": 8.962694421705063,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.6273,
      "step": 38316
    },
    {
      "epoch": 8.962928312478073,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.3951,
      "step": 38317
    },
    {
      "epoch": 8.963162203251082,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.676,
      "step": 38318
    },
    {
      "epoch": 8.963396094024091,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 38319
    },
    {
      "epoch": 8.9636299847971,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7531,
      "step": 38320
    },
    {
      "epoch": 8.963863875570109,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7069,
      "step": 38321
    },
    {
      "epoch": 8.964097766343118,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 38322
    },
    {
      "epoch": 8.964331657116126,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 38323
    },
    {
      "epoch": 8.964565547889135,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6893,
      "step": 38324
    },
    {
      "epoch": 8.964799438662144,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 38325
    },
    {
      "epoch": 8.965033329435153,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.647,
      "step": 38326
    },
    {
      "epoch": 8.965267220208164,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4672,
      "step": 38327
    },
    {
      "epoch": 8.965501110981172,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 38328
    },
    {
      "epoch": 8.965735001754181,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 38329
    },
    {
      "epoch": 8.96596889252719,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 38330
    },
    {
      "epoch": 8.966202783300199,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 38331
    },
    {
      "epoch": 8.966436674073208,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 38332
    },
    {
      "epoch": 8.966670564846217,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5997,
      "step": 38333
    },
    {
      "epoch": 8.966904455619225,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.3896,
      "step": 38334
    },
    {
      "epoch": 8.967138346392234,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7731,
      "step": 38335
    },
    {
      "epoch": 8.967372237165243,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8849,
      "step": 38336
    },
    {
      "epoch": 8.967606127938254,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6454,
      "step": 38337
    },
    {
      "epoch": 8.967840018711263,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 38338
    },
    {
      "epoch": 8.968073909484271,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 38339
    },
    {
      "epoch": 8.96830780025728,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 38340
    },
    {
      "epoch": 8.968541691030289,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 38341
    },
    {
      "epoch": 8.968775581803298,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.647,
      "step": 38342
    },
    {
      "epoch": 8.969009472576307,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.5388,
      "step": 38343
    },
    {
      "epoch": 8.969243363349316,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 38344
    },
    {
      "epoch": 8.969477254122324,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 38345
    },
    {
      "epoch": 8.969711144895333,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 38346
    },
    {
      "epoch": 8.969945035668342,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.5997,
      "step": 38347
    },
    {
      "epoch": 8.970178926441353,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 38348
    },
    {
      "epoch": 8.970412817214362,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 38349
    },
    {
      "epoch": 8.97064670798737,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 38350
    },
    {
      "epoch": 8.97088059876038,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6571,
      "step": 38351
    },
    {
      "epoch": 8.971114489533388,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5281,
      "step": 38352
    },
    {
      "epoch": 8.971348380306397,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.589,
      "step": 38353
    },
    {
      "epoch": 8.971582271079406,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5597,
      "step": 38354
    },
    {
      "epoch": 8.971816161852415,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6215,
      "step": 38355
    },
    {
      "epoch": 8.972050052625423,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6741,
      "step": 38356
    },
    {
      "epoch": 8.972283943398432,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6382,
      "step": 38357
    },
    {
      "epoch": 8.972517834171441,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 38358
    },
    {
      "epoch": 8.972751724944452,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.6299,
      "step": 38359
    },
    {
      "epoch": 8.97298561571746,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 38360
    },
    {
      "epoch": 8.97321950649047,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5813,
      "step": 38361
    },
    {
      "epoch": 8.973453397263478,
      "grad_norm": 9.3125,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 38362
    },
    {
      "epoch": 8.973687288036487,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 38363
    },
    {
      "epoch": 8.973921178809496,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 38364
    },
    {
      "epoch": 8.974155069582505,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 38365
    },
    {
      "epoch": 8.974388960355514,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6295,
      "step": 38366
    },
    {
      "epoch": 8.974622851128522,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.2591,
      "step": 38367
    },
    {
      "epoch": 8.974856741901531,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7406,
      "step": 38368
    },
    {
      "epoch": 8.97509063267454,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5582,
      "step": 38369
    },
    {
      "epoch": 8.97532452344755,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 38370
    },
    {
      "epoch": 8.97555841422056,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6785,
      "step": 38371
    },
    {
      "epoch": 8.975792304993568,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 38372
    },
    {
      "epoch": 8.976026195766577,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 38373
    },
    {
      "epoch": 8.976260086539586,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4628,
      "step": 38374
    },
    {
      "epoch": 8.976493977312595,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 38375
    },
    {
      "epoch": 8.976727868085604,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 38376
    },
    {
      "epoch": 8.976961758858613,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 38377
    },
    {
      "epoch": 8.977195649631621,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7936,
      "step": 38378
    },
    {
      "epoch": 8.97742954040463,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5985,
      "step": 38379
    },
    {
      "epoch": 8.97766343117764,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.3554,
      "step": 38380
    },
    {
      "epoch": 8.97789732195065,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6808,
      "step": 38381
    },
    {
      "epoch": 8.978131212723659,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 38382
    },
    {
      "epoch": 8.978365103496667,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 38383
    },
    {
      "epoch": 8.978598994269676,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 38384
    },
    {
      "epoch": 8.978832885042685,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.4283,
      "step": 38385
    },
    {
      "epoch": 8.979066775815694,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 38386
    },
    {
      "epoch": 8.979300666588703,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 38387
    },
    {
      "epoch": 8.979534557361712,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 38388
    },
    {
      "epoch": 8.97976844813472,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6104,
      "step": 38389
    },
    {
      "epoch": 8.980002338907731,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6658,
      "step": 38390
    },
    {
      "epoch": 8.98023622968074,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.708,
      "step": 38391
    },
    {
      "epoch": 8.980470120453749,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6035,
      "step": 38392
    },
    {
      "epoch": 8.980704011226758,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6341,
      "step": 38393
    },
    {
      "epoch": 8.980937901999766,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 38394
    },
    {
      "epoch": 8.981171792772775,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6485,
      "step": 38395
    },
    {
      "epoch": 8.981405683545784,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4332,
      "step": 38396
    },
    {
      "epoch": 8.981639574318793,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 38397
    },
    {
      "epoch": 8.981873465091802,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 38398
    },
    {
      "epoch": 8.98210735586481,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 38399
    },
    {
      "epoch": 8.98234124663782,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6812,
      "step": 38400
    },
    {
      "epoch": 8.98234124663782,
      "eval_runtime": 4.5912,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 38400
    },
    {
      "epoch": 8.98257513741083,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 38401
    },
    {
      "epoch": 8.982809028183839,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6225,
      "step": 38402
    },
    {
      "epoch": 8.983042918956848,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 38403
    },
    {
      "epoch": 8.983276809729857,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 38404
    },
    {
      "epoch": 8.983510700502865,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8851,
      "step": 38405
    },
    {
      "epoch": 8.983744591275874,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 38406
    },
    {
      "epoch": 8.983978482048883,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 38407
    },
    {
      "epoch": 8.984212372821892,
      "grad_norm": 7.28125,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 38408
    },
    {
      "epoch": 8.9844462635949,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 38409
    },
    {
      "epoch": 8.98468015436791,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6217,
      "step": 38410
    },
    {
      "epoch": 8.984914045140918,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7644,
      "step": 38411
    },
    {
      "epoch": 8.985147935913929,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 38412
    },
    {
      "epoch": 8.985381826686938,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 38413
    },
    {
      "epoch": 8.985615717459947,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 38414
    },
    {
      "epoch": 8.985849608232956,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6214,
      "step": 38415
    },
    {
      "epoch": 8.986083499005964,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6203,
      "step": 38416
    },
    {
      "epoch": 8.986317389778973,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5768,
      "step": 38417
    },
    {
      "epoch": 8.986551280551982,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 38418
    },
    {
      "epoch": 8.98678517132499,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 38419
    },
    {
      "epoch": 8.987019062098,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7585,
      "step": 38420
    },
    {
      "epoch": 8.987252952871009,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.462,
      "step": 38421
    },
    {
      "epoch": 8.987486843644017,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7424,
      "step": 38422
    },
    {
      "epoch": 8.987720734417028,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 38423
    },
    {
      "epoch": 8.987954625190037,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 38424
    },
    {
      "epoch": 8.988188515963046,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 38425
    },
    {
      "epoch": 8.988422406736055,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 38426
    },
    {
      "epoch": 8.988656297509063,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.4408,
      "step": 38427
    },
    {
      "epoch": 8.988890188282072,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 38428
    },
    {
      "epoch": 8.989124079055081,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 38429
    },
    {
      "epoch": 8.98935796982809,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8645,
      "step": 38430
    },
    {
      "epoch": 8.989591860601099,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 38431
    },
    {
      "epoch": 8.989825751374108,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6978,
      "step": 38432
    },
    {
      "epoch": 8.990059642147116,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 38433
    },
    {
      "epoch": 8.990293532920127,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 38434
    },
    {
      "epoch": 8.990527423693136,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.546,
      "step": 38435
    },
    {
      "epoch": 8.990761314466145,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 38436
    },
    {
      "epoch": 8.990995205239154,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 38437
    },
    {
      "epoch": 8.991229096012162,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.399,
      "step": 38438
    },
    {
      "epoch": 8.991462986785171,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 38439
    },
    {
      "epoch": 8.99169687755818,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 38440
    },
    {
      "epoch": 8.991930768331189,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 38441
    },
    {
      "epoch": 8.992164659104198,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 38442
    },
    {
      "epoch": 8.992398549877207,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5859,
      "step": 38443
    },
    {
      "epoch": 8.992632440650217,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 38444
    },
    {
      "epoch": 8.992866331423226,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2578,
      "step": 38445
    },
    {
      "epoch": 8.993100222196235,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 38446
    },
    {
      "epoch": 8.993334112969244,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 38447
    },
    {
      "epoch": 8.993568003742253,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 38448
    },
    {
      "epoch": 8.993801894515261,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5317,
      "step": 38449
    },
    {
      "epoch": 8.99403578528827,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 38450
    },
    {
      "epoch": 8.994269676061279,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 38451
    },
    {
      "epoch": 8.994503566834288,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8059,
      "step": 38452
    },
    {
      "epoch": 8.994737457607297,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 38453
    },
    {
      "epoch": 8.994971348380307,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 38454
    },
    {
      "epoch": 8.995205239153316,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7205,
      "step": 38455
    },
    {
      "epoch": 8.995439129926325,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6763,
      "step": 38456
    },
    {
      "epoch": 8.995673020699334,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6571,
      "step": 38457
    },
    {
      "epoch": 8.995906911472343,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.802,
      "step": 38458
    },
    {
      "epoch": 8.996140802245352,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6772,
      "step": 38459
    },
    {
      "epoch": 8.99637469301836,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5256,
      "step": 38460
    },
    {
      "epoch": 8.99660858379137,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.714,
      "step": 38461
    },
    {
      "epoch": 8.996842474564378,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 38462
    },
    {
      "epoch": 8.997076365337387,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7713,
      "step": 38463
    },
    {
      "epoch": 8.997310256110396,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 38464
    },
    {
      "epoch": 8.997544146883406,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5479,
      "step": 38465
    },
    {
      "epoch": 8.997778037656415,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7537,
      "step": 38466
    },
    {
      "epoch": 8.998011928429424,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 38467
    },
    {
      "epoch": 8.998245819202433,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 38468
    },
    {
      "epoch": 8.998479709975442,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 38469
    },
    {
      "epoch": 8.99871360074845,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.522,
      "step": 38470
    },
    {
      "epoch": 8.99894749152146,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6933,
      "step": 38471
    },
    {
      "epoch": 8.999181382294468,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 38472
    },
    {
      "epoch": 8.999415273067477,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 38473
    },
    {
      "epoch": 8.999649163840486,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 38474
    },
    {
      "epoch": 8.999883054613495,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 38475
    },
    {
      "epoch": 9.000233890773009,
      "grad_norm": 15.0,
      "learning_rate": 3e-05,
      "loss": 3.2977,
      "step": 38476
    },
    {
      "epoch": 9.000467781546018,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 38477
    },
    {
      "epoch": 9.000701672319027,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 38478
    },
    {
      "epoch": 9.000935563092035,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 38479
    },
    {
      "epoch": 9.001169453865044,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8075,
      "step": 38480
    },
    {
      "epoch": 9.001403344638055,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6525,
      "step": 38481
    },
    {
      "epoch": 9.001637235411064,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 38482
    },
    {
      "epoch": 9.001871126184072,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 38483
    },
    {
      "epoch": 9.002105016957081,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6967,
      "step": 38484
    },
    {
      "epoch": 9.00233890773009,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.4854,
      "step": 38485
    },
    {
      "epoch": 9.002572798503099,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5659,
      "step": 38486
    },
    {
      "epoch": 9.002806689276108,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 38487
    },
    {
      "epoch": 9.003040580049117,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 38488
    },
    {
      "epoch": 9.003274470822126,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 38489
    },
    {
      "epoch": 9.003508361595134,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 38490
    },
    {
      "epoch": 9.003742252368145,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 38491
    },
    {
      "epoch": 9.003976143141154,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4526,
      "step": 38492
    },
    {
      "epoch": 9.004210033914163,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5059,
      "step": 38493
    },
    {
      "epoch": 9.004443924687171,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 38494
    },
    {
      "epoch": 9.00467781546018,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5668,
      "step": 38495
    },
    {
      "epoch": 9.00491170623319,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 38496
    },
    {
      "epoch": 9.005145597006198,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 38497
    },
    {
      "epoch": 9.005379487779207,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 38498
    },
    {
      "epoch": 9.005613378552216,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 38499
    },
    {
      "epoch": 9.005847269325225,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 38500
    },
    {
      "epoch": 9.005847269325225,
      "eval_runtime": 4.6607,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 38500
    },
    {
      "epoch": 9.006081160098233,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 38501
    },
    {
      "epoch": 9.006315050871244,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7212,
      "step": 38502
    },
    {
      "epoch": 9.006548941644253,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 38503
    },
    {
      "epoch": 9.006782832417262,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5471,
      "step": 38504
    },
    {
      "epoch": 9.00701672319027,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4591,
      "step": 38505
    },
    {
      "epoch": 9.00725061396328,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 38506
    },
    {
      "epoch": 9.007484504736288,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 38507
    },
    {
      "epoch": 9.007718395509297,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 38508
    },
    {
      "epoch": 9.007952286282306,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4476,
      "step": 38509
    },
    {
      "epoch": 9.008186177055315,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6138,
      "step": 38510
    },
    {
      "epoch": 9.008420067828323,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 38511
    },
    {
      "epoch": 9.008653958601332,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 38512
    },
    {
      "epoch": 9.008887849374343,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6,
      "step": 38513
    },
    {
      "epoch": 9.009121740147352,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6216,
      "step": 38514
    },
    {
      "epoch": 9.00935563092036,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.6401,
      "step": 38515
    },
    {
      "epoch": 9.00958952169337,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6372,
      "step": 38516
    },
    {
      "epoch": 9.009823412466378,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 38517
    },
    {
      "epoch": 9.010057303239387,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.5482,
      "step": 38518
    },
    {
      "epoch": 9.010291194012396,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 38519
    },
    {
      "epoch": 9.010525084785405,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 38520
    },
    {
      "epoch": 9.010758975558414,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6908,
      "step": 38521
    },
    {
      "epoch": 9.010992866331422,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7781,
      "step": 38522
    },
    {
      "epoch": 9.011226757104433,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 38523
    },
    {
      "epoch": 9.011460647877442,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7019,
      "step": 38524
    },
    {
      "epoch": 9.01169453865045,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5635,
      "step": 38525
    },
    {
      "epoch": 9.01192842942346,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 38526
    },
    {
      "epoch": 9.012162320196468,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 38527
    },
    {
      "epoch": 9.012396210969477,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7594,
      "step": 38528
    },
    {
      "epoch": 9.012630101742486,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 38529
    },
    {
      "epoch": 9.012863992515495,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5451,
      "step": 38530
    },
    {
      "epoch": 9.013097883288504,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 38531
    },
    {
      "epoch": 9.013331774061513,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8409,
      "step": 38532
    },
    {
      "epoch": 9.013565664834521,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7573,
      "step": 38533
    },
    {
      "epoch": 9.013799555607532,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 38534
    },
    {
      "epoch": 9.014033446380541,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.47,
      "step": 38535
    },
    {
      "epoch": 9.01426733715355,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 38536
    },
    {
      "epoch": 9.014501227926559,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5175,
      "step": 38537
    },
    {
      "epoch": 9.014735118699567,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7644,
      "step": 38538
    },
    {
      "epoch": 9.014969009472576,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6453,
      "step": 38539
    },
    {
      "epoch": 9.015202900245585,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 38540
    },
    {
      "epoch": 9.015436791018594,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5137,
      "step": 38541
    },
    {
      "epoch": 9.015670681791603,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.553,
      "step": 38542
    },
    {
      "epoch": 9.015904572564612,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4659,
      "step": 38543
    },
    {
      "epoch": 9.01613846333762,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 38544
    },
    {
      "epoch": 9.016372354110631,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.7384,
      "step": 38545
    },
    {
      "epoch": 9.01660624488364,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9087,
      "step": 38546
    },
    {
      "epoch": 9.016840135656649,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 38547
    },
    {
      "epoch": 9.017074026429658,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8846,
      "step": 38548
    },
    {
      "epoch": 9.017307917202666,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 38549
    },
    {
      "epoch": 9.017541807975675,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.3447,
      "step": 38550
    },
    {
      "epoch": 9.017775698748684,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 38551
    },
    {
      "epoch": 9.018009589521693,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 38552
    },
    {
      "epoch": 9.018243480294702,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.688,
      "step": 38553
    },
    {
      "epoch": 9.01847737106771,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 38554
    },
    {
      "epoch": 9.018711261840721,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 38555
    },
    {
      "epoch": 9.01894515261373,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 38556
    },
    {
      "epoch": 9.019179043386739,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5529,
      "step": 38557
    },
    {
      "epoch": 9.019412934159748,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8416,
      "step": 38558
    },
    {
      "epoch": 9.019646824932757,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 38559
    },
    {
      "epoch": 9.019880715705765,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 38560
    },
    {
      "epoch": 9.020114606478774,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 38561
    },
    {
      "epoch": 9.020348497251783,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 38562
    },
    {
      "epoch": 9.020582388024792,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5946,
      "step": 38563
    },
    {
      "epoch": 9.0208162787978,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5192,
      "step": 38564
    },
    {
      "epoch": 9.02105016957081,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6483,
      "step": 38565
    },
    {
      "epoch": 9.02128406034382,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6506,
      "step": 38566
    },
    {
      "epoch": 9.021517951116829,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 38567
    },
    {
      "epoch": 9.021751841889838,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 38568
    },
    {
      "epoch": 9.021985732662847,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.4451,
      "step": 38569
    },
    {
      "epoch": 9.022219623435856,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 38570
    },
    {
      "epoch": 9.022453514208864,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 38571
    },
    {
      "epoch": 9.022687404981873,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 38572
    },
    {
      "epoch": 9.022921295754882,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 38573
    },
    {
      "epoch": 9.023155186527891,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 38574
    },
    {
      "epoch": 9.0233890773009,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 38575
    },
    {
      "epoch": 9.023622968073909,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4636,
      "step": 38576
    },
    {
      "epoch": 9.02385685884692,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6653,
      "step": 38577
    },
    {
      "epoch": 9.024090749619928,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 38578
    },
    {
      "epoch": 9.024324640392937,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 38579
    },
    {
      "epoch": 9.024558531165946,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.72,
      "step": 38580
    },
    {
      "epoch": 9.024792421938955,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7653,
      "step": 38581
    },
    {
      "epoch": 9.025026312711963,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6558,
      "step": 38582
    },
    {
      "epoch": 9.025260203484972,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6836,
      "step": 38583
    },
    {
      "epoch": 9.025494094257981,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.5212,
      "step": 38584
    },
    {
      "epoch": 9.02572798503099,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4646,
      "step": 38585
    },
    {
      "epoch": 9.025961875803999,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7463,
      "step": 38586
    },
    {
      "epoch": 9.02619576657701,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5498,
      "step": 38587
    },
    {
      "epoch": 9.026429657350018,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 38588
    },
    {
      "epoch": 9.026663548123027,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 38589
    },
    {
      "epoch": 9.026897438896036,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7191,
      "step": 38590
    },
    {
      "epoch": 9.027131329669045,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 38591
    },
    {
      "epoch": 9.027365220442054,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5729,
      "step": 38592
    },
    {
      "epoch": 9.027599111215062,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.7321,
      "step": 38593
    },
    {
      "epoch": 9.027833001988071,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.4355,
      "step": 38594
    },
    {
      "epoch": 9.02806689276108,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 38595
    },
    {
      "epoch": 9.028300783534089,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6763,
      "step": 38596
    },
    {
      "epoch": 9.028534674307098,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6507,
      "step": 38597
    },
    {
      "epoch": 9.028768565080108,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.699,
      "step": 38598
    },
    {
      "epoch": 9.029002455853117,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 38599
    },
    {
      "epoch": 9.029236346626126,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 38600
    },
    {
      "epoch": 9.029236346626126,
      "eval_runtime": 4.647,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 38600
    },
    {
      "epoch": 9.029470237399135,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 38601
    },
    {
      "epoch": 9.029704128172144,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6455,
      "step": 38602
    },
    {
      "epoch": 9.029938018945153,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 38603
    },
    {
      "epoch": 9.030171909718161,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 38604
    },
    {
      "epoch": 9.03040580049117,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6891,
      "step": 38605
    },
    {
      "epoch": 9.030639691264179,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.3911,
      "step": 38606
    },
    {
      "epoch": 9.030873582037188,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5385,
      "step": 38607
    },
    {
      "epoch": 9.031107472810197,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5433,
      "step": 38608
    },
    {
      "epoch": 9.031341363583207,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 38609
    },
    {
      "epoch": 9.031575254356216,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6078,
      "step": 38610
    },
    {
      "epoch": 9.031809145129225,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 38611
    },
    {
      "epoch": 9.032043035902234,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8359,
      "step": 38612
    },
    {
      "epoch": 9.032276926675243,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7426,
      "step": 38613
    },
    {
      "epoch": 9.032510817448252,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 38614
    },
    {
      "epoch": 9.03274470822126,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 38615
    },
    {
      "epoch": 9.03297859899427,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6039,
      "step": 38616
    },
    {
      "epoch": 9.033212489767278,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.497,
      "step": 38617
    },
    {
      "epoch": 9.033446380540287,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5201,
      "step": 38618
    },
    {
      "epoch": 9.033680271313298,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5823,
      "step": 38619
    },
    {
      "epoch": 9.033914162086306,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 38620
    },
    {
      "epoch": 9.034148052859315,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 38621
    },
    {
      "epoch": 9.034381943632324,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 38622
    },
    {
      "epoch": 9.034615834405333,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 38623
    },
    {
      "epoch": 9.034849725178342,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6416,
      "step": 38624
    },
    {
      "epoch": 9.03508361595135,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5809,
      "step": 38625
    },
    {
      "epoch": 9.03531750672436,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 38626
    },
    {
      "epoch": 9.035551397497368,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 38627
    },
    {
      "epoch": 9.035785288270377,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6447,
      "step": 38628
    },
    {
      "epoch": 9.036019179043386,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 38629
    },
    {
      "epoch": 9.036253069816397,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5525,
      "step": 38630
    },
    {
      "epoch": 9.036486960589405,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5881,
      "step": 38631
    },
    {
      "epoch": 9.036720851362414,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.517,
      "step": 38632
    },
    {
      "epoch": 9.036954742135423,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5416,
      "step": 38633
    },
    {
      "epoch": 9.037188632908432,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.5514,
      "step": 38634
    },
    {
      "epoch": 9.03742252368144,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5528,
      "step": 38635
    },
    {
      "epoch": 9.03765641445445,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 38636
    },
    {
      "epoch": 9.037890305227458,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4065,
      "step": 38637
    },
    {
      "epoch": 9.038124196000467,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5415,
      "step": 38638
    },
    {
      "epoch": 9.038358086773476,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.4555,
      "step": 38639
    },
    {
      "epoch": 9.038591977546485,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 38640
    },
    {
      "epoch": 9.038825868319496,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5956,
      "step": 38641
    },
    {
      "epoch": 9.039059759092504,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 38642
    },
    {
      "epoch": 9.039293649865513,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 38643
    },
    {
      "epoch": 9.039527540638522,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 38644
    },
    {
      "epoch": 9.03976143141153,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7718,
      "step": 38645
    },
    {
      "epoch": 9.03999532218454,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 38646
    },
    {
      "epoch": 9.040229212957549,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 38647
    },
    {
      "epoch": 9.040463103730557,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 38648
    },
    {
      "epoch": 9.040696994503566,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 38649
    },
    {
      "epoch": 9.040930885276575,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 38650
    },
    {
      "epoch": 9.041164776049586,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5215,
      "step": 38651
    },
    {
      "epoch": 9.041398666822595,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6672,
      "step": 38652
    },
    {
      "epoch": 9.041632557595603,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 38653
    },
    {
      "epoch": 9.041866448368612,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 38654
    },
    {
      "epoch": 9.042100339141621,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6857,
      "step": 38655
    },
    {
      "epoch": 9.04233422991463,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7907,
      "step": 38656
    },
    {
      "epoch": 9.042568120687639,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 38657
    },
    {
      "epoch": 9.042802011460648,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 38658
    },
    {
      "epoch": 9.043035902233656,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 1.5879,
      "step": 38659
    },
    {
      "epoch": 9.043269793006665,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8233,
      "step": 38660
    },
    {
      "epoch": 9.043503683779674,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 38661
    },
    {
      "epoch": 9.043737574552685,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5745,
      "step": 38662
    },
    {
      "epoch": 9.043971465325694,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8101,
      "step": 38663
    },
    {
      "epoch": 9.044205356098702,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6817,
      "step": 38664
    },
    {
      "epoch": 9.044439246871711,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7813,
      "step": 38665
    },
    {
      "epoch": 9.04467313764472,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5956,
      "step": 38666
    },
    {
      "epoch": 9.044907028417729,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 38667
    },
    {
      "epoch": 9.045140919190738,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 38668
    },
    {
      "epoch": 9.045374809963747,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.7758,
      "step": 38669
    },
    {
      "epoch": 9.045608700736755,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 38670
    },
    {
      "epoch": 9.045842591509764,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.701,
      "step": 38671
    },
    {
      "epoch": 9.046076482282775,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 38672
    },
    {
      "epoch": 9.046310373055784,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.4708,
      "step": 38673
    },
    {
      "epoch": 9.046544263828793,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 38674
    },
    {
      "epoch": 9.046778154601801,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 38675
    },
    {
      "epoch": 9.04701204537481,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5883,
      "step": 38676
    },
    {
      "epoch": 9.047245936147819,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 38677
    },
    {
      "epoch": 9.047479826920828,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 38678
    },
    {
      "epoch": 9.047713717693837,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5615,
      "step": 38679
    },
    {
      "epoch": 9.047947608466846,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 38680
    },
    {
      "epoch": 9.048181499239854,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 38681
    },
    {
      "epoch": 9.048415390012863,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5871,
      "step": 38682
    },
    {
      "epoch": 9.048649280785874,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7511,
      "step": 38683
    },
    {
      "epoch": 9.048883171558883,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6589,
      "step": 38684
    },
    {
      "epoch": 9.049117062331892,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 38685
    },
    {
      "epoch": 9.0493509531049,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 38686
    },
    {
      "epoch": 9.04958484387791,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5542,
      "step": 38687
    },
    {
      "epoch": 9.049818734650918,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.572,
      "step": 38688
    },
    {
      "epoch": 9.050052625423927,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 38689
    },
    {
      "epoch": 9.050286516196936,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6643,
      "step": 38690
    },
    {
      "epoch": 9.050520406969945,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5077,
      "step": 38691
    },
    {
      "epoch": 9.050754297742953,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5544,
      "step": 38692
    },
    {
      "epoch": 9.050988188515962,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4787,
      "step": 38693
    },
    {
      "epoch": 9.051222079288973,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 38694
    },
    {
      "epoch": 9.051455970061982,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5871,
      "step": 38695
    },
    {
      "epoch": 9.05168986083499,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5168,
      "step": 38696
    },
    {
      "epoch": 9.051923751608,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 38697
    },
    {
      "epoch": 9.052157642381008,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7224,
      "step": 38698
    },
    {
      "epoch": 9.052391533154017,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 38699
    },
    {
      "epoch": 9.052625423927026,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 38700
    },
    {
      "epoch": 9.052625423927026,
      "eval_runtime": 4.6027,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 38700
    },
    {
      "epoch": 9.052859314700035,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5343,
      "step": 38701
    },
    {
      "epoch": 9.053093205473044,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5555,
      "step": 38702
    },
    {
      "epoch": 9.053327096246052,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6553,
      "step": 38703
    },
    {
      "epoch": 9.053560987019063,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5515,
      "step": 38704
    },
    {
      "epoch": 9.053794877792072,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.4946,
      "step": 38705
    },
    {
      "epoch": 9.05402876856508,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 38706
    },
    {
      "epoch": 9.05426265933809,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 38707
    },
    {
      "epoch": 9.054496550111098,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6425,
      "step": 38708
    },
    {
      "epoch": 9.054730440884107,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.61,
      "step": 38709
    },
    {
      "epoch": 9.054964331657116,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 38710
    },
    {
      "epoch": 9.055198222430125,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.574,
      "step": 38711
    },
    {
      "epoch": 9.055432113203134,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.838,
      "step": 38712
    },
    {
      "epoch": 9.055666003976143,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6072,
      "step": 38713
    },
    {
      "epoch": 9.055899894749151,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5437,
      "step": 38714
    },
    {
      "epoch": 9.056133785522162,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6237,
      "step": 38715
    },
    {
      "epoch": 9.05636767629517,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7482,
      "step": 38716
    },
    {
      "epoch": 9.05660156706818,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5517,
      "step": 38717
    },
    {
      "epoch": 9.056835457841188,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 38718
    },
    {
      "epoch": 9.057069348614197,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7201,
      "step": 38719
    },
    {
      "epoch": 9.057303239387206,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.552,
      "step": 38720
    },
    {
      "epoch": 9.057537130160215,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.689,
      "step": 38721
    },
    {
      "epoch": 9.057771020933224,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 38722
    },
    {
      "epoch": 9.058004911706233,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5934,
      "step": 38723
    },
    {
      "epoch": 9.058238802479242,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.511,
      "step": 38724
    },
    {
      "epoch": 9.05847269325225,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 38725
    },
    {
      "epoch": 9.058706584025261,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 38726
    },
    {
      "epoch": 9.05894047479827,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4571,
      "step": 38727
    },
    {
      "epoch": 9.059174365571279,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 38728
    },
    {
      "epoch": 9.059408256344287,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 38729
    },
    {
      "epoch": 9.059642147117296,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8251,
      "step": 38730
    },
    {
      "epoch": 9.059876037890305,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5653,
      "step": 38731
    },
    {
      "epoch": 9.060109928663314,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6705,
      "step": 38732
    },
    {
      "epoch": 9.060343819436323,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6395,
      "step": 38733
    },
    {
      "epoch": 9.060577710209332,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 38734
    },
    {
      "epoch": 9.06081160098234,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.728,
      "step": 38735
    },
    {
      "epoch": 9.061045491755351,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 38736
    },
    {
      "epoch": 9.06127938252836,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 38737
    },
    {
      "epoch": 9.061513273301369,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 38738
    },
    {
      "epoch": 9.061747164074378,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 38739
    },
    {
      "epoch": 9.061981054847386,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 38740
    },
    {
      "epoch": 9.062214945620395,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 38741
    },
    {
      "epoch": 9.062448836393404,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 38742
    },
    {
      "epoch": 9.062682727166413,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.473,
      "step": 38743
    },
    {
      "epoch": 9.062916617939422,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 38744
    },
    {
      "epoch": 9.06315050871243,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 38745
    },
    {
      "epoch": 9.06338439948544,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 38746
    },
    {
      "epoch": 9.06361829025845,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7407,
      "step": 38747
    },
    {
      "epoch": 9.063852181031459,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 38748
    },
    {
      "epoch": 9.064086071804468,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 38749
    },
    {
      "epoch": 9.064319962577477,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5806,
      "step": 38750
    },
    {
      "epoch": 9.064553853350485,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 38751
    },
    {
      "epoch": 9.064787744123494,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 38752
    },
    {
      "epoch": 9.065021634896503,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5394,
      "step": 38753
    },
    {
      "epoch": 9.065255525669512,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 38754
    },
    {
      "epoch": 9.06548941644252,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7751,
      "step": 38755
    },
    {
      "epoch": 9.06572330721553,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2347,
      "step": 38756
    },
    {
      "epoch": 9.065957197988538,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 38757
    },
    {
      "epoch": 9.066191088761549,
      "grad_norm": 9.75,
      "learning_rate": 3e-05,
      "loss": 1.5314,
      "step": 38758
    },
    {
      "epoch": 9.066424979534558,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.718,
      "step": 38759
    },
    {
      "epoch": 9.066658870307567,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6377,
      "step": 38760
    },
    {
      "epoch": 9.066892761080576,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 38761
    },
    {
      "epoch": 9.067126651853584,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7922,
      "step": 38762
    },
    {
      "epoch": 9.067360542626593,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 38763
    },
    {
      "epoch": 9.067594433399602,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.4443,
      "step": 38764
    },
    {
      "epoch": 9.067828324172611,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7512,
      "step": 38765
    },
    {
      "epoch": 9.06806221494562,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 38766
    },
    {
      "epoch": 9.068296105718629,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 38767
    },
    {
      "epoch": 9.06852999649164,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 38768
    },
    {
      "epoch": 9.068763887264648,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.678,
      "step": 38769
    },
    {
      "epoch": 9.068997778037657,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6383,
      "step": 38770
    },
    {
      "epoch": 9.069231668810666,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 38771
    },
    {
      "epoch": 9.069465559583675,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 38772
    },
    {
      "epoch": 9.069699450356683,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5991,
      "step": 38773
    },
    {
      "epoch": 9.069933341129692,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5515,
      "step": 38774
    },
    {
      "epoch": 9.070167231902701,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 38775
    },
    {
      "epoch": 9.07040112267571,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 38776
    },
    {
      "epoch": 9.070635013448719,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 38777
    },
    {
      "epoch": 9.070868904221728,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 38778
    },
    {
      "epoch": 9.071102794994738,
      "grad_norm": 8.4375,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 38779
    },
    {
      "epoch": 9.071336685767747,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 38780
    },
    {
      "epoch": 9.071570576540756,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6218,
      "step": 38781
    },
    {
      "epoch": 9.071804467313765,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7644,
      "step": 38782
    },
    {
      "epoch": 9.072038358086774,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4274,
      "step": 38783
    },
    {
      "epoch": 9.072272248859782,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 38784
    },
    {
      "epoch": 9.072506139632791,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 38785
    },
    {
      "epoch": 9.0727400304058,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6365,
      "step": 38786
    },
    {
      "epoch": 9.072973921178809,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7883,
      "step": 38787
    },
    {
      "epoch": 9.073207811951818,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.7162,
      "step": 38788
    },
    {
      "epoch": 9.073441702724827,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.587,
      "step": 38789
    },
    {
      "epoch": 9.073675593497837,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5228,
      "step": 38790
    },
    {
      "epoch": 9.073909484270846,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 38791
    },
    {
      "epoch": 9.074143375043855,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 38792
    },
    {
      "epoch": 9.074377265816864,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7318,
      "step": 38793
    },
    {
      "epoch": 9.074611156589873,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.729,
      "step": 38794
    },
    {
      "epoch": 9.074845047362881,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5839,
      "step": 38795
    },
    {
      "epoch": 9.07507893813589,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 38796
    },
    {
      "epoch": 9.0753128289089,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 38797
    },
    {
      "epoch": 9.075546719681908,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 38798
    },
    {
      "epoch": 9.075780610454917,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 38799
    },
    {
      "epoch": 9.076014501227927,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.441,
      "step": 38800
    },
    {
      "epoch": 9.076014501227927,
      "eval_runtime": 4.6026,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 38800
    },
    {
      "epoch": 9.076248392000936,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8838,
      "step": 38801
    },
    {
      "epoch": 9.076482282773945,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 38802
    },
    {
      "epoch": 9.076716173546954,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 38803
    },
    {
      "epoch": 9.076950064319963,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 38804
    },
    {
      "epoch": 9.077183955092972,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6339,
      "step": 38805
    },
    {
      "epoch": 9.07741784586598,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.7081,
      "step": 38806
    },
    {
      "epoch": 9.07765173663899,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 38807
    },
    {
      "epoch": 9.077885627411998,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6452,
      "step": 38808
    },
    {
      "epoch": 9.078119518185007,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7423,
      "step": 38809
    },
    {
      "epoch": 9.078353408958016,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 38810
    },
    {
      "epoch": 9.078587299731026,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 38811
    },
    {
      "epoch": 9.078821190504035,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8508,
      "step": 38812
    },
    {
      "epoch": 9.079055081277044,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7846,
      "step": 38813
    },
    {
      "epoch": 9.079288972050053,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6149,
      "step": 38814
    },
    {
      "epoch": 9.079522862823062,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 1.5751,
      "step": 38815
    },
    {
      "epoch": 9.07975675359607,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6823,
      "step": 38816
    },
    {
      "epoch": 9.07999064436908,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 38817
    },
    {
      "epoch": 9.080224535142088,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 38818
    },
    {
      "epoch": 9.080458425915097,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6409,
      "step": 38819
    },
    {
      "epoch": 9.080692316688106,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6031,
      "step": 38820
    },
    {
      "epoch": 9.080926207461115,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.597,
      "step": 38821
    },
    {
      "epoch": 9.081160098234125,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 38822
    },
    {
      "epoch": 9.081393989007134,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 38823
    },
    {
      "epoch": 9.081627879780143,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 38824
    },
    {
      "epoch": 9.081861770553152,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.3506,
      "step": 38825
    },
    {
      "epoch": 9.08209566132616,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 38826
    },
    {
      "epoch": 9.08232955209917,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.536,
      "step": 38827
    },
    {
      "epoch": 9.082563442872178,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 38828
    },
    {
      "epoch": 9.082797333645187,
      "grad_norm": 2.546875,
      "learning_rate": 3e-05,
      "loss": 1.5739,
      "step": 38829
    },
    {
      "epoch": 9.083031224418196,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 38830
    },
    {
      "epoch": 9.083265115191205,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5608,
      "step": 38831
    },
    {
      "epoch": 9.083499005964216,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5586,
      "step": 38832
    },
    {
      "epoch": 9.083732896737224,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6225,
      "step": 38833
    },
    {
      "epoch": 9.083966787510233,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8707,
      "step": 38834
    },
    {
      "epoch": 9.084200678283242,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 38835
    },
    {
      "epoch": 9.084434569056251,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5021,
      "step": 38836
    },
    {
      "epoch": 9.08466845982926,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 38837
    },
    {
      "epoch": 9.084902350602269,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6565,
      "step": 38838
    },
    {
      "epoch": 9.085136241375277,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5503,
      "step": 38839
    },
    {
      "epoch": 9.085370132148286,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 38840
    },
    {
      "epoch": 9.085604022921295,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 38841
    },
    {
      "epoch": 9.085837913694304,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6389,
      "step": 38842
    },
    {
      "epoch": 9.086071804467315,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7664,
      "step": 38843
    },
    {
      "epoch": 9.086305695240323,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6583,
      "step": 38844
    },
    {
      "epoch": 9.086539586013332,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6197,
      "step": 38845
    },
    {
      "epoch": 9.086773476786341,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 38846
    },
    {
      "epoch": 9.08700736755935,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5764,
      "step": 38847
    },
    {
      "epoch": 9.087241258332359,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 38848
    },
    {
      "epoch": 9.087475149105368,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 38849
    },
    {
      "epoch": 9.087709039878376,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.3415,
      "step": 38850
    },
    {
      "epoch": 9.087942930651385,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 38851
    },
    {
      "epoch": 9.088176821424394,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5625,
      "step": 38852
    },
    {
      "epoch": 9.088410712197405,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.484,
      "step": 38853
    },
    {
      "epoch": 9.088644602970414,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 38854
    },
    {
      "epoch": 9.088878493743422,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 38855
    },
    {
      "epoch": 9.089112384516431,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 38856
    },
    {
      "epoch": 9.08934627528944,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 38857
    },
    {
      "epoch": 9.089580166062449,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6363,
      "step": 38858
    },
    {
      "epoch": 9.089814056835458,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 38859
    },
    {
      "epoch": 9.090047947608467,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 1.5577,
      "step": 38860
    },
    {
      "epoch": 9.090281838381475,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 38861
    },
    {
      "epoch": 9.090515729154484,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 38862
    },
    {
      "epoch": 9.090749619927493,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 38863
    },
    {
      "epoch": 9.090983510700504,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5788,
      "step": 38864
    },
    {
      "epoch": 9.091217401473513,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 38865
    },
    {
      "epoch": 9.091451292246521,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 38866
    },
    {
      "epoch": 9.09168518301953,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.646,
      "step": 38867
    },
    {
      "epoch": 9.091919073792539,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 38868
    },
    {
      "epoch": 9.092152964565548,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 38869
    },
    {
      "epoch": 9.092386855338557,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.3778,
      "step": 38870
    },
    {
      "epoch": 9.092620746111566,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 38871
    },
    {
      "epoch": 9.092854636884574,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5877,
      "step": 38872
    },
    {
      "epoch": 9.093088527657583,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 38873
    },
    {
      "epoch": 9.093322418430592,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.554,
      "step": 38874
    },
    {
      "epoch": 9.093556309203603,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5761,
      "step": 38875
    },
    {
      "epoch": 9.093790199976612,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 38876
    },
    {
      "epoch": 9.09402409074962,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 38877
    },
    {
      "epoch": 9.09425798152263,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 38878
    },
    {
      "epoch": 9.094491872295638,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 38879
    },
    {
      "epoch": 9.094725763068647,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.6689,
      "step": 38880
    },
    {
      "epoch": 9.094959653841656,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 38881
    },
    {
      "epoch": 9.095193544614665,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5676,
      "step": 38882
    },
    {
      "epoch": 9.095427435387673,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 38883
    },
    {
      "epoch": 9.095661326160682,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 38884
    },
    {
      "epoch": 9.095895216933691,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6633,
      "step": 38885
    },
    {
      "epoch": 9.096129107706702,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6186,
      "step": 38886
    },
    {
      "epoch": 9.09636299847971,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6512,
      "step": 38887
    },
    {
      "epoch": 9.09659688925272,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 38888
    },
    {
      "epoch": 9.096830780025728,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 38889
    },
    {
      "epoch": 9.097064670798737,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5517,
      "step": 38890
    },
    {
      "epoch": 9.097298561571746,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 38891
    },
    {
      "epoch": 9.097532452344755,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 38892
    },
    {
      "epoch": 9.097766343117764,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5518,
      "step": 38893
    },
    {
      "epoch": 9.098000233890772,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 38894
    },
    {
      "epoch": 9.098234124663781,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.5048,
      "step": 38895
    },
    {
      "epoch": 9.098468015436792,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5606,
      "step": 38896
    },
    {
      "epoch": 9.0987019062098,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 38897
    },
    {
      "epoch": 9.09893579698281,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5398,
      "step": 38898
    },
    {
      "epoch": 9.099169687755818,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 38899
    },
    {
      "epoch": 9.099403578528827,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7615,
      "step": 38900
    },
    {
      "epoch": 9.099403578528827,
      "eval_runtime": 4.6888,
      "eval_samples_per_second": 0.213,
      "eval_steps_per_second": 0.213,
      "step": 38900
    },
    {
      "epoch": 9.099637469301836,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.2163,
      "step": 38901
    },
    {
      "epoch": 9.099871360074845,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4797,
      "step": 38902
    },
    {
      "epoch": 9.100105250847854,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5293,
      "step": 38903
    },
    {
      "epoch": 9.100339141620863,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5671,
      "step": 38904
    },
    {
      "epoch": 9.100573032393871,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4634,
      "step": 38905
    },
    {
      "epoch": 9.10080692316688,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6553,
      "step": 38906
    },
    {
      "epoch": 9.10104081393989,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 38907
    },
    {
      "epoch": 9.1012747047129,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 38908
    },
    {
      "epoch": 9.101508595485909,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7282,
      "step": 38909
    },
    {
      "epoch": 9.101742486258917,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.5291,
      "step": 38910
    },
    {
      "epoch": 9.101976377031926,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 38911
    },
    {
      "epoch": 9.102210267804935,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.499,
      "step": 38912
    },
    {
      "epoch": 9.102444158577944,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7727,
      "step": 38913
    },
    {
      "epoch": 9.102678049350953,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7821,
      "step": 38914
    },
    {
      "epoch": 9.102911940123962,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4885,
      "step": 38915
    },
    {
      "epoch": 9.10314583089697,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 38916
    },
    {
      "epoch": 9.103379721669981,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 38917
    },
    {
      "epoch": 9.10361361244299,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6958,
      "step": 38918
    },
    {
      "epoch": 9.103847503215999,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 38919
    },
    {
      "epoch": 9.104081393989008,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 38920
    },
    {
      "epoch": 9.104315284762016,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6682,
      "step": 38921
    },
    {
      "epoch": 9.104549175535025,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5405,
      "step": 38922
    },
    {
      "epoch": 9.104783066308034,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7155,
      "step": 38923
    },
    {
      "epoch": 9.105016957081043,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6286,
      "step": 38924
    },
    {
      "epoch": 9.105250847854052,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 38925
    },
    {
      "epoch": 9.10548473862706,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5978,
      "step": 38926
    },
    {
      "epoch": 9.10571862940007,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 38927
    },
    {
      "epoch": 9.10595252017308,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5779,
      "step": 38928
    },
    {
      "epoch": 9.106186410946089,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 38929
    },
    {
      "epoch": 9.106420301719098,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5879,
      "step": 38930
    },
    {
      "epoch": 9.106654192492107,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6739,
      "step": 38931
    },
    {
      "epoch": 9.106888083265115,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6387,
      "step": 38932
    },
    {
      "epoch": 9.107121974038124,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.487,
      "step": 38933
    },
    {
      "epoch": 9.107355864811133,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 38934
    },
    {
      "epoch": 9.107589755584142,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 38935
    },
    {
      "epoch": 9.10782364635715,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 38936
    },
    {
      "epoch": 9.10805753713016,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 38937
    },
    {
      "epoch": 9.108291427903168,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 38938
    },
    {
      "epoch": 9.108525318676179,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 38939
    },
    {
      "epoch": 9.108759209449188,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5464,
      "step": 38940
    },
    {
      "epoch": 9.108993100222197,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6481,
      "step": 38941
    },
    {
      "epoch": 9.109226990995205,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 38942
    },
    {
      "epoch": 9.109460881768214,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7732,
      "step": 38943
    },
    {
      "epoch": 9.109694772541223,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6965,
      "step": 38944
    },
    {
      "epoch": 9.109928663314232,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 38945
    },
    {
      "epoch": 9.11016255408724,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 38946
    },
    {
      "epoch": 9.11039644486025,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 38947
    },
    {
      "epoch": 9.110630335633259,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 38948
    },
    {
      "epoch": 9.11086422640627,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5437,
      "step": 38949
    },
    {
      "epoch": 9.111098117179278,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 38950
    },
    {
      "epoch": 9.111332007952287,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6913,
      "step": 38951
    },
    {
      "epoch": 9.111565898725296,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0932,
      "step": 38952
    },
    {
      "epoch": 9.111799789498304,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5372,
      "step": 38953
    },
    {
      "epoch": 9.112033680271313,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 38954
    },
    {
      "epoch": 9.112267571044322,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7296,
      "step": 38955
    },
    {
      "epoch": 9.112501461817331,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6496,
      "step": 38956
    },
    {
      "epoch": 9.11273535259034,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 38957
    },
    {
      "epoch": 9.112969243363349,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4455,
      "step": 38958
    },
    {
      "epoch": 9.113203134136358,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 38959
    },
    {
      "epoch": 9.113437024909368,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5938,
      "step": 38960
    },
    {
      "epoch": 9.113670915682377,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6186,
      "step": 38961
    },
    {
      "epoch": 9.113904806455386,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 38962
    },
    {
      "epoch": 9.114138697228395,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5953,
      "step": 38963
    },
    {
      "epoch": 9.114372588001403,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 38964
    },
    {
      "epoch": 9.114606478774412,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.621,
      "step": 38965
    },
    {
      "epoch": 9.114840369547421,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 38966
    },
    {
      "epoch": 9.11507426032043,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.6314,
      "step": 38967
    },
    {
      "epoch": 9.115308151093439,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 38968
    },
    {
      "epoch": 9.115542041866448,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 38969
    },
    {
      "epoch": 9.115775932639457,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9094,
      "step": 38970
    },
    {
      "epoch": 9.116009823412467,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 38971
    },
    {
      "epoch": 9.116243714185476,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5616,
      "step": 38972
    },
    {
      "epoch": 9.116477604958485,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4245,
      "step": 38973
    },
    {
      "epoch": 9.116711495731494,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6288,
      "step": 38974
    },
    {
      "epoch": 9.116945386504502,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 38975
    },
    {
      "epoch": 9.117179277277511,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.4903,
      "step": 38976
    },
    {
      "epoch": 9.11741316805052,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5743,
      "step": 38977
    },
    {
      "epoch": 9.117647058823529,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6856,
      "step": 38978
    },
    {
      "epoch": 9.117880949596538,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 38979
    },
    {
      "epoch": 9.118114840369547,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6824,
      "step": 38980
    },
    {
      "epoch": 9.118348731142557,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 38981
    },
    {
      "epoch": 9.118582621915566,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 38982
    },
    {
      "epoch": 9.118816512688575,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 38983
    },
    {
      "epoch": 9.119050403461584,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7662,
      "step": 38984
    },
    {
      "epoch": 9.119284294234593,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.618,
      "step": 38985
    },
    {
      "epoch": 9.119518185007601,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7265,
      "step": 38986
    },
    {
      "epoch": 9.11975207578061,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 38987
    },
    {
      "epoch": 9.11998596655362,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8817,
      "step": 38988
    },
    {
      "epoch": 9.120219857326628,
      "grad_norm": 7.59375,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 38989
    },
    {
      "epoch": 9.120453748099637,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.3879,
      "step": 38990
    },
    {
      "epoch": 9.120687638872646,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 38991
    },
    {
      "epoch": 9.120921529645656,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 38992
    },
    {
      "epoch": 9.121155420418665,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5326,
      "step": 38993
    },
    {
      "epoch": 9.121389311191674,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5383,
      "step": 38994
    },
    {
      "epoch": 9.121623201964683,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 38995
    },
    {
      "epoch": 9.121857092737692,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 38996
    },
    {
      "epoch": 9.1220909835107,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 38997
    },
    {
      "epoch": 9.12232487428371,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7444,
      "step": 38998
    },
    {
      "epoch": 9.122558765056718,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7587,
      "step": 38999
    },
    {
      "epoch": 9.122792655829727,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 39000
    },
    {
      "epoch": 9.122792655829727,
      "eval_runtime": 4.6293,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 39000
    },
    {
      "epoch": 9.123026546602736,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7329,
      "step": 39001
    },
    {
      "epoch": 9.123260437375745,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 39002
    },
    {
      "epoch": 9.123494328148755,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6118,
      "step": 39003
    },
    {
      "epoch": 9.123728218921764,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.8095,
      "step": 39004
    },
    {
      "epoch": 9.123962109694773,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 39005
    },
    {
      "epoch": 9.124196000467782,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6897,
      "step": 39006
    },
    {
      "epoch": 9.12442989124079,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7253,
      "step": 39007
    },
    {
      "epoch": 9.1246637820138,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 39008
    },
    {
      "epoch": 9.124897672786808,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 39009
    },
    {
      "epoch": 9.125131563559817,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 39010
    },
    {
      "epoch": 9.125365454332826,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 39011
    },
    {
      "epoch": 9.125599345105835,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 39012
    },
    {
      "epoch": 9.125833235878845,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 39013
    },
    {
      "epoch": 9.126067126651854,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7045,
      "step": 39014
    },
    {
      "epoch": 9.126301017424863,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5232,
      "step": 39015
    },
    {
      "epoch": 9.126534908197872,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7398,
      "step": 39016
    },
    {
      "epoch": 9.12676879897088,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 39017
    },
    {
      "epoch": 9.12700268974389,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6033,
      "step": 39018
    },
    {
      "epoch": 9.127236580516898,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 39019
    },
    {
      "epoch": 9.127470471289907,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 39020
    },
    {
      "epoch": 9.127704362062916,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6078,
      "step": 39021
    },
    {
      "epoch": 9.127938252835925,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 39022
    },
    {
      "epoch": 9.128172143608934,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.747,
      "step": 39023
    },
    {
      "epoch": 9.128406034381944,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6093,
      "step": 39024
    },
    {
      "epoch": 9.128639925154953,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 39025
    },
    {
      "epoch": 9.128873815927962,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6962,
      "step": 39026
    },
    {
      "epoch": 9.129107706700971,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 39027
    },
    {
      "epoch": 9.12934159747398,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5243,
      "step": 39028
    },
    {
      "epoch": 9.129575488246989,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.692,
      "step": 39029
    },
    {
      "epoch": 9.129809379019997,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5163,
      "step": 39030
    },
    {
      "epoch": 9.130043269793006,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 39031
    },
    {
      "epoch": 9.130277160566015,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 39032
    },
    {
      "epoch": 9.130511051339024,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7503,
      "step": 39033
    },
    {
      "epoch": 9.130744942112035,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 39034
    },
    {
      "epoch": 9.130978832885043,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 39035
    },
    {
      "epoch": 9.131212723658052,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 39036
    },
    {
      "epoch": 9.131446614431061,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 39037
    },
    {
      "epoch": 9.13168050520407,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 39038
    },
    {
      "epoch": 9.131914395977079,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6599,
      "step": 39039
    },
    {
      "epoch": 9.132148286750088,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 39040
    },
    {
      "epoch": 9.132382177523096,
      "grad_norm": 10.875,
      "learning_rate": 3e-05,
      "loss": 1.7876,
      "step": 39041
    },
    {
      "epoch": 9.132616068296105,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.681,
      "step": 39042
    },
    {
      "epoch": 9.132849959069114,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 39043
    },
    {
      "epoch": 9.133083849842123,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5104,
      "step": 39044
    },
    {
      "epoch": 9.133317740615134,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.702,
      "step": 39045
    },
    {
      "epoch": 9.133551631388142,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 39046
    },
    {
      "epoch": 9.133785522161151,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7074,
      "step": 39047
    },
    {
      "epoch": 9.13401941293416,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 39048
    },
    {
      "epoch": 9.134253303707169,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7141,
      "step": 39049
    },
    {
      "epoch": 9.134487194480178,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 39050
    },
    {
      "epoch": 9.134721085253187,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6144,
      "step": 39051
    },
    {
      "epoch": 9.134954976026195,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 39052
    },
    {
      "epoch": 9.135188866799204,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 39053
    },
    {
      "epoch": 9.135422757572213,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 39054
    },
    {
      "epoch": 9.135656648345222,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 39055
    },
    {
      "epoch": 9.135890539118233,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7731,
      "step": 39056
    },
    {
      "epoch": 9.136124429891241,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5185,
      "step": 39057
    },
    {
      "epoch": 9.13635832066425,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8025,
      "step": 39058
    },
    {
      "epoch": 9.136592211437259,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7367,
      "step": 39059
    },
    {
      "epoch": 9.136826102210268,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7581,
      "step": 39060
    },
    {
      "epoch": 9.137059992983277,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 39061
    },
    {
      "epoch": 9.137293883756286,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.3779,
      "step": 39062
    },
    {
      "epoch": 9.137527774529294,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 39063
    },
    {
      "epoch": 9.137761665302303,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5083,
      "step": 39064
    },
    {
      "epoch": 9.137995556075312,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 39065
    },
    {
      "epoch": 9.138229446848321,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5737,
      "step": 39066
    },
    {
      "epoch": 9.138463337621332,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.6787,
      "step": 39067
    },
    {
      "epoch": 9.13869722839434,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 39068
    },
    {
      "epoch": 9.13893111916735,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 39069
    },
    {
      "epoch": 9.139165009940358,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6756,
      "step": 39070
    },
    {
      "epoch": 9.139398900713367,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 39071
    },
    {
      "epoch": 9.139632791486376,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 39072
    },
    {
      "epoch": 9.139866682259385,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 39073
    },
    {
      "epoch": 9.140100573032393,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7116,
      "step": 39074
    },
    {
      "epoch": 9.140334463805402,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 39075
    },
    {
      "epoch": 9.140568354578411,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 39076
    },
    {
      "epoch": 9.140802245351422,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.47,
      "step": 39077
    },
    {
      "epoch": 9.14103613612443,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 39078
    },
    {
      "epoch": 9.14127002689744,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5891,
      "step": 39079
    },
    {
      "epoch": 9.141503917670448,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 39080
    },
    {
      "epoch": 9.141737808443457,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.731,
      "step": 39081
    },
    {
      "epoch": 9.141971699216466,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5844,
      "step": 39082
    },
    {
      "epoch": 9.142205589989475,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7369,
      "step": 39083
    },
    {
      "epoch": 9.142439480762484,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 39084
    },
    {
      "epoch": 9.142673371535492,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 39085
    },
    {
      "epoch": 9.142907262308501,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 39086
    },
    {
      "epoch": 9.14314115308151,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 39087
    },
    {
      "epoch": 9.14337504385452,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 39088
    },
    {
      "epoch": 9.14360893462753,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 39089
    },
    {
      "epoch": 9.143842825400538,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 39090
    },
    {
      "epoch": 9.144076716173547,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 39091
    },
    {
      "epoch": 9.144310606946556,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 39092
    },
    {
      "epoch": 9.144544497719565,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7592,
      "step": 39093
    },
    {
      "epoch": 9.144778388492574,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4225,
      "step": 39094
    },
    {
      "epoch": 9.145012279265583,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9196,
      "step": 39095
    },
    {
      "epoch": 9.145246170038591,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 39096
    },
    {
      "epoch": 9.1454800608116,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5511,
      "step": 39097
    },
    {
      "epoch": 9.14571395158461,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 39098
    },
    {
      "epoch": 9.14594784235762,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 39099
    },
    {
      "epoch": 9.146181733130629,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 39100
    },
    {
      "epoch": 9.146181733130629,
      "eval_runtime": 4.6388,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 39100
    },
    {
      "epoch": 9.146415623903637,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.709,
      "step": 39101
    },
    {
      "epoch": 9.146649514676646,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 39102
    },
    {
      "epoch": 9.146883405449655,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 39103
    },
    {
      "epoch": 9.147117296222664,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 39104
    },
    {
      "epoch": 9.147351186995673,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5642,
      "step": 39105
    },
    {
      "epoch": 9.147585077768682,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5244,
      "step": 39106
    },
    {
      "epoch": 9.14781896854169,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6773,
      "step": 39107
    },
    {
      "epoch": 9.1480528593147,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 39108
    },
    {
      "epoch": 9.14828675008771,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5735,
      "step": 39109
    },
    {
      "epoch": 9.148520640860719,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 39110
    },
    {
      "epoch": 9.148754531633728,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6719,
      "step": 39111
    },
    {
      "epoch": 9.148988422406736,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 39112
    },
    {
      "epoch": 9.149222313179745,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 39113
    },
    {
      "epoch": 9.149456203952754,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6511,
      "step": 39114
    },
    {
      "epoch": 9.149690094725763,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5768,
      "step": 39115
    },
    {
      "epoch": 9.149923985498772,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 39116
    },
    {
      "epoch": 9.15015787627178,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7228,
      "step": 39117
    },
    {
      "epoch": 9.15039176704479,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 39118
    },
    {
      "epoch": 9.150625657817798,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 39119
    },
    {
      "epoch": 9.150859548590809,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 39120
    },
    {
      "epoch": 9.151093439363818,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 39121
    },
    {
      "epoch": 9.151327330136827,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 39122
    },
    {
      "epoch": 9.151561220909835,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 39123
    },
    {
      "epoch": 9.151795111682844,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 39124
    },
    {
      "epoch": 9.152029002455853,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 39125
    },
    {
      "epoch": 9.152262893228862,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 39126
    },
    {
      "epoch": 9.15249678400187,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 39127
    },
    {
      "epoch": 9.15273067477488,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 39128
    },
    {
      "epoch": 9.152964565547888,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6892,
      "step": 39129
    },
    {
      "epoch": 9.153198456320899,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.3646,
      "step": 39130
    },
    {
      "epoch": 9.153432347093908,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.775,
      "step": 39131
    },
    {
      "epoch": 9.153666237866917,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 39132
    },
    {
      "epoch": 9.153900128639926,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 39133
    },
    {
      "epoch": 9.154134019412934,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5934,
      "step": 39134
    },
    {
      "epoch": 9.154367910185943,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6048,
      "step": 39135
    },
    {
      "epoch": 9.154601800958952,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 39136
    },
    {
      "epoch": 9.15483569173196,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.6646,
      "step": 39137
    },
    {
      "epoch": 9.15506958250497,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6533,
      "step": 39138
    },
    {
      "epoch": 9.155303473277979,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6904,
      "step": 39139
    },
    {
      "epoch": 9.155537364050987,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.684,
      "step": 39140
    },
    {
      "epoch": 9.155771254823998,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.528,
      "step": 39141
    },
    {
      "epoch": 9.156005145597007,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.503,
      "step": 39142
    },
    {
      "epoch": 9.156239036370016,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 39143
    },
    {
      "epoch": 9.156472927143025,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.3516,
      "step": 39144
    },
    {
      "epoch": 9.156706817916033,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.68,
      "step": 39145
    },
    {
      "epoch": 9.156940708689042,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6536,
      "step": 39146
    },
    {
      "epoch": 9.157174599462051,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 39147
    },
    {
      "epoch": 9.15740849023506,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6159,
      "step": 39148
    },
    {
      "epoch": 9.157642381008069,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 39149
    },
    {
      "epoch": 9.157876271781078,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5588,
      "step": 39150
    },
    {
      "epoch": 9.158110162554086,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6589,
      "step": 39151
    },
    {
      "epoch": 9.158344053327097,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 39152
    },
    {
      "epoch": 9.158577944100106,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 39153
    },
    {
      "epoch": 9.158811834873115,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 39154
    },
    {
      "epoch": 9.159045725646124,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 39155
    },
    {
      "epoch": 9.159279616419132,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 39156
    },
    {
      "epoch": 9.159513507192141,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6325,
      "step": 39157
    },
    {
      "epoch": 9.15974739796515,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8073,
      "step": 39158
    },
    {
      "epoch": 9.159981288738159,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7086,
      "step": 39159
    },
    {
      "epoch": 9.160215179511168,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5134,
      "step": 39160
    },
    {
      "epoch": 9.160449070284177,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 39161
    },
    {
      "epoch": 9.160682961057187,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 39162
    },
    {
      "epoch": 9.160916851830196,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 39163
    },
    {
      "epoch": 9.161150742603205,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.608,
      "step": 39164
    },
    {
      "epoch": 9.161384633376214,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7908,
      "step": 39165
    },
    {
      "epoch": 9.161618524149223,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5948,
      "step": 39166
    },
    {
      "epoch": 9.161852414922231,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7608,
      "step": 39167
    },
    {
      "epoch": 9.16208630569524,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 39168
    },
    {
      "epoch": 9.162320196468249,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6938,
      "step": 39169
    },
    {
      "epoch": 9.162554087241258,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 39170
    },
    {
      "epoch": 9.162787978014267,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2535,
      "step": 39171
    },
    {
      "epoch": 9.163021868787276,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.4171,
      "step": 39172
    },
    {
      "epoch": 9.163255759560286,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.526,
      "step": 39173
    },
    {
      "epoch": 9.163489650333295,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 39174
    },
    {
      "epoch": 9.163723541106304,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 39175
    },
    {
      "epoch": 9.163957431879313,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8205,
      "step": 39176
    },
    {
      "epoch": 9.164191322652322,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6325,
      "step": 39177
    },
    {
      "epoch": 9.16442521342533,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 39178
    },
    {
      "epoch": 9.16465910419834,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7713,
      "step": 39179
    },
    {
      "epoch": 9.164892994971348,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 39180
    },
    {
      "epoch": 9.165126885744357,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.614,
      "step": 39181
    },
    {
      "epoch": 9.165360776517366,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.687,
      "step": 39182
    },
    {
      "epoch": 9.165594667290375,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 39183
    },
    {
      "epoch": 9.165828558063385,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6221,
      "step": 39184
    },
    {
      "epoch": 9.166062448836394,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4699,
      "step": 39185
    },
    {
      "epoch": 9.166296339609403,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 39186
    },
    {
      "epoch": 9.166530230382412,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 39187
    },
    {
      "epoch": 9.16676412115542,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 39188
    },
    {
      "epoch": 9.16699801192843,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 39189
    },
    {
      "epoch": 9.167231902701438,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5561,
      "step": 39190
    },
    {
      "epoch": 9.167465793474447,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7078,
      "step": 39191
    },
    {
      "epoch": 9.167699684247456,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 39192
    },
    {
      "epoch": 9.167933575020465,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6377,
      "step": 39193
    },
    {
      "epoch": 9.168167465793475,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 39194
    },
    {
      "epoch": 9.168401356566484,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5644,
      "step": 39195
    },
    {
      "epoch": 9.168635247339493,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7543,
      "step": 39196
    },
    {
      "epoch": 9.168869138112502,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6465,
      "step": 39197
    },
    {
      "epoch": 9.16910302888551,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 39198
    },
    {
      "epoch": 9.16933691965852,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8093,
      "step": 39199
    },
    {
      "epoch": 9.169570810431528,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.5586,
      "step": 39200
    },
    {
      "epoch": 9.169570810431528,
      "eval_runtime": 4.5912,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 39200
    },
    {
      "epoch": 9.169804701204537,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 39201
    },
    {
      "epoch": 9.170038591977546,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 39202
    },
    {
      "epoch": 9.170272482750555,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6691,
      "step": 39203
    },
    {
      "epoch": 9.170506373523564,
      "grad_norm": 9.5625,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 39204
    },
    {
      "epoch": 9.170740264296574,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 39205
    },
    {
      "epoch": 9.170974155069583,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.642,
      "step": 39206
    },
    {
      "epoch": 9.171208045842592,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 39207
    },
    {
      "epoch": 9.1714419366156,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6401,
      "step": 39208
    },
    {
      "epoch": 9.17167582738861,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 39209
    },
    {
      "epoch": 9.171909718161618,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.3836,
      "step": 39210
    },
    {
      "epoch": 9.172143608934627,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6715,
      "step": 39211
    },
    {
      "epoch": 9.172377499707636,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 39212
    },
    {
      "epoch": 9.172611390480645,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 39213
    },
    {
      "epoch": 9.172845281253654,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5622,
      "step": 39214
    },
    {
      "epoch": 9.173079172026663,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4569,
      "step": 39215
    },
    {
      "epoch": 9.173313062799673,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6216,
      "step": 39216
    },
    {
      "epoch": 9.173546953572682,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8254,
      "step": 39217
    },
    {
      "epoch": 9.173780844345691,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9478,
      "step": 39218
    },
    {
      "epoch": 9.1740147351187,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.5606,
      "step": 39219
    },
    {
      "epoch": 9.174248625891709,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5478,
      "step": 39220
    },
    {
      "epoch": 9.174482516664717,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7732,
      "step": 39221
    },
    {
      "epoch": 9.174716407437726,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5562,
      "step": 39222
    },
    {
      "epoch": 9.174950298210735,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 39223
    },
    {
      "epoch": 9.175184188983744,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6962,
      "step": 39224
    },
    {
      "epoch": 9.175418079756753,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 39225
    },
    {
      "epoch": 9.175651970529763,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 39226
    },
    {
      "epoch": 9.175885861302772,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6068,
      "step": 39227
    },
    {
      "epoch": 9.176119752075781,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 39228
    },
    {
      "epoch": 9.17635364284879,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 39229
    },
    {
      "epoch": 9.176587533621799,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 39230
    },
    {
      "epoch": 9.176821424394808,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 39231
    },
    {
      "epoch": 9.177055315167816,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5396,
      "step": 39232
    },
    {
      "epoch": 9.177289205940825,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 39233
    },
    {
      "epoch": 9.177523096713834,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 39234
    },
    {
      "epoch": 9.177756987486843,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7131,
      "step": 39235
    },
    {
      "epoch": 9.177990878259852,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6794,
      "step": 39236
    },
    {
      "epoch": 9.178224769032862,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.3813,
      "step": 39237
    },
    {
      "epoch": 9.178458659805871,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.5337,
      "step": 39238
    },
    {
      "epoch": 9.17869255057888,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6421,
      "step": 39239
    },
    {
      "epoch": 9.178926441351889,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7028,
      "step": 39240
    },
    {
      "epoch": 9.179160332124898,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5854,
      "step": 39241
    },
    {
      "epoch": 9.179394222897907,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6614,
      "step": 39242
    },
    {
      "epoch": 9.179628113670915,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 39243
    },
    {
      "epoch": 9.179862004443924,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6882,
      "step": 39244
    },
    {
      "epoch": 9.180095895216933,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8131,
      "step": 39245
    },
    {
      "epoch": 9.180329785989942,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 39246
    },
    {
      "epoch": 9.18056367676295,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6852,
      "step": 39247
    },
    {
      "epoch": 9.180797567535961,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 39248
    },
    {
      "epoch": 9.18103145830897,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5682,
      "step": 39249
    },
    {
      "epoch": 9.181265349081979,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5047,
      "step": 39250
    },
    {
      "epoch": 9.181499239854988,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 39251
    },
    {
      "epoch": 9.181733130627997,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 39252
    },
    {
      "epoch": 9.181967021401006,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 39253
    },
    {
      "epoch": 9.182200912174014,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6201,
      "step": 39254
    },
    {
      "epoch": 9.182434802947023,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6283,
      "step": 39255
    },
    {
      "epoch": 9.182668693720032,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 39256
    },
    {
      "epoch": 9.182902584493041,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6235,
      "step": 39257
    },
    {
      "epoch": 9.183136475266052,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.645,
      "step": 39258
    },
    {
      "epoch": 9.18337036603906,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 39259
    },
    {
      "epoch": 9.18360425681207,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7148,
      "step": 39260
    },
    {
      "epoch": 9.183838147585078,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5895,
      "step": 39261
    },
    {
      "epoch": 9.184072038358087,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7515,
      "step": 39262
    },
    {
      "epoch": 9.184305929131096,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 39263
    },
    {
      "epoch": 9.184539819904105,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7409,
      "step": 39264
    },
    {
      "epoch": 9.184773710677113,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6821,
      "step": 39265
    },
    {
      "epoch": 9.185007601450122,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 39266
    },
    {
      "epoch": 9.185241492223131,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4882,
      "step": 39267
    },
    {
      "epoch": 9.18547538299614,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7315,
      "step": 39268
    },
    {
      "epoch": 9.18570927376915,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 39269
    },
    {
      "epoch": 9.18594316454216,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 39270
    },
    {
      "epoch": 9.186177055315168,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 39271
    },
    {
      "epoch": 9.186410946088177,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 39272
    },
    {
      "epoch": 9.186644836861186,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5215,
      "step": 39273
    },
    {
      "epoch": 9.186878727634195,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 39274
    },
    {
      "epoch": 9.187112618407204,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 39275
    },
    {
      "epoch": 9.187346509180212,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6815,
      "step": 39276
    },
    {
      "epoch": 9.187580399953221,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 39277
    },
    {
      "epoch": 9.18781429072623,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5851,
      "step": 39278
    },
    {
      "epoch": 9.18804818149924,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 39279
    },
    {
      "epoch": 9.18828207227225,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 39280
    },
    {
      "epoch": 9.188515963045258,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5921,
      "step": 39281
    },
    {
      "epoch": 9.188749853818267,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 39282
    },
    {
      "epoch": 9.188983744591276,
      "grad_norm": 7.875,
      "learning_rate": 3e-05,
      "loss": 1.647,
      "step": 39283
    },
    {
      "epoch": 9.189217635364285,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.571,
      "step": 39284
    },
    {
      "epoch": 9.189451526137294,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 39285
    },
    {
      "epoch": 9.189685416910303,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 39286
    },
    {
      "epoch": 9.189919307683311,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 39287
    },
    {
      "epoch": 9.19015319845632,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.2833,
      "step": 39288
    },
    {
      "epoch": 9.19038708922933,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 39289
    },
    {
      "epoch": 9.19062098000234,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5084,
      "step": 39290
    },
    {
      "epoch": 9.190854870775349,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 39291
    },
    {
      "epoch": 9.191088761548357,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.72,
      "step": 39292
    },
    {
      "epoch": 9.191322652321366,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6526,
      "step": 39293
    },
    {
      "epoch": 9.191556543094375,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 39294
    },
    {
      "epoch": 9.191790433867384,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 39295
    },
    {
      "epoch": 9.192024324640393,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 39296
    },
    {
      "epoch": 9.192258215413402,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 39297
    },
    {
      "epoch": 9.19249210618641,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 39298
    },
    {
      "epoch": 9.19272599695942,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 39299
    },
    {
      "epoch": 9.192959887732428,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6931,
      "step": 39300
    },
    {
      "epoch": 9.192959887732428,
      "eval_runtime": 4.6429,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 39300
    },
    {
      "epoch": 9.193193778505439,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7307,
      "step": 39301
    },
    {
      "epoch": 9.193427669278448,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 39302
    },
    {
      "epoch": 9.193661560051456,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 39303
    },
    {
      "epoch": 9.193895450824465,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.6729,
      "step": 39304
    },
    {
      "epoch": 9.194129341597474,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6024,
      "step": 39305
    },
    {
      "epoch": 9.194363232370483,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 39306
    },
    {
      "epoch": 9.194597123143492,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5073,
      "step": 39307
    },
    {
      "epoch": 9.1948310139165,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 39308
    },
    {
      "epoch": 9.19506490468951,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 39309
    },
    {
      "epoch": 9.195298795462518,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5758,
      "step": 39310
    },
    {
      "epoch": 9.195532686235527,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7327,
      "step": 39311
    },
    {
      "epoch": 9.195766577008538,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 39312
    },
    {
      "epoch": 9.196000467781547,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5139,
      "step": 39313
    },
    {
      "epoch": 9.196234358554555,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 39314
    },
    {
      "epoch": 9.196468249327564,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 39315
    },
    {
      "epoch": 9.196702140100573,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6019,
      "step": 39316
    },
    {
      "epoch": 9.196936030873582,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 39317
    },
    {
      "epoch": 9.19716992164659,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7628,
      "step": 39318
    },
    {
      "epoch": 9.1974038124196,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6909,
      "step": 39319
    },
    {
      "epoch": 9.197637703192608,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.599,
      "step": 39320
    },
    {
      "epoch": 9.197871593965617,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 39321
    },
    {
      "epoch": 9.198105484738628,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 39322
    },
    {
      "epoch": 9.198339375511637,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6476,
      "step": 39323
    },
    {
      "epoch": 9.198573266284646,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 39324
    },
    {
      "epoch": 9.198807157057654,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 39325
    },
    {
      "epoch": 9.199041047830663,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5074,
      "step": 39326
    },
    {
      "epoch": 9.199274938603672,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 39327
    },
    {
      "epoch": 9.199508829376681,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5744,
      "step": 39328
    },
    {
      "epoch": 9.19974272014969,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.718,
      "step": 39329
    },
    {
      "epoch": 9.199976610922699,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 39330
    },
    {
      "epoch": 9.200210501695707,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 39331
    },
    {
      "epoch": 9.200444392468716,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.578,
      "step": 39332
    },
    {
      "epoch": 9.200678283241727,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7028,
      "step": 39333
    },
    {
      "epoch": 9.200912174014736,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6996,
      "step": 39334
    },
    {
      "epoch": 9.201146064787745,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7614,
      "step": 39335
    },
    {
      "epoch": 9.201379955560753,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 39336
    },
    {
      "epoch": 9.201613846333762,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 39337
    },
    {
      "epoch": 9.201847737106771,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6032,
      "step": 39338
    },
    {
      "epoch": 9.20208162787978,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 39339
    },
    {
      "epoch": 9.202315518652789,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6554,
      "step": 39340
    },
    {
      "epoch": 9.202549409425798,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5899,
      "step": 39341
    },
    {
      "epoch": 9.202783300198806,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6958,
      "step": 39342
    },
    {
      "epoch": 9.203017190971817,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 39343
    },
    {
      "epoch": 9.203251081744826,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 39344
    },
    {
      "epoch": 9.203484972517835,
      "grad_norm": 2.390625,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 39345
    },
    {
      "epoch": 9.203718863290844,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 39346
    },
    {
      "epoch": 9.203952754063852,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8175,
      "step": 39347
    },
    {
      "epoch": 9.204186644836861,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 39348
    },
    {
      "epoch": 9.20442053560987,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 39349
    },
    {
      "epoch": 9.204654426382879,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.622,
      "step": 39350
    },
    {
      "epoch": 9.204888317155888,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6083,
      "step": 39351
    },
    {
      "epoch": 9.205122207928897,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5607,
      "step": 39352
    },
    {
      "epoch": 9.205356098701905,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8423,
      "step": 39353
    },
    {
      "epoch": 9.205589989474916,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5054,
      "step": 39354
    },
    {
      "epoch": 9.205823880247925,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 39355
    },
    {
      "epoch": 9.206057771020934,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 39356
    },
    {
      "epoch": 9.206291661793943,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4991,
      "step": 39357
    },
    {
      "epoch": 9.206525552566951,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4094,
      "step": 39358
    },
    {
      "epoch": 9.20675944333996,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6532,
      "step": 39359
    },
    {
      "epoch": 9.206993334112969,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 39360
    },
    {
      "epoch": 9.207227224885978,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 39361
    },
    {
      "epoch": 9.207461115658987,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7078,
      "step": 39362
    },
    {
      "epoch": 9.207695006431996,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 39363
    },
    {
      "epoch": 9.207928897205004,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 39364
    },
    {
      "epoch": 9.208162787978015,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6695,
      "step": 39365
    },
    {
      "epoch": 9.208396678751024,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 39366
    },
    {
      "epoch": 9.208630569524033,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 39367
    },
    {
      "epoch": 9.208864460297042,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 39368
    },
    {
      "epoch": 9.20909835107005,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 39369
    },
    {
      "epoch": 9.20933224184306,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 39370
    },
    {
      "epoch": 9.209566132616068,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 39371
    },
    {
      "epoch": 9.209800023389077,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 39372
    },
    {
      "epoch": 9.210033914162086,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 39373
    },
    {
      "epoch": 9.210267804935095,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 39374
    },
    {
      "epoch": 9.210501695708105,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 39375
    },
    {
      "epoch": 9.210735586481114,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7061,
      "step": 39376
    },
    {
      "epoch": 9.210969477254123,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5064,
      "step": 39377
    },
    {
      "epoch": 9.211203368027132,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4166,
      "step": 39378
    },
    {
      "epoch": 9.21143725880014,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 39379
    },
    {
      "epoch": 9.21167114957315,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 39380
    },
    {
      "epoch": 9.211905040346158,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.747,
      "step": 39381
    },
    {
      "epoch": 9.212138931119167,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 39382
    },
    {
      "epoch": 9.212372821892176,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5549,
      "step": 39383
    },
    {
      "epoch": 9.212606712665185,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7415,
      "step": 39384
    },
    {
      "epoch": 9.212840603438194,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 39385
    },
    {
      "epoch": 9.213074494211204,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.7608,
      "step": 39386
    },
    {
      "epoch": 9.213308384984213,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 39387
    },
    {
      "epoch": 9.213542275757222,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 39388
    },
    {
      "epoch": 9.21377616653023,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 39389
    },
    {
      "epoch": 9.21401005730324,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.645,
      "step": 39390
    },
    {
      "epoch": 9.214243948076248,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6325,
      "step": 39391
    },
    {
      "epoch": 9.214477838849257,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 39392
    },
    {
      "epoch": 9.214711729622266,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 39393
    },
    {
      "epoch": 9.214945620395275,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 39394
    },
    {
      "epoch": 9.215179511168284,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6425,
      "step": 39395
    },
    {
      "epoch": 9.215413401941293,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 39396
    },
    {
      "epoch": 9.215647292714303,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.704,
      "step": 39397
    },
    {
      "epoch": 9.215881183487312,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.727,
      "step": 39398
    },
    {
      "epoch": 9.21611507426032,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5642,
      "step": 39399
    },
    {
      "epoch": 9.21634896503333,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 39400
    },
    {
      "epoch": 9.21634896503333,
      "eval_runtime": 4.6613,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 39400
    },
    {
      "epoch": 9.216582855806339,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6314,
      "step": 39401
    },
    {
      "epoch": 9.216816746579347,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 39402
    },
    {
      "epoch": 9.217050637352356,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5828,
      "step": 39403
    },
    {
      "epoch": 9.217284528125365,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6856,
      "step": 39404
    },
    {
      "epoch": 9.217518418898374,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7,
      "step": 39405
    },
    {
      "epoch": 9.217752309671383,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.5283,
      "step": 39406
    },
    {
      "epoch": 9.217986200444393,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4339,
      "step": 39407
    },
    {
      "epoch": 9.218220091217402,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 39408
    },
    {
      "epoch": 9.218453981990411,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5204,
      "step": 39409
    },
    {
      "epoch": 9.21868787276342,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7515,
      "step": 39410
    },
    {
      "epoch": 9.218921763536429,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 39411
    },
    {
      "epoch": 9.219155654309438,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4926,
      "step": 39412
    },
    {
      "epoch": 9.219389545082446,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.594,
      "step": 39413
    },
    {
      "epoch": 9.219623435855455,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.631,
      "step": 39414
    },
    {
      "epoch": 9.219857326628464,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5136,
      "step": 39415
    },
    {
      "epoch": 9.220091217401473,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 39416
    },
    {
      "epoch": 9.220325108174482,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6565,
      "step": 39417
    },
    {
      "epoch": 9.220558998947492,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 39418
    },
    {
      "epoch": 9.220792889720501,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 39419
    },
    {
      "epoch": 9.22102678049351,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 39420
    },
    {
      "epoch": 9.221260671266519,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5311,
      "step": 39421
    },
    {
      "epoch": 9.221494562039528,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 39422
    },
    {
      "epoch": 9.221728452812537,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 39423
    },
    {
      "epoch": 9.221962343585545,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6975,
      "step": 39424
    },
    {
      "epoch": 9.222196234358554,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 39425
    },
    {
      "epoch": 9.222430125131563,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6627,
      "step": 39426
    },
    {
      "epoch": 9.222664015904572,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 39427
    },
    {
      "epoch": 9.22289790667758,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 39428
    },
    {
      "epoch": 9.223131797450591,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 39429
    },
    {
      "epoch": 9.2233656882236,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5601,
      "step": 39430
    },
    {
      "epoch": 9.223599578996609,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 39431
    },
    {
      "epoch": 9.223833469769618,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6927,
      "step": 39432
    },
    {
      "epoch": 9.224067360542627,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 39433
    },
    {
      "epoch": 9.224301251315635,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 39434
    },
    {
      "epoch": 9.224535142088644,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5493,
      "step": 39435
    },
    {
      "epoch": 9.224769032861653,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 39436
    },
    {
      "epoch": 9.225002923634662,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5762,
      "step": 39437
    },
    {
      "epoch": 9.22523681440767,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7011,
      "step": 39438
    },
    {
      "epoch": 9.225470705180681,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 39439
    },
    {
      "epoch": 9.22570459595369,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 39440
    },
    {
      "epoch": 9.2259384867267,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6138,
      "step": 39441
    },
    {
      "epoch": 9.226172377499708,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4352,
      "step": 39442
    },
    {
      "epoch": 9.226406268272717,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 39443
    },
    {
      "epoch": 9.226640159045726,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5112,
      "step": 39444
    },
    {
      "epoch": 9.226874049818734,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 39445
    },
    {
      "epoch": 9.227107940591743,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 39446
    },
    {
      "epoch": 9.227341831364752,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 39447
    },
    {
      "epoch": 9.227575722137761,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6223,
      "step": 39448
    },
    {
      "epoch": 9.22780961291077,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.177,
      "step": 39449
    },
    {
      "epoch": 9.22804350368378,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 39450
    },
    {
      "epoch": 9.22827739445679,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 39451
    },
    {
      "epoch": 9.228511285229798,
      "grad_norm": 12.0625,
      "learning_rate": 3e-05,
      "loss": 1.6523,
      "step": 39452
    },
    {
      "epoch": 9.228745176002807,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 39453
    },
    {
      "epoch": 9.228979066775816,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 39454
    },
    {
      "epoch": 9.229212957548825,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 39455
    },
    {
      "epoch": 9.229446848321833,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 39456
    },
    {
      "epoch": 9.229680739094842,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6105,
      "step": 39457
    },
    {
      "epoch": 9.229914629867851,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6105,
      "step": 39458
    },
    {
      "epoch": 9.23014852064086,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.6767,
      "step": 39459
    },
    {
      "epoch": 9.23038241141387,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 39460
    },
    {
      "epoch": 9.23061630218688,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 39461
    },
    {
      "epoch": 9.230850192959888,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 39462
    },
    {
      "epoch": 9.231084083732897,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.545,
      "step": 39463
    },
    {
      "epoch": 9.231317974505906,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.4586,
      "step": 39464
    },
    {
      "epoch": 9.231551865278915,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 39465
    },
    {
      "epoch": 9.231785756051924,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 39466
    },
    {
      "epoch": 9.232019646824932,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 39467
    },
    {
      "epoch": 9.232253537597941,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6864,
      "step": 39468
    },
    {
      "epoch": 9.23248742837095,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5648,
      "step": 39469
    },
    {
      "epoch": 9.232721319143959,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 39470
    },
    {
      "epoch": 9.23295520991697,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6822,
      "step": 39471
    },
    {
      "epoch": 9.233189100689978,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 39472
    },
    {
      "epoch": 9.233422991462987,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5533,
      "step": 39473
    },
    {
      "epoch": 9.233656882235996,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 39474
    },
    {
      "epoch": 9.233890773009005,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4808,
      "step": 39475
    },
    {
      "epoch": 9.234124663782014,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 39476
    },
    {
      "epoch": 9.234358554555023,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 39477
    },
    {
      "epoch": 9.234592445328031,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 39478
    },
    {
      "epoch": 9.23482633610104,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 39479
    },
    {
      "epoch": 9.23506022687405,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 39480
    },
    {
      "epoch": 9.235294117647058,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7399,
      "step": 39481
    },
    {
      "epoch": 9.235528008420069,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 39482
    },
    {
      "epoch": 9.235761899193077,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 39483
    },
    {
      "epoch": 9.235995789966086,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6855,
      "step": 39484
    },
    {
      "epoch": 9.236229680739095,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4912,
      "step": 39485
    },
    {
      "epoch": 9.236463571512104,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 39486
    },
    {
      "epoch": 9.236697462285113,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 39487
    },
    {
      "epoch": 9.236931353058122,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5932,
      "step": 39488
    },
    {
      "epoch": 9.23716524383113,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 39489
    },
    {
      "epoch": 9.23739913460414,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 39490
    },
    {
      "epoch": 9.237633025377148,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.717,
      "step": 39491
    },
    {
      "epoch": 9.237866916150157,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5635,
      "step": 39492
    },
    {
      "epoch": 9.238100806923168,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.696,
      "step": 39493
    },
    {
      "epoch": 9.238334697696176,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7499,
      "step": 39494
    },
    {
      "epoch": 9.238568588469185,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5504,
      "step": 39495
    },
    {
      "epoch": 9.238802479242194,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 39496
    },
    {
      "epoch": 9.239036370015203,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 39497
    },
    {
      "epoch": 9.239270260788212,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5368,
      "step": 39498
    },
    {
      "epoch": 9.23950415156122,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 39499
    },
    {
      "epoch": 9.23973804233423,
      "grad_norm": 13.125,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 39500
    },
    {
      "epoch": 9.23973804233423,
      "eval_runtime": 4.6355,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 39500
    },
    {
      "epoch": 9.239971933107238,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 39501
    },
    {
      "epoch": 9.240205823880247,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7443,
      "step": 39502
    },
    {
      "epoch": 9.240439714653258,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 39503
    },
    {
      "epoch": 9.240673605426267,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.7488,
      "step": 39504
    },
    {
      "epoch": 9.240907496199275,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5404,
      "step": 39505
    },
    {
      "epoch": 9.241141386972284,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 39506
    },
    {
      "epoch": 9.241375277745293,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5357,
      "step": 39507
    },
    {
      "epoch": 9.241609168518302,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 39508
    },
    {
      "epoch": 9.24184305929131,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7039,
      "step": 39509
    },
    {
      "epoch": 9.24207695006432,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 39510
    },
    {
      "epoch": 9.242310840837328,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.4946,
      "step": 39511
    },
    {
      "epoch": 9.242544731610337,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5068,
      "step": 39512
    },
    {
      "epoch": 9.242778622383346,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 39513
    },
    {
      "epoch": 9.243012513156357,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 39514
    },
    {
      "epoch": 9.243246403929366,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5522,
      "step": 39515
    },
    {
      "epoch": 9.243480294702374,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7777,
      "step": 39516
    },
    {
      "epoch": 9.243714185475383,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 39517
    },
    {
      "epoch": 9.243948076248392,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 39518
    },
    {
      "epoch": 9.244181967021401,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 39519
    },
    {
      "epoch": 9.24441585779441,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 39520
    },
    {
      "epoch": 9.244649748567419,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6928,
      "step": 39521
    },
    {
      "epoch": 9.244883639340427,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.8991,
      "step": 39522
    },
    {
      "epoch": 9.245117530113436,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.383,
      "step": 39523
    },
    {
      "epoch": 9.245351420886447,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6469,
      "step": 39524
    },
    {
      "epoch": 9.245585311659456,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 39525
    },
    {
      "epoch": 9.245819202432465,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 39526
    },
    {
      "epoch": 9.246053093205473,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 39527
    },
    {
      "epoch": 9.246286983978482,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7457,
      "step": 39528
    },
    {
      "epoch": 9.246520874751491,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.688,
      "step": 39529
    },
    {
      "epoch": 9.2467547655245,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 39530
    },
    {
      "epoch": 9.246988656297509,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5542,
      "step": 39531
    },
    {
      "epoch": 9.247222547070518,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 39532
    },
    {
      "epoch": 9.247456437843526,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 39533
    },
    {
      "epoch": 9.247690328616535,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5541,
      "step": 39534
    },
    {
      "epoch": 9.247924219389546,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4227,
      "step": 39535
    },
    {
      "epoch": 9.248158110162555,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6592,
      "step": 39536
    },
    {
      "epoch": 9.248392000935564,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7627,
      "step": 39537
    },
    {
      "epoch": 9.248625891708572,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5476,
      "step": 39538
    },
    {
      "epoch": 9.248859782481581,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 39539
    },
    {
      "epoch": 9.24909367325459,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 39540
    },
    {
      "epoch": 9.249327564027599,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6362,
      "step": 39541
    },
    {
      "epoch": 9.249561454800608,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5924,
      "step": 39542
    },
    {
      "epoch": 9.249795345573617,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 39543
    },
    {
      "epoch": 9.250029236346625,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9455,
      "step": 39544
    },
    {
      "epoch": 9.250263127119634,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 39545
    },
    {
      "epoch": 9.250497017892645,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6144,
      "step": 39546
    },
    {
      "epoch": 9.250730908665654,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 39547
    },
    {
      "epoch": 9.250964799438663,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 39548
    },
    {
      "epoch": 9.251198690211671,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6634,
      "step": 39549
    },
    {
      "epoch": 9.25143258098468,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8143,
      "step": 39550
    },
    {
      "epoch": 9.251666471757689,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 39551
    },
    {
      "epoch": 9.251900362530698,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 39552
    },
    {
      "epoch": 9.252134253303707,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 39553
    },
    {
      "epoch": 9.252368144076716,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6393,
      "step": 39554
    },
    {
      "epoch": 9.252602034849724,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.793,
      "step": 39555
    },
    {
      "epoch": 9.252835925622733,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7028,
      "step": 39556
    },
    {
      "epoch": 9.253069816395744,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5853,
      "step": 39557
    },
    {
      "epoch": 9.253303707168753,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 39558
    },
    {
      "epoch": 9.253537597941762,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 39559
    },
    {
      "epoch": 9.25377148871477,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6922,
      "step": 39560
    },
    {
      "epoch": 9.25400537948778,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.2996,
      "step": 39561
    },
    {
      "epoch": 9.254239270260788,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 39562
    },
    {
      "epoch": 9.254473161033797,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 39563
    },
    {
      "epoch": 9.254707051806806,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 39564
    },
    {
      "epoch": 9.254940942579815,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 39565
    },
    {
      "epoch": 9.255174833352823,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6374,
      "step": 39566
    },
    {
      "epoch": 9.255408724125834,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 39567
    },
    {
      "epoch": 9.255642614898843,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 39568
    },
    {
      "epoch": 9.255876505671852,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.772,
      "step": 39569
    },
    {
      "epoch": 9.25611039644486,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 39570
    },
    {
      "epoch": 9.25634428721787,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 39571
    },
    {
      "epoch": 9.256578177990878,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 39572
    },
    {
      "epoch": 9.256812068763887,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 39573
    },
    {
      "epoch": 9.257045959536896,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6632,
      "step": 39574
    },
    {
      "epoch": 9.257279850309905,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 39575
    },
    {
      "epoch": 9.257513741082914,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7403,
      "step": 39576
    },
    {
      "epoch": 9.257747631855924,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7453,
      "step": 39577
    },
    {
      "epoch": 9.257981522628933,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 39578
    },
    {
      "epoch": 9.258215413401942,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6712,
      "step": 39579
    },
    {
      "epoch": 9.25844930417495,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6383,
      "step": 39580
    },
    {
      "epoch": 9.25868319494796,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 39581
    },
    {
      "epoch": 9.258917085720968,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 39582
    },
    {
      "epoch": 9.259150976493977,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 39583
    },
    {
      "epoch": 9.259384867266986,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7081,
      "step": 39584
    },
    {
      "epoch": 9.259618758039995,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.4684,
      "step": 39585
    },
    {
      "epoch": 9.259852648813004,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6603,
      "step": 39586
    },
    {
      "epoch": 9.260086539586013,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 39587
    },
    {
      "epoch": 9.260320430359023,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.3922,
      "step": 39588
    },
    {
      "epoch": 9.260554321132032,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.4761,
      "step": 39589
    },
    {
      "epoch": 9.26078821190504,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 39590
    },
    {
      "epoch": 9.26102210267805,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.3648,
      "step": 39591
    },
    {
      "epoch": 9.261255993451059,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 39592
    },
    {
      "epoch": 9.261489884224067,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.501,
      "step": 39593
    },
    {
      "epoch": 9.261723774997076,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5134,
      "step": 39594
    },
    {
      "epoch": 9.261957665770085,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6041,
      "step": 39595
    },
    {
      "epoch": 9.262191556543094,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 39596
    },
    {
      "epoch": 9.262425447316103,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6429,
      "step": 39597
    },
    {
      "epoch": 9.262659338089112,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6248,
      "step": 39598
    },
    {
      "epoch": 9.262893228862122,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9087,
      "step": 39599
    },
    {
      "epoch": 9.263127119635131,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 39600
    },
    {
      "epoch": 9.263127119635131,
      "eval_runtime": 4.5806,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 39600
    },
    {
      "epoch": 9.26336101040814,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 39601
    },
    {
      "epoch": 9.263594901181149,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6591,
      "step": 39602
    },
    {
      "epoch": 9.263828791954158,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 39603
    },
    {
      "epoch": 9.264062682727166,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6657,
      "step": 39604
    },
    {
      "epoch": 9.264296573500175,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 39605
    },
    {
      "epoch": 9.264530464273184,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 39606
    },
    {
      "epoch": 9.264764355046193,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 39607
    },
    {
      "epoch": 9.264998245819202,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6126,
      "step": 39608
    },
    {
      "epoch": 9.26523213659221,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 39609
    },
    {
      "epoch": 9.265466027365221,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 39610
    },
    {
      "epoch": 9.26569991813823,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7041,
      "step": 39611
    },
    {
      "epoch": 9.265933808911239,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.3144,
      "step": 39612
    },
    {
      "epoch": 9.266167699684248,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 39613
    },
    {
      "epoch": 9.266401590457257,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6232,
      "step": 39614
    },
    {
      "epoch": 9.266635481230265,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 39615
    },
    {
      "epoch": 9.266869372003274,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 39616
    },
    {
      "epoch": 9.267103262776283,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 39617
    },
    {
      "epoch": 9.267337153549292,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6394,
      "step": 39618
    },
    {
      "epoch": 9.2675710443223,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 39619
    },
    {
      "epoch": 9.267804935095311,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 39620
    },
    {
      "epoch": 9.26803882586832,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6788,
      "step": 39621
    },
    {
      "epoch": 9.268272716641329,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 39622
    },
    {
      "epoch": 9.268506607414338,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 39623
    },
    {
      "epoch": 9.268740498187347,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 39624
    },
    {
      "epoch": 9.268974388960356,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 39625
    },
    {
      "epoch": 9.269208279733364,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 39626
    },
    {
      "epoch": 9.269442170506373,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4898,
      "step": 39627
    },
    {
      "epoch": 9.269676061279382,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 39628
    },
    {
      "epoch": 9.26990995205239,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6493,
      "step": 39629
    },
    {
      "epoch": 9.2701438428254,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.3425,
      "step": 39630
    },
    {
      "epoch": 9.27037773359841,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 39631
    },
    {
      "epoch": 9.27061162437142,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6168,
      "step": 39632
    },
    {
      "epoch": 9.270845515144428,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 39633
    },
    {
      "epoch": 9.271079405917437,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 39634
    },
    {
      "epoch": 9.271313296690446,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 39635
    },
    {
      "epoch": 9.271547187463455,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 39636
    },
    {
      "epoch": 9.271781078236463,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 39637
    },
    {
      "epoch": 9.272014969009472,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6928,
      "step": 39638
    },
    {
      "epoch": 9.272248859782481,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4797,
      "step": 39639
    },
    {
      "epoch": 9.27248275055549,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 39640
    },
    {
      "epoch": 9.2727166413285,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 39641
    },
    {
      "epoch": 9.27295053210151,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7535,
      "step": 39642
    },
    {
      "epoch": 9.273184422874518,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 39643
    },
    {
      "epoch": 9.273418313647527,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5977,
      "step": 39644
    },
    {
      "epoch": 9.273652204420536,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5705,
      "step": 39645
    },
    {
      "epoch": 9.273886095193545,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4834,
      "step": 39646
    },
    {
      "epoch": 9.274119985966554,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 39647
    },
    {
      "epoch": 9.274353876739562,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6973,
      "step": 39648
    },
    {
      "epoch": 9.274587767512571,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 39649
    },
    {
      "epoch": 9.27482165828558,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 39650
    },
    {
      "epoch": 9.275055549058589,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7181,
      "step": 39651
    },
    {
      "epoch": 9.2752894398316,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 39652
    },
    {
      "epoch": 9.275523330604608,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 39653
    },
    {
      "epoch": 9.275757221377617,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 39654
    },
    {
      "epoch": 9.275991112150626,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.5079,
      "step": 39655
    },
    {
      "epoch": 9.276225002923635,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4248,
      "step": 39656
    },
    {
      "epoch": 9.276458893696644,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 39657
    },
    {
      "epoch": 9.276692784469653,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 39658
    },
    {
      "epoch": 9.276926675242661,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 39659
    },
    {
      "epoch": 9.27716056601567,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 39660
    },
    {
      "epoch": 9.277394456788679,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 39661
    },
    {
      "epoch": 9.277628347561688,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 39662
    },
    {
      "epoch": 9.277862238334698,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 39663
    },
    {
      "epoch": 9.278096129107707,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 39664
    },
    {
      "epoch": 9.278330019880716,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 39665
    },
    {
      "epoch": 9.278563910653725,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7682,
      "step": 39666
    },
    {
      "epoch": 9.278797801426734,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0032,
      "step": 39667
    },
    {
      "epoch": 9.279031692199743,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.63,
      "step": 39668
    },
    {
      "epoch": 9.279265582972752,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7084,
      "step": 39669
    },
    {
      "epoch": 9.27949947374576,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 39670
    },
    {
      "epoch": 9.27973336451877,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 39671
    },
    {
      "epoch": 9.279967255291778,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 39672
    },
    {
      "epoch": 9.280201146064787,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5259,
      "step": 39673
    },
    {
      "epoch": 9.280435036837797,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 39674
    },
    {
      "epoch": 9.280668927610806,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.55,
      "step": 39675
    },
    {
      "epoch": 9.280902818383815,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6666,
      "step": 39676
    },
    {
      "epoch": 9.281136709156824,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 39677
    },
    {
      "epoch": 9.281370599929833,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0492,
      "step": 39678
    },
    {
      "epoch": 9.281604490702842,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6997,
      "step": 39679
    },
    {
      "epoch": 9.28183838147585,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7081,
      "step": 39680
    },
    {
      "epoch": 9.28207227224886,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.687,
      "step": 39681
    },
    {
      "epoch": 9.282306163021868,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.4205,
      "step": 39682
    },
    {
      "epoch": 9.282540053794877,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 39683
    },
    {
      "epoch": 9.282773944567888,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5494,
      "step": 39684
    },
    {
      "epoch": 9.283007835340896,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 39685
    },
    {
      "epoch": 9.283241726113905,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6803,
      "step": 39686
    },
    {
      "epoch": 9.283475616886914,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6698,
      "step": 39687
    },
    {
      "epoch": 9.283709507659923,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 39688
    },
    {
      "epoch": 9.283943398432932,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 39689
    },
    {
      "epoch": 9.28417728920594,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.3973,
      "step": 39690
    },
    {
      "epoch": 9.28441117997895,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 39691
    },
    {
      "epoch": 9.284645070751958,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5434,
      "step": 39692
    },
    {
      "epoch": 9.284878961524967,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6261,
      "step": 39693
    },
    {
      "epoch": 9.285112852297976,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.4553,
      "step": 39694
    },
    {
      "epoch": 9.285346743070987,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 39695
    },
    {
      "epoch": 9.285580633843995,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7356,
      "step": 39696
    },
    {
      "epoch": 9.285814524617004,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 39697
    },
    {
      "epoch": 9.286048415390013,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 39698
    },
    {
      "epoch": 9.286282306163022,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 39699
    },
    {
      "epoch": 9.28651619693603,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 39700
    },
    {
      "epoch": 9.28651619693603,
      "eval_runtime": 4.621,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 39700
    },
    {
      "epoch": 9.28675008770904,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 39701
    },
    {
      "epoch": 9.286983978482048,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6183,
      "step": 39702
    },
    {
      "epoch": 9.287217869255057,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.564,
      "step": 39703
    },
    {
      "epoch": 9.287451760028066,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7513,
      "step": 39704
    },
    {
      "epoch": 9.287685650801077,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6785,
      "step": 39705
    },
    {
      "epoch": 9.287919541574086,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 39706
    },
    {
      "epoch": 9.288153432347094,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 39707
    },
    {
      "epoch": 9.288387323120103,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7817,
      "step": 39708
    },
    {
      "epoch": 9.288621213893112,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8156,
      "step": 39709
    },
    {
      "epoch": 9.288855104666121,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 39710
    },
    {
      "epoch": 9.28908899543913,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6689,
      "step": 39711
    },
    {
      "epoch": 9.289322886212139,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.7531,
      "step": 39712
    },
    {
      "epoch": 9.289556776985147,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7781,
      "step": 39713
    },
    {
      "epoch": 9.289790667758156,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 39714
    },
    {
      "epoch": 9.290024558531165,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 39715
    },
    {
      "epoch": 9.290258449304176,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 39716
    },
    {
      "epoch": 9.290492340077185,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.5144,
      "step": 39717
    },
    {
      "epoch": 9.290726230850193,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 39718
    },
    {
      "epoch": 9.290960121623202,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6313,
      "step": 39719
    },
    {
      "epoch": 9.291194012396211,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 39720
    },
    {
      "epoch": 9.29142790316922,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 39721
    },
    {
      "epoch": 9.291661793942229,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7698,
      "step": 39722
    },
    {
      "epoch": 9.291895684715238,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6537,
      "step": 39723
    },
    {
      "epoch": 9.292129575488246,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.3948,
      "step": 39724
    },
    {
      "epoch": 9.292363466261255,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6052,
      "step": 39725
    },
    {
      "epoch": 9.292597357034264,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8146,
      "step": 39726
    },
    {
      "epoch": 9.292831247807275,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 39727
    },
    {
      "epoch": 9.293065138580284,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 39728
    },
    {
      "epoch": 9.293299029353292,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7257,
      "step": 39729
    },
    {
      "epoch": 9.293532920126301,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.3737,
      "step": 39730
    },
    {
      "epoch": 9.29376681089931,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6212,
      "step": 39731
    },
    {
      "epoch": 9.294000701672319,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6316,
      "step": 39732
    },
    {
      "epoch": 9.294234592445328,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.1967,
      "step": 39733
    },
    {
      "epoch": 9.294468483218337,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6523,
      "step": 39734
    },
    {
      "epoch": 9.294702373991345,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5278,
      "step": 39735
    },
    {
      "epoch": 9.294936264764354,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4576,
      "step": 39736
    },
    {
      "epoch": 9.295170155537363,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 39737
    },
    {
      "epoch": 9.295404046310374,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 39738
    },
    {
      "epoch": 9.295637937083383,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6015,
      "step": 39739
    },
    {
      "epoch": 9.295871827856391,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 39740
    },
    {
      "epoch": 9.2961057186294,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 39741
    },
    {
      "epoch": 9.296339609402409,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 39742
    },
    {
      "epoch": 9.296573500175418,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 39743
    },
    {
      "epoch": 9.296807390948427,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.575,
      "step": 39744
    },
    {
      "epoch": 9.297041281721436,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 39745
    },
    {
      "epoch": 9.297275172494444,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 39746
    },
    {
      "epoch": 9.297509063267453,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 39747
    },
    {
      "epoch": 9.297742954040464,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 39748
    },
    {
      "epoch": 9.297976844813473,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 39749
    },
    {
      "epoch": 9.298210735586482,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 39750
    },
    {
      "epoch": 9.29844462635949,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 39751
    },
    {
      "epoch": 9.2986785171325,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5117,
      "step": 39752
    },
    {
      "epoch": 9.298912407905508,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 39753
    },
    {
      "epoch": 9.299146298678517,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9432,
      "step": 39754
    },
    {
      "epoch": 9.299380189451526,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.5411,
      "step": 39755
    },
    {
      "epoch": 9.299614080224535,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 39756
    },
    {
      "epoch": 9.299847970997543,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 39757
    },
    {
      "epoch": 9.300081861770552,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7749,
      "step": 39758
    },
    {
      "epoch": 9.300315752543563,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6231,
      "step": 39759
    },
    {
      "epoch": 9.300549643316572,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7185,
      "step": 39760
    },
    {
      "epoch": 9.30078353408958,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.7579,
      "step": 39761
    },
    {
      "epoch": 9.30101742486259,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7609,
      "step": 39762
    },
    {
      "epoch": 9.301251315635598,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5494,
      "step": 39763
    },
    {
      "epoch": 9.301485206408607,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.7242,
      "step": 39764
    },
    {
      "epoch": 9.301719097181616,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7398,
      "step": 39765
    },
    {
      "epoch": 9.301952987954625,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 39766
    },
    {
      "epoch": 9.302186878727634,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7116,
      "step": 39767
    },
    {
      "epoch": 9.302420769500642,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 39768
    },
    {
      "epoch": 9.302654660273653,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 39769
    },
    {
      "epoch": 9.302888551046662,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 39770
    },
    {
      "epoch": 9.30312244181967,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7767,
      "step": 39771
    },
    {
      "epoch": 9.30335633259268,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 39772
    },
    {
      "epoch": 9.303590223365688,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.8061,
      "step": 39773
    },
    {
      "epoch": 9.303824114138697,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6179,
      "step": 39774
    },
    {
      "epoch": 9.304058004911706,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 39775
    },
    {
      "epoch": 9.304291895684715,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 39776
    },
    {
      "epoch": 9.304525786457724,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 39777
    },
    {
      "epoch": 9.304759677230733,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4055,
      "step": 39778
    },
    {
      "epoch": 9.304993568003741,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4005,
      "step": 39779
    },
    {
      "epoch": 9.305227458776752,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2607,
      "step": 39780
    },
    {
      "epoch": 9.305461349549761,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 39781
    },
    {
      "epoch": 9.30569524032277,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 39782
    },
    {
      "epoch": 9.305929131095779,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.7785,
      "step": 39783
    },
    {
      "epoch": 9.306163021868787,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6403,
      "step": 39784
    },
    {
      "epoch": 9.306396912641796,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7099,
      "step": 39785
    },
    {
      "epoch": 9.306630803414805,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 39786
    },
    {
      "epoch": 9.306864694187814,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8011,
      "step": 39787
    },
    {
      "epoch": 9.307098584960823,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 39788
    },
    {
      "epoch": 9.307332475733832,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 39789
    },
    {
      "epoch": 9.30756636650684,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 39790
    },
    {
      "epoch": 9.307800257279851,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 39791
    },
    {
      "epoch": 9.30803414805286,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.632,
      "step": 39792
    },
    {
      "epoch": 9.308268038825869,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6602,
      "step": 39793
    },
    {
      "epoch": 9.308501929598878,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 39794
    },
    {
      "epoch": 9.308735820371886,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5808,
      "step": 39795
    },
    {
      "epoch": 9.308969711144895,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6709,
      "step": 39796
    },
    {
      "epoch": 9.309203601917904,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 39797
    },
    {
      "epoch": 9.309437492690913,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 39798
    },
    {
      "epoch": 9.309671383463922,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 39799
    },
    {
      "epoch": 9.30990527423693,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 39800
    },
    {
      "epoch": 9.30990527423693,
      "eval_runtime": 4.6121,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 39800
    },
    {
      "epoch": 9.310139165009941,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 39801
    },
    {
      "epoch": 9.31037305578295,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7893,
      "step": 39802
    },
    {
      "epoch": 9.310606946555959,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 39803
    },
    {
      "epoch": 9.310840837328968,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4774,
      "step": 39804
    },
    {
      "epoch": 9.311074728101977,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5337,
      "step": 39805
    },
    {
      "epoch": 9.311308618874985,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 39806
    },
    {
      "epoch": 9.311542509647994,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 39807
    },
    {
      "epoch": 9.311776400421003,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6547,
      "step": 39808
    },
    {
      "epoch": 9.312010291194012,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.3847,
      "step": 39809
    },
    {
      "epoch": 9.31224418196702,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 39810
    },
    {
      "epoch": 9.31247807274003,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7572,
      "step": 39811
    },
    {
      "epoch": 9.31271196351304,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 39812
    },
    {
      "epoch": 9.312945854286049,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 39813
    },
    {
      "epoch": 9.313179745059058,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8693,
      "step": 39814
    },
    {
      "epoch": 9.313413635832067,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 39815
    },
    {
      "epoch": 9.313647526605076,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 39816
    },
    {
      "epoch": 9.313881417378084,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 39817
    },
    {
      "epoch": 9.314115308151093,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6176,
      "step": 39818
    },
    {
      "epoch": 9.314349198924102,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6229,
      "step": 39819
    },
    {
      "epoch": 9.314583089697111,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4167,
      "step": 39820
    },
    {
      "epoch": 9.31481698047012,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 39821
    },
    {
      "epoch": 9.31505087124313,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 39822
    },
    {
      "epoch": 9.31528476201614,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 39823
    },
    {
      "epoch": 9.315518652789148,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5185,
      "step": 39824
    },
    {
      "epoch": 9.315752543562157,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5649,
      "step": 39825
    },
    {
      "epoch": 9.315986434335166,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 39826
    },
    {
      "epoch": 9.316220325108175,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6695,
      "step": 39827
    },
    {
      "epoch": 9.316454215881183,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 39828
    },
    {
      "epoch": 9.316688106654192,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5407,
      "step": 39829
    },
    {
      "epoch": 9.316921997427201,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 39830
    },
    {
      "epoch": 9.31715588820021,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7113,
      "step": 39831
    },
    {
      "epoch": 9.317389778973219,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 39832
    },
    {
      "epoch": 9.31762366974623,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8315,
      "step": 39833
    },
    {
      "epoch": 9.317857560519238,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6756,
      "step": 39834
    },
    {
      "epoch": 9.318091451292247,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7198,
      "step": 39835
    },
    {
      "epoch": 9.318325342065256,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6247,
      "step": 39836
    },
    {
      "epoch": 9.318559232838265,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6475,
      "step": 39837
    },
    {
      "epoch": 9.318793123611274,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6372,
      "step": 39838
    },
    {
      "epoch": 9.319027014384282,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 39839
    },
    {
      "epoch": 9.319260905157291,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.6492,
      "step": 39840
    },
    {
      "epoch": 9.3194947959303,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5486,
      "step": 39841
    },
    {
      "epoch": 9.319728686703309,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5417,
      "step": 39842
    },
    {
      "epoch": 9.319962577476318,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6018,
      "step": 39843
    },
    {
      "epoch": 9.320196468249328,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 39844
    },
    {
      "epoch": 9.320430359022337,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 39845
    },
    {
      "epoch": 9.320664249795346,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 39846
    },
    {
      "epoch": 9.320898140568355,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 39847
    },
    {
      "epoch": 9.321132031341364,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.558,
      "step": 39848
    },
    {
      "epoch": 9.321365922114373,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4745,
      "step": 39849
    },
    {
      "epoch": 9.321599812887381,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 39850
    },
    {
      "epoch": 9.32183370366039,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 39851
    },
    {
      "epoch": 9.322067594433399,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 39852
    },
    {
      "epoch": 9.322301485206408,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5028,
      "step": 39853
    },
    {
      "epoch": 9.322535375979417,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 39854
    },
    {
      "epoch": 9.322769266752427,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 39855
    },
    {
      "epoch": 9.323003157525436,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5918,
      "step": 39856
    },
    {
      "epoch": 9.323237048298445,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 39857
    },
    {
      "epoch": 9.323470939071454,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4856,
      "step": 39858
    },
    {
      "epoch": 9.323704829844463,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 39859
    },
    {
      "epoch": 9.323938720617472,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.572,
      "step": 39860
    },
    {
      "epoch": 9.32417261139048,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 39861
    },
    {
      "epoch": 9.32440650216349,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6542,
      "step": 39862
    },
    {
      "epoch": 9.324640392936498,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7098,
      "step": 39863
    },
    {
      "epoch": 9.324874283709507,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 39864
    },
    {
      "epoch": 9.325108174482517,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 39865
    },
    {
      "epoch": 9.325342065255526,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 39866
    },
    {
      "epoch": 9.325575956028535,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 39867
    },
    {
      "epoch": 9.325809846801544,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 39868
    },
    {
      "epoch": 9.326043737574553,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 39869
    },
    {
      "epoch": 9.326277628347562,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.646,
      "step": 39870
    },
    {
      "epoch": 9.32651151912057,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 39871
    },
    {
      "epoch": 9.32674540989358,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4528,
      "step": 39872
    },
    {
      "epoch": 9.326979300666588,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 39873
    },
    {
      "epoch": 9.327213191439597,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 39874
    },
    {
      "epoch": 9.327447082212606,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 39875
    },
    {
      "epoch": 9.327680972985616,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 39876
    },
    {
      "epoch": 9.327914863758625,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 39877
    },
    {
      "epoch": 9.328148754531634,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 39878
    },
    {
      "epoch": 9.328382645304643,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 39879
    },
    {
      "epoch": 9.328616536077652,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 39880
    },
    {
      "epoch": 9.32885042685066,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 39881
    },
    {
      "epoch": 9.32908431762367,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 39882
    },
    {
      "epoch": 9.329318208396678,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 39883
    },
    {
      "epoch": 9.329552099169687,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 39884
    },
    {
      "epoch": 9.329785989942696,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 39885
    },
    {
      "epoch": 9.330019880715707,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4801,
      "step": 39886
    },
    {
      "epoch": 9.330253771488715,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6404,
      "step": 39887
    },
    {
      "epoch": 9.330487662261724,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 39888
    },
    {
      "epoch": 9.330721553034733,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7896,
      "step": 39889
    },
    {
      "epoch": 9.330955443807742,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 39890
    },
    {
      "epoch": 9.33118933458075,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 39891
    },
    {
      "epoch": 9.33142322535376,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.69,
      "step": 39892
    },
    {
      "epoch": 9.331657116126769,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7208,
      "step": 39893
    },
    {
      "epoch": 9.331891006899777,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7367,
      "step": 39894
    },
    {
      "epoch": 9.332124897672786,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8858,
      "step": 39895
    },
    {
      "epoch": 9.332358788445795,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6549,
      "step": 39896
    },
    {
      "epoch": 9.332592679218806,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.533,
      "step": 39897
    },
    {
      "epoch": 9.332826569991814,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 39898
    },
    {
      "epoch": 9.333060460764823,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 39899
    },
    {
      "epoch": 9.333294351537832,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 39900
    },
    {
      "epoch": 9.333294351537832,
      "eval_runtime": 4.609,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 39900
    },
    {
      "epoch": 9.333528242310841,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.3335,
      "step": 39901
    },
    {
      "epoch": 9.33376213308385,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 39902
    },
    {
      "epoch": 9.333996023856859,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 39903
    },
    {
      "epoch": 9.334229914629868,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 39904
    },
    {
      "epoch": 9.334463805402876,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 39905
    },
    {
      "epoch": 9.334697696175885,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 39906
    },
    {
      "epoch": 9.334931586948894,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6083,
      "step": 39907
    },
    {
      "epoch": 9.335165477721905,
      "grad_norm": 13.9375,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 39908
    },
    {
      "epoch": 9.335399368494913,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6756,
      "step": 39909
    },
    {
      "epoch": 9.335633259267922,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 39910
    },
    {
      "epoch": 9.335867150040931,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 39911
    },
    {
      "epoch": 9.33610104081394,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 39912
    },
    {
      "epoch": 9.336334931586949,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 39913
    },
    {
      "epoch": 9.336568822359958,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 39914
    },
    {
      "epoch": 9.336802713132967,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.5535,
      "step": 39915
    },
    {
      "epoch": 9.337036603905975,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 39916
    },
    {
      "epoch": 9.337270494678984,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7085,
      "step": 39917
    },
    {
      "epoch": 9.337504385451993,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8054,
      "step": 39918
    },
    {
      "epoch": 9.337738276225004,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 39919
    },
    {
      "epoch": 9.337972166998012,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4729,
      "step": 39920
    },
    {
      "epoch": 9.338206057771021,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 39921
    },
    {
      "epoch": 9.33843994854403,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 39922
    },
    {
      "epoch": 9.338673839317039,
      "grad_norm": 8.625,
      "learning_rate": 3e-05,
      "loss": 1.7634,
      "step": 39923
    },
    {
      "epoch": 9.338907730090048,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 39924
    },
    {
      "epoch": 9.339141620863057,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6051,
      "step": 39925
    },
    {
      "epoch": 9.339375511636065,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 39926
    },
    {
      "epoch": 9.339609402409074,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5308,
      "step": 39927
    },
    {
      "epoch": 9.339843293182083,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 39928
    },
    {
      "epoch": 9.340077183955094,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6528,
      "step": 39929
    },
    {
      "epoch": 9.340311074728103,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6824,
      "step": 39930
    },
    {
      "epoch": 9.340544965501111,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.4581,
      "step": 39931
    },
    {
      "epoch": 9.34077885627412,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6702,
      "step": 39932
    },
    {
      "epoch": 9.34101274704713,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7568,
      "step": 39933
    },
    {
      "epoch": 9.341246637820138,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6804,
      "step": 39934
    },
    {
      "epoch": 9.341480528593147,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 39935
    },
    {
      "epoch": 9.341714419366156,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.4041,
      "step": 39936
    },
    {
      "epoch": 9.341948310139164,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5942,
      "step": 39937
    },
    {
      "epoch": 9.342182200912173,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 39938
    },
    {
      "epoch": 9.342416091685182,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 39939
    },
    {
      "epoch": 9.342649982458193,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 39940
    },
    {
      "epoch": 9.342883873231202,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 39941
    },
    {
      "epoch": 9.34311776400421,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.609,
      "step": 39942
    },
    {
      "epoch": 9.34335165477722,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8257,
      "step": 39943
    },
    {
      "epoch": 9.343585545550228,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 39944
    },
    {
      "epoch": 9.343819436323237,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 39945
    },
    {
      "epoch": 9.344053327096246,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7846,
      "step": 39946
    },
    {
      "epoch": 9.344287217869255,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6368,
      "step": 39947
    },
    {
      "epoch": 9.344521108642263,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6464,
      "step": 39948
    },
    {
      "epoch": 9.344754999415272,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7071,
      "step": 39949
    },
    {
      "epoch": 9.344988890188283,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.542,
      "step": 39950
    },
    {
      "epoch": 9.345222780961292,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 39951
    },
    {
      "epoch": 9.3454566717343,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4943,
      "step": 39952
    },
    {
      "epoch": 9.34569056250731,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8027,
      "step": 39953
    },
    {
      "epoch": 9.345924453280318,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 39954
    },
    {
      "epoch": 9.346158344053327,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 39955
    },
    {
      "epoch": 9.346392234826336,
      "grad_norm": 7.21875,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 39956
    },
    {
      "epoch": 9.346626125599345,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7094,
      "step": 39957
    },
    {
      "epoch": 9.346860016372354,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 39958
    },
    {
      "epoch": 9.347093907145362,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.4121,
      "step": 39959
    },
    {
      "epoch": 9.347327797918371,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 39960
    },
    {
      "epoch": 9.347561688691382,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 39961
    },
    {
      "epoch": 9.34779557946439,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7477,
      "step": 39962
    },
    {
      "epoch": 9.3480294702374,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4185,
      "step": 39963
    },
    {
      "epoch": 9.348263361010408,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6416,
      "step": 39964
    },
    {
      "epoch": 9.348497251783417,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 39965
    },
    {
      "epoch": 9.348731142556426,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7126,
      "step": 39966
    },
    {
      "epoch": 9.348965033329435,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.63,
      "step": 39967
    },
    {
      "epoch": 9.349198924102444,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 39968
    },
    {
      "epoch": 9.349432814875453,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 39969
    },
    {
      "epoch": 9.349666705648461,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7053,
      "step": 39970
    },
    {
      "epoch": 9.34990059642147,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 39971
    },
    {
      "epoch": 9.350134487194481,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 39972
    },
    {
      "epoch": 9.35036837796749,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 39973
    },
    {
      "epoch": 9.350602268740499,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6833,
      "step": 39974
    },
    {
      "epoch": 9.350836159513507,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 39975
    },
    {
      "epoch": 9.351070050286516,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5242,
      "step": 39976
    },
    {
      "epoch": 9.351303941059525,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9087,
      "step": 39977
    },
    {
      "epoch": 9.351537831832534,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.4822,
      "step": 39978
    },
    {
      "epoch": 9.351771722605543,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 39979
    },
    {
      "epoch": 9.352005613378552,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5342,
      "step": 39980
    },
    {
      "epoch": 9.35223950415156,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 39981
    },
    {
      "epoch": 9.352473394924571,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8645,
      "step": 39982
    },
    {
      "epoch": 9.35270728569758,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6452,
      "step": 39983
    },
    {
      "epoch": 9.352941176470589,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 39984
    },
    {
      "epoch": 9.353175067243598,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 39985
    },
    {
      "epoch": 9.353408958016606,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 39986
    },
    {
      "epoch": 9.353642848789615,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.3929,
      "step": 39987
    },
    {
      "epoch": 9.353876739562624,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.564,
      "step": 39988
    },
    {
      "epoch": 9.354110630335633,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7394,
      "step": 39989
    },
    {
      "epoch": 9.354344521108642,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 39990
    },
    {
      "epoch": 9.35457841188165,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 39991
    },
    {
      "epoch": 9.35481230265466,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7699,
      "step": 39992
    },
    {
      "epoch": 9.35504619342767,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6713,
      "step": 39993
    },
    {
      "epoch": 9.355280084200679,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 39994
    },
    {
      "epoch": 9.355513974973688,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4888,
      "step": 39995
    },
    {
      "epoch": 9.355747865746697,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 39996
    },
    {
      "epoch": 9.355981756519705,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.794,
      "step": 39997
    },
    {
      "epoch": 9.356215647292714,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 39998
    },
    {
      "epoch": 9.356449538065723,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5636,
      "step": 39999
    },
    {
      "epoch": 9.356683428838732,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.654,
      "step": 40000
    },
    {
      "epoch": 9.356683428838732,
      "eval_runtime": 4.6406,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 40000
    },
    {
      "epoch": 9.35691731961174,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 40001
    },
    {
      "epoch": 9.35715121038475,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5906,
      "step": 40002
    },
    {
      "epoch": 9.35738510115776,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 40003
    },
    {
      "epoch": 9.357618991930769,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6982,
      "step": 40004
    },
    {
      "epoch": 9.357852882703778,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.4693,
      "step": 40005
    },
    {
      "epoch": 9.358086773476787,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6582,
      "step": 40006
    },
    {
      "epoch": 9.358320664249796,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6957,
      "step": 40007
    },
    {
      "epoch": 9.358554555022804,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 40008
    },
    {
      "epoch": 9.358788445795813,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 40009
    },
    {
      "epoch": 9.359022336568822,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 40010
    },
    {
      "epoch": 9.359256227341831,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7052,
      "step": 40011
    },
    {
      "epoch": 9.35949011811484,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 40012
    },
    {
      "epoch": 9.359724008887849,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 40013
    },
    {
      "epoch": 9.35995789966086,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 40014
    },
    {
      "epoch": 9.360191790433868,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 40015
    },
    {
      "epoch": 9.360425681206877,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6285,
      "step": 40016
    },
    {
      "epoch": 9.360659571979886,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 40017
    },
    {
      "epoch": 9.360893462752895,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 40018
    },
    {
      "epoch": 9.361127353525903,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6364,
      "step": 40019
    },
    {
      "epoch": 9.361361244298912,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6812,
      "step": 40020
    },
    {
      "epoch": 9.361595135071921,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8824,
      "step": 40021
    },
    {
      "epoch": 9.36182902584493,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 40022
    },
    {
      "epoch": 9.362062916617939,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.3675,
      "step": 40023
    },
    {
      "epoch": 9.362296807390948,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 40024
    },
    {
      "epoch": 9.362530698163958,
      "grad_norm": 7.96875,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 40025
    },
    {
      "epoch": 9.362764588936967,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5549,
      "step": 40026
    },
    {
      "epoch": 9.362998479709976,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 40027
    },
    {
      "epoch": 9.363232370482985,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6802,
      "step": 40028
    },
    {
      "epoch": 9.363466261255994,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 40029
    },
    {
      "epoch": 9.363700152029002,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6569,
      "step": 40030
    },
    {
      "epoch": 9.363934042802011,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6178,
      "step": 40031
    },
    {
      "epoch": 9.36416793357502,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5882,
      "step": 40032
    },
    {
      "epoch": 9.364401824348029,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.763,
      "step": 40033
    },
    {
      "epoch": 9.364635715121038,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 40034
    },
    {
      "epoch": 9.364869605894047,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 40035
    },
    {
      "epoch": 9.365103496667057,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6314,
      "step": 40036
    },
    {
      "epoch": 9.365337387440066,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 40037
    },
    {
      "epoch": 9.365571278213075,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 40038
    },
    {
      "epoch": 9.365805168986084,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6165,
      "step": 40039
    },
    {
      "epoch": 9.366039059759093,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3888,
      "step": 40040
    },
    {
      "epoch": 9.366272950532101,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.5367,
      "step": 40041
    },
    {
      "epoch": 9.36650684130511,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5967,
      "step": 40042
    },
    {
      "epoch": 9.366740732078119,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.736,
      "step": 40043
    },
    {
      "epoch": 9.366974622851128,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7634,
      "step": 40044
    },
    {
      "epoch": 9.367208513624137,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 40045
    },
    {
      "epoch": 9.367442404397147,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 40046
    },
    {
      "epoch": 9.367676295170156,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8195,
      "step": 40047
    },
    {
      "epoch": 9.367910185943165,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 40048
    },
    {
      "epoch": 9.368144076716174,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7064,
      "step": 40049
    },
    {
      "epoch": 9.368377967489183,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 40050
    },
    {
      "epoch": 9.368611858262192,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 40051
    },
    {
      "epoch": 9.3688457490352,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.3682,
      "step": 40052
    },
    {
      "epoch": 9.36907963980821,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4451,
      "step": 40053
    },
    {
      "epoch": 9.369313530581218,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 40054
    },
    {
      "epoch": 9.369547421354227,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 40055
    },
    {
      "epoch": 9.369781312127236,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 40056
    },
    {
      "epoch": 9.370015202900246,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 40057
    },
    {
      "epoch": 9.370249093673255,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5625,
      "step": 40058
    },
    {
      "epoch": 9.370482984446264,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5939,
      "step": 40059
    },
    {
      "epoch": 9.370716875219273,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 40060
    },
    {
      "epoch": 9.370950765992282,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 40061
    },
    {
      "epoch": 9.37118465676529,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 40062
    },
    {
      "epoch": 9.3714185475383,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.5878,
      "step": 40063
    },
    {
      "epoch": 9.371652438311308,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 40064
    },
    {
      "epoch": 9.371886329084317,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 40065
    },
    {
      "epoch": 9.372120219857326,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 40066
    },
    {
      "epoch": 9.372354110630337,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5882,
      "step": 40067
    },
    {
      "epoch": 9.372588001403345,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 40068
    },
    {
      "epoch": 9.372821892176354,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 40069
    },
    {
      "epoch": 9.373055782949363,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 40070
    },
    {
      "epoch": 9.373289673722372,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 40071
    },
    {
      "epoch": 9.37352356449538,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0172,
      "step": 40072
    },
    {
      "epoch": 9.37375745526839,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 40073
    },
    {
      "epoch": 9.373991346041398,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 40074
    },
    {
      "epoch": 9.374225236814407,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 40075
    },
    {
      "epoch": 9.374459127587416,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.65,
      "step": 40076
    },
    {
      "epoch": 9.374693018360425,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6943,
      "step": 40077
    },
    {
      "epoch": 9.374926909133436,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 40078
    },
    {
      "epoch": 9.375160799906444,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 40079
    },
    {
      "epoch": 9.375394690679453,
      "grad_norm": 10.375,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 40080
    },
    {
      "epoch": 9.375628581452462,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8382,
      "step": 40081
    },
    {
      "epoch": 9.37586247222547,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 40082
    },
    {
      "epoch": 9.37609636299848,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6236,
      "step": 40083
    },
    {
      "epoch": 9.376330253771489,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 40084
    },
    {
      "epoch": 9.376564144544497,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 40085
    },
    {
      "epoch": 9.376798035317506,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 40086
    },
    {
      "epoch": 9.377031926090515,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4931,
      "step": 40087
    },
    {
      "epoch": 9.377265816863524,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 40088
    },
    {
      "epoch": 9.377499707636535,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 40089
    },
    {
      "epoch": 9.377733598409543,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5802,
      "step": 40090
    },
    {
      "epoch": 9.377967489182552,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 40091
    },
    {
      "epoch": 9.378201379955561,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 40092
    },
    {
      "epoch": 9.37843527072857,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 40093
    },
    {
      "epoch": 9.378669161501579,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.5307,
      "step": 40094
    },
    {
      "epoch": 9.378903052274588,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 40095
    },
    {
      "epoch": 9.379136943047596,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 40096
    },
    {
      "epoch": 9.379370833820605,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 40097
    },
    {
      "epoch": 9.379604724593614,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 40098
    },
    {
      "epoch": 9.379838615366623,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5518,
      "step": 40099
    },
    {
      "epoch": 9.380072506139634,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 40100
    },
    {
      "epoch": 9.380072506139634,
      "eval_runtime": 4.6837,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 40100
    },
    {
      "epoch": 9.380306396912642,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 40101
    },
    {
      "epoch": 9.380540287685651,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7561,
      "step": 40102
    },
    {
      "epoch": 9.38077417845866,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 40103
    },
    {
      "epoch": 9.381008069231669,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6436,
      "step": 40104
    },
    {
      "epoch": 9.381241960004678,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 40105
    },
    {
      "epoch": 9.381475850777687,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6752,
      "step": 40106
    },
    {
      "epoch": 9.381709741550695,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 40107
    },
    {
      "epoch": 9.381943632323704,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7846,
      "step": 40108
    },
    {
      "epoch": 9.382177523096713,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 40109
    },
    {
      "epoch": 9.382411413869724,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 1.5061,
      "step": 40110
    },
    {
      "epoch": 9.382645304642732,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 40111
    },
    {
      "epoch": 9.382879195415741,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7158,
      "step": 40112
    },
    {
      "epoch": 9.38311308618875,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5804,
      "step": 40113
    },
    {
      "epoch": 9.383346976961759,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 40114
    },
    {
      "epoch": 9.383580867734768,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7682,
      "step": 40115
    },
    {
      "epoch": 9.383814758507777,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 40116
    },
    {
      "epoch": 9.384048649280786,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 40117
    },
    {
      "epoch": 9.384282540053794,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5269,
      "step": 40118
    },
    {
      "epoch": 9.384516430826803,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5071,
      "step": 40119
    },
    {
      "epoch": 9.384750321599812,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.7079,
      "step": 40120
    },
    {
      "epoch": 9.384984212372823,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6699,
      "step": 40121
    },
    {
      "epoch": 9.385218103145831,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 40122
    },
    {
      "epoch": 9.38545199391884,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 40123
    },
    {
      "epoch": 9.38568588469185,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 40124
    },
    {
      "epoch": 9.385919775464858,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5501,
      "step": 40125
    },
    {
      "epoch": 9.386153666237867,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 40126
    },
    {
      "epoch": 9.386387557010876,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 40127
    },
    {
      "epoch": 9.386621447783885,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7664,
      "step": 40128
    },
    {
      "epoch": 9.386855338556893,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4879,
      "step": 40129
    },
    {
      "epoch": 9.387089229329902,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 40130
    },
    {
      "epoch": 9.387323120102913,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6721,
      "step": 40131
    },
    {
      "epoch": 9.387557010875922,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 40132
    },
    {
      "epoch": 9.38779090164893,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 40133
    },
    {
      "epoch": 9.38802479242194,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6958,
      "step": 40134
    },
    {
      "epoch": 9.388258683194948,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6223,
      "step": 40135
    },
    {
      "epoch": 9.388492573967957,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 40136
    },
    {
      "epoch": 9.388726464740966,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7182,
      "step": 40137
    },
    {
      "epoch": 9.388960355513975,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 40138
    },
    {
      "epoch": 9.389194246286984,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5011,
      "step": 40139
    },
    {
      "epoch": 9.389428137059992,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 40140
    },
    {
      "epoch": 9.389662027833001,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 40141
    },
    {
      "epoch": 9.389895918606012,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6828,
      "step": 40142
    },
    {
      "epoch": 9.39012980937902,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7051,
      "step": 40143
    },
    {
      "epoch": 9.39036370015203,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5817,
      "step": 40144
    },
    {
      "epoch": 9.390597590925038,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7521,
      "step": 40145
    },
    {
      "epoch": 9.390831481698047,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6574,
      "step": 40146
    },
    {
      "epoch": 9.391065372471056,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5989,
      "step": 40147
    },
    {
      "epoch": 9.391299263244065,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7206,
      "step": 40148
    },
    {
      "epoch": 9.391533154017074,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 40149
    },
    {
      "epoch": 9.391767044790083,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 40150
    },
    {
      "epoch": 9.392000935563091,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 40151
    },
    {
      "epoch": 9.3922348263361,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 40152
    },
    {
      "epoch": 9.39246871710911,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6145,
      "step": 40153
    },
    {
      "epoch": 9.39270260788212,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 40154
    },
    {
      "epoch": 9.392936498655128,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 40155
    },
    {
      "epoch": 9.393170389428137,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 40156
    },
    {
      "epoch": 9.393404280201146,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 40157
    },
    {
      "epoch": 9.393638170974155,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 40158
    },
    {
      "epoch": 9.393872061747164,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 40159
    },
    {
      "epoch": 9.394105952520173,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5902,
      "step": 40160
    },
    {
      "epoch": 9.394339843293182,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.574,
      "step": 40161
    },
    {
      "epoch": 9.39457373406619,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 40162
    },
    {
      "epoch": 9.3948076248392,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 40163
    },
    {
      "epoch": 9.39504151561221,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 40164
    },
    {
      "epoch": 9.395275406385219,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 40165
    },
    {
      "epoch": 9.395509297158227,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 40166
    },
    {
      "epoch": 9.395743187931236,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 40167
    },
    {
      "epoch": 9.395977078704245,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5641,
      "step": 40168
    },
    {
      "epoch": 9.396210969477254,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8169,
      "step": 40169
    },
    {
      "epoch": 9.396444860250263,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 40170
    },
    {
      "epoch": 9.396678751023272,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 40171
    },
    {
      "epoch": 9.39691264179628,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 40172
    },
    {
      "epoch": 9.39714653256929,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 40173
    },
    {
      "epoch": 9.3973804233423,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7745,
      "step": 40174
    },
    {
      "epoch": 9.397614314115309,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6718,
      "step": 40175
    },
    {
      "epoch": 9.397848204888318,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 40176
    },
    {
      "epoch": 9.398082095661326,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 40177
    },
    {
      "epoch": 9.398315986434335,
      "grad_norm": 7.25,
      "learning_rate": 3e-05,
      "loss": 1.4952,
      "step": 40178
    },
    {
      "epoch": 9.398549877207344,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9461,
      "step": 40179
    },
    {
      "epoch": 9.398783767980353,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.5061,
      "step": 40180
    },
    {
      "epoch": 9.399017658753362,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 40181
    },
    {
      "epoch": 9.39925154952637,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6966,
      "step": 40182
    },
    {
      "epoch": 9.39948544029938,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5573,
      "step": 40183
    },
    {
      "epoch": 9.39971933107239,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 40184
    },
    {
      "epoch": 9.399953221845399,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 40185
    },
    {
      "epoch": 9.400187112618408,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4739,
      "step": 40186
    },
    {
      "epoch": 9.400421003391417,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6323,
      "step": 40187
    },
    {
      "epoch": 9.400654894164425,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.568,
      "step": 40188
    },
    {
      "epoch": 9.400888784937434,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 40189
    },
    {
      "epoch": 9.401122675710443,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.39,
      "step": 40190
    },
    {
      "epoch": 9.401356566483452,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 40191
    },
    {
      "epoch": 9.40159045725646,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7532,
      "step": 40192
    },
    {
      "epoch": 9.40182434802947,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 40193
    },
    {
      "epoch": 9.402058238802478,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 40194
    },
    {
      "epoch": 9.402292129575489,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 40195
    },
    {
      "epoch": 9.402526020348498,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7108,
      "step": 40196
    },
    {
      "epoch": 9.402759911121507,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 40197
    },
    {
      "epoch": 9.402993801894516,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 40198
    },
    {
      "epoch": 9.403227692667524,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.3633,
      "step": 40199
    },
    {
      "epoch": 9.403461583440533,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 40200
    },
    {
      "epoch": 9.403461583440533,
      "eval_runtime": 4.6292,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 40200
    },
    {
      "epoch": 9.403695474213542,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 40201
    },
    {
      "epoch": 9.403929364986551,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 40202
    },
    {
      "epoch": 9.40416325575956,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 40203
    },
    {
      "epoch": 9.404397146532569,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 40204
    },
    {
      "epoch": 9.404631037305577,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6346,
      "step": 40205
    },
    {
      "epoch": 9.404864928078588,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 40206
    },
    {
      "epoch": 9.405098818851597,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9184,
      "step": 40207
    },
    {
      "epoch": 9.405332709624606,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 40208
    },
    {
      "epoch": 9.405566600397615,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 40209
    },
    {
      "epoch": 9.405800491170623,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 40210
    },
    {
      "epoch": 9.406034381943632,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7588,
      "step": 40211
    },
    {
      "epoch": 9.406268272716641,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 40212
    },
    {
      "epoch": 9.40650216348965,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 40213
    },
    {
      "epoch": 9.406736054262659,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6665,
      "step": 40214
    },
    {
      "epoch": 9.406969945035668,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7222,
      "step": 40215
    },
    {
      "epoch": 9.407203835808676,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 40216
    },
    {
      "epoch": 9.407437726581687,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 40217
    },
    {
      "epoch": 9.407671617354696,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8242,
      "step": 40218
    },
    {
      "epoch": 9.407905508127705,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.57,
      "step": 40219
    },
    {
      "epoch": 9.408139398900714,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6573,
      "step": 40220
    },
    {
      "epoch": 9.408373289673722,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 40221
    },
    {
      "epoch": 9.408607180446731,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5813,
      "step": 40222
    },
    {
      "epoch": 9.40884107121974,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 40223
    },
    {
      "epoch": 9.409074961992749,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6237,
      "step": 40224
    },
    {
      "epoch": 9.409308852765758,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 40225
    },
    {
      "epoch": 9.409542743538767,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8176,
      "step": 40226
    },
    {
      "epoch": 9.409776634311777,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 40227
    },
    {
      "epoch": 9.410010525084786,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 40228
    },
    {
      "epoch": 9.410244415857795,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7274,
      "step": 40229
    },
    {
      "epoch": 9.410478306630804,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5828,
      "step": 40230
    },
    {
      "epoch": 9.410712197403813,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6521,
      "step": 40231
    },
    {
      "epoch": 9.410946088176821,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 40232
    },
    {
      "epoch": 9.41117997894983,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6568,
      "step": 40233
    },
    {
      "epoch": 9.411413869722839,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 40234
    },
    {
      "epoch": 9.411647760495848,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6234,
      "step": 40235
    },
    {
      "epoch": 9.411881651268857,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5325,
      "step": 40236
    },
    {
      "epoch": 9.412115542041866,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4803,
      "step": 40237
    },
    {
      "epoch": 9.412349432814876,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.758,
      "step": 40238
    },
    {
      "epoch": 9.412583323587885,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.381,
      "step": 40239
    },
    {
      "epoch": 9.412817214360894,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.776,
      "step": 40240
    },
    {
      "epoch": 9.413051105133903,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 40241
    },
    {
      "epoch": 9.413284995906912,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 40242
    },
    {
      "epoch": 9.41351888667992,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 40243
    },
    {
      "epoch": 9.41375277745293,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6268,
      "step": 40244
    },
    {
      "epoch": 9.413986668225938,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.4674,
      "step": 40245
    },
    {
      "epoch": 9.414220558998947,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5384,
      "step": 40246
    },
    {
      "epoch": 9.414454449771956,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 40247
    },
    {
      "epoch": 9.414688340544966,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5946,
      "step": 40248
    },
    {
      "epoch": 9.414922231317975,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 40249
    },
    {
      "epoch": 9.415156122090984,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 40250
    },
    {
      "epoch": 9.415390012863993,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7265,
      "step": 40251
    },
    {
      "epoch": 9.415623903637002,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 40252
    },
    {
      "epoch": 9.41585779441001,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7084,
      "step": 40253
    },
    {
      "epoch": 9.41609168518302,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 40254
    },
    {
      "epoch": 9.416325575956028,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8315,
      "step": 40255
    },
    {
      "epoch": 9.416559466729037,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 40256
    },
    {
      "epoch": 9.416793357502046,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4956,
      "step": 40257
    },
    {
      "epoch": 9.417027248275055,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6193,
      "step": 40258
    },
    {
      "epoch": 9.417261139048065,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 40259
    },
    {
      "epoch": 9.417495029821074,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 40260
    },
    {
      "epoch": 9.417728920594083,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 40261
    },
    {
      "epoch": 9.417962811367092,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 40262
    },
    {
      "epoch": 9.4181967021401,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 40263
    },
    {
      "epoch": 9.41843059291311,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 40264
    },
    {
      "epoch": 9.418664483686118,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 40265
    },
    {
      "epoch": 9.418898374459127,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 40266
    },
    {
      "epoch": 9.419132265232136,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.668,
      "step": 40267
    },
    {
      "epoch": 9.419366156005145,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6597,
      "step": 40268
    },
    {
      "epoch": 9.419600046778154,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 40269
    },
    {
      "epoch": 9.419833937551164,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 40270
    },
    {
      "epoch": 9.420067828324173,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 40271
    },
    {
      "epoch": 9.420301719097182,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 40272
    },
    {
      "epoch": 9.420535609870191,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.5224,
      "step": 40273
    },
    {
      "epoch": 9.4207695006432,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7004,
      "step": 40274
    },
    {
      "epoch": 9.421003391416209,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 40275
    },
    {
      "epoch": 9.421237282189217,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5261,
      "step": 40276
    },
    {
      "epoch": 9.421471172962226,
      "grad_norm": 8.3125,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 40277
    },
    {
      "epoch": 9.421705063735235,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 40278
    },
    {
      "epoch": 9.421938954508244,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7724,
      "step": 40279
    },
    {
      "epoch": 9.422172845281253,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 40280
    },
    {
      "epoch": 9.422406736054263,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5828,
      "step": 40281
    },
    {
      "epoch": 9.422640626827272,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 40282
    },
    {
      "epoch": 9.422874517600281,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 40283
    },
    {
      "epoch": 9.42310840837329,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 40284
    },
    {
      "epoch": 9.423342299146299,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 40285
    },
    {
      "epoch": 9.423576189919308,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5479,
      "step": 40286
    },
    {
      "epoch": 9.423810080692316,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 40287
    },
    {
      "epoch": 9.424043971465325,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 40288
    },
    {
      "epoch": 9.424277862238334,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6081,
      "step": 40289
    },
    {
      "epoch": 9.424511753011343,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 40290
    },
    {
      "epoch": 9.424745643784354,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.353,
      "step": 40291
    },
    {
      "epoch": 9.424979534557362,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6391,
      "step": 40292
    },
    {
      "epoch": 9.425213425330371,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7517,
      "step": 40293
    },
    {
      "epoch": 9.42544731610338,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 40294
    },
    {
      "epoch": 9.425681206876389,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.679,
      "step": 40295
    },
    {
      "epoch": 9.425915097649398,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6892,
      "step": 40296
    },
    {
      "epoch": 9.426148988422407,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7709,
      "step": 40297
    },
    {
      "epoch": 9.426382879195415,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6739,
      "step": 40298
    },
    {
      "epoch": 9.426616769968424,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.5998,
      "step": 40299
    },
    {
      "epoch": 9.426850660741433,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6999,
      "step": 40300
    },
    {
      "epoch": 9.426850660741433,
      "eval_runtime": 4.6299,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 40300
    },
    {
      "epoch": 9.427084551514442,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5146,
      "step": 40301
    },
    {
      "epoch": 9.427318442287453,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8019,
      "step": 40302
    },
    {
      "epoch": 9.427552333060461,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6002,
      "step": 40303
    },
    {
      "epoch": 9.42778622383347,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 40304
    },
    {
      "epoch": 9.428020114606479,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5723,
      "step": 40305
    },
    {
      "epoch": 9.428254005379488,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 40306
    },
    {
      "epoch": 9.428487896152497,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6494,
      "step": 40307
    },
    {
      "epoch": 9.428721786925506,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6801,
      "step": 40308
    },
    {
      "epoch": 9.428955677698514,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6633,
      "step": 40309
    },
    {
      "epoch": 9.429189568471523,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6743,
      "step": 40310
    },
    {
      "epoch": 9.429423459244532,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 40311
    },
    {
      "epoch": 9.429657350017543,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.7,
      "step": 40312
    },
    {
      "epoch": 9.429891240790552,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6246,
      "step": 40313
    },
    {
      "epoch": 9.43012513156356,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 40314
    },
    {
      "epoch": 9.43035902233657,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 40315
    },
    {
      "epoch": 9.430592913109578,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5362,
      "step": 40316
    },
    {
      "epoch": 9.430826803882587,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6787,
      "step": 40317
    },
    {
      "epoch": 9.431060694655596,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8461,
      "step": 40318
    },
    {
      "epoch": 9.431294585428605,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6572,
      "step": 40319
    },
    {
      "epoch": 9.431528476201613,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 40320
    },
    {
      "epoch": 9.431762366974622,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 40321
    },
    {
      "epoch": 9.431996257747631,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 40322
    },
    {
      "epoch": 9.432230148520642,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.4261,
      "step": 40323
    },
    {
      "epoch": 9.43246403929365,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 40324
    },
    {
      "epoch": 9.43269793006666,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6133,
      "step": 40325
    },
    {
      "epoch": 9.432931820839668,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 40326
    },
    {
      "epoch": 9.433165711612677,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 40327
    },
    {
      "epoch": 9.433399602385686,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 40328
    },
    {
      "epoch": 9.433633493158695,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 40329
    },
    {
      "epoch": 9.433867383931704,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 40330
    },
    {
      "epoch": 9.434101274704712,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 40331
    },
    {
      "epoch": 9.434335165477721,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 40332
    },
    {
      "epoch": 9.43456905625073,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 40333
    },
    {
      "epoch": 9.43480294702374,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 40334
    },
    {
      "epoch": 9.43503683779675,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5332,
      "step": 40335
    },
    {
      "epoch": 9.435270728569758,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 40336
    },
    {
      "epoch": 9.435504619342767,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6758,
      "step": 40337
    },
    {
      "epoch": 9.435738510115776,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 40338
    },
    {
      "epoch": 9.435972400888785,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 40339
    },
    {
      "epoch": 9.436206291661794,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9184,
      "step": 40340
    },
    {
      "epoch": 9.436440182434803,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5093,
      "step": 40341
    },
    {
      "epoch": 9.436674073207811,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 40342
    },
    {
      "epoch": 9.43690796398082,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 40343
    },
    {
      "epoch": 9.437141854753829,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5614,
      "step": 40344
    },
    {
      "epoch": 9.43737574552684,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 40345
    },
    {
      "epoch": 9.437609636299849,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6761,
      "step": 40346
    },
    {
      "epoch": 9.437843527072857,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 40347
    },
    {
      "epoch": 9.438077417845866,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 40348
    },
    {
      "epoch": 9.438311308618875,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 40349
    },
    {
      "epoch": 9.438545199391884,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 40350
    },
    {
      "epoch": 9.438779090164893,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5388,
      "step": 40351
    },
    {
      "epoch": 9.439012980937902,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6974,
      "step": 40352
    },
    {
      "epoch": 9.43924687171091,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 40353
    },
    {
      "epoch": 9.43948076248392,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4461,
      "step": 40354
    },
    {
      "epoch": 9.43971465325693,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4292,
      "step": 40355
    },
    {
      "epoch": 9.439948544029939,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6185,
      "step": 40356
    },
    {
      "epoch": 9.440182434802947,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4973,
      "step": 40357
    },
    {
      "epoch": 9.440416325575956,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 40358
    },
    {
      "epoch": 9.440650216348965,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5389,
      "step": 40359
    },
    {
      "epoch": 9.440884107121974,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.6961,
      "step": 40360
    },
    {
      "epoch": 9.441117997894983,
      "grad_norm": 10.1875,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 40361
    },
    {
      "epoch": 9.441351888667992,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 40362
    },
    {
      "epoch": 9.441585779441,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 40363
    },
    {
      "epoch": 9.44181967021401,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 40364
    },
    {
      "epoch": 9.44205356098702,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6943,
      "step": 40365
    },
    {
      "epoch": 9.442287451760029,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.501,
      "step": 40366
    },
    {
      "epoch": 9.442521342533038,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 40367
    },
    {
      "epoch": 9.442755233306046,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 40368
    },
    {
      "epoch": 9.442989124079055,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 40369
    },
    {
      "epoch": 9.443223014852064,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 40370
    },
    {
      "epoch": 9.443456905625073,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4814,
      "step": 40371
    },
    {
      "epoch": 9.443690796398082,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 40372
    },
    {
      "epoch": 9.44392468717109,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 40373
    },
    {
      "epoch": 9.4441585779441,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 40374
    },
    {
      "epoch": 9.444392468717108,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7352,
      "step": 40375
    },
    {
      "epoch": 9.444626359490119,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 40376
    },
    {
      "epoch": 9.444860250263128,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 40377
    },
    {
      "epoch": 9.445094141036137,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 40378
    },
    {
      "epoch": 9.445328031809145,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6103,
      "step": 40379
    },
    {
      "epoch": 9.445561922582154,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4507,
      "step": 40380
    },
    {
      "epoch": 9.445795813355163,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6507,
      "step": 40381
    },
    {
      "epoch": 9.446029704128172,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6385,
      "step": 40382
    },
    {
      "epoch": 9.44626359490118,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 40383
    },
    {
      "epoch": 9.44649748567419,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 40384
    },
    {
      "epoch": 9.446731376447199,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 40385
    },
    {
      "epoch": 9.446965267220207,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 40386
    },
    {
      "epoch": 9.447199157993218,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 40387
    },
    {
      "epoch": 9.447433048766227,
      "grad_norm": 10.4375,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 40388
    },
    {
      "epoch": 9.447666939539236,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6642,
      "step": 40389
    },
    {
      "epoch": 9.447900830312244,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6127,
      "step": 40390
    },
    {
      "epoch": 9.448134721085253,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 40391
    },
    {
      "epoch": 9.448368611858262,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6664,
      "step": 40392
    },
    {
      "epoch": 9.448602502631271,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 40393
    },
    {
      "epoch": 9.44883639340428,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8093,
      "step": 40394
    },
    {
      "epoch": 9.449070284177289,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 40395
    },
    {
      "epoch": 9.449304174950298,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6782,
      "step": 40396
    },
    {
      "epoch": 9.449538065723306,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 40397
    },
    {
      "epoch": 9.449771956496317,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 40398
    },
    {
      "epoch": 9.450005847269326,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 40399
    },
    {
      "epoch": 9.450239738042335,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 40400
    },
    {
      "epoch": 9.450239738042335,
      "eval_runtime": 4.6211,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 40400
    },
    {
      "epoch": 9.450473628815343,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 40401
    },
    {
      "epoch": 9.450707519588352,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 40402
    },
    {
      "epoch": 9.450941410361361,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.5069,
      "step": 40403
    },
    {
      "epoch": 9.45117530113437,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 40404
    },
    {
      "epoch": 9.451409191907379,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6918,
      "step": 40405
    },
    {
      "epoch": 9.451643082680388,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 40406
    },
    {
      "epoch": 9.451876973453397,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 40407
    },
    {
      "epoch": 9.452110864226407,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6444,
      "step": 40408
    },
    {
      "epoch": 9.452344754999416,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 40409
    },
    {
      "epoch": 9.452578645772425,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 40410
    },
    {
      "epoch": 9.452812536545434,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5534,
      "step": 40411
    },
    {
      "epoch": 9.453046427318442,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 40412
    },
    {
      "epoch": 9.453280318091451,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 40413
    },
    {
      "epoch": 9.45351420886446,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6151,
      "step": 40414
    },
    {
      "epoch": 9.453748099637469,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5118,
      "step": 40415
    },
    {
      "epoch": 9.453981990410478,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.4914,
      "step": 40416
    },
    {
      "epoch": 9.454215881183487,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 40417
    },
    {
      "epoch": 9.454449771956495,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8276,
      "step": 40418
    },
    {
      "epoch": 9.454683662729506,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 40419
    },
    {
      "epoch": 9.454917553502515,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 40420
    },
    {
      "epoch": 9.455151444275524,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 40421
    },
    {
      "epoch": 9.455385335048533,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6,
      "step": 40422
    },
    {
      "epoch": 9.455619225821541,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 40423
    },
    {
      "epoch": 9.45585311659455,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7468,
      "step": 40424
    },
    {
      "epoch": 9.45608700736756,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7666,
      "step": 40425
    },
    {
      "epoch": 9.456320898140568,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5747,
      "step": 40426
    },
    {
      "epoch": 9.456554788913577,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.669,
      "step": 40427
    },
    {
      "epoch": 9.456788679686586,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 40428
    },
    {
      "epoch": 9.457022570459596,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 40429
    },
    {
      "epoch": 9.457256461232605,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 40430
    },
    {
      "epoch": 9.457490352005614,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5457,
      "step": 40431
    },
    {
      "epoch": 9.457724242778623,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 40432
    },
    {
      "epoch": 9.457958133551632,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7083,
      "step": 40433
    },
    {
      "epoch": 9.45819202432464,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.6714,
      "step": 40434
    },
    {
      "epoch": 9.45842591509765,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.6661,
      "step": 40435
    },
    {
      "epoch": 9.458659805870658,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6154,
      "step": 40436
    },
    {
      "epoch": 9.458893696643667,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 40437
    },
    {
      "epoch": 9.459127587416676,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.4855,
      "step": 40438
    },
    {
      "epoch": 9.459361478189685,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.3137,
      "step": 40439
    },
    {
      "epoch": 9.459595368962695,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9002,
      "step": 40440
    },
    {
      "epoch": 9.459829259735704,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 40441
    },
    {
      "epoch": 9.460063150508713,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6708,
      "step": 40442
    },
    {
      "epoch": 9.460297041281722,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 40443
    },
    {
      "epoch": 9.46053093205473,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 40444
    },
    {
      "epoch": 9.46076482282774,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 40445
    },
    {
      "epoch": 9.460998713600748,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6748,
      "step": 40446
    },
    {
      "epoch": 9.461232604373757,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.5833,
      "step": 40447
    },
    {
      "epoch": 9.461466495146766,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.5699,
      "step": 40448
    },
    {
      "epoch": 9.461700385919775,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 40449
    },
    {
      "epoch": 9.461934276692784,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 40450
    },
    {
      "epoch": 9.462168167465794,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5426,
      "step": 40451
    },
    {
      "epoch": 9.462402058238803,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7951,
      "step": 40452
    },
    {
      "epoch": 9.462635949011812,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4928,
      "step": 40453
    },
    {
      "epoch": 9.46286983978482,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 40454
    },
    {
      "epoch": 9.46310373055783,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.3189,
      "step": 40455
    },
    {
      "epoch": 9.463337621330838,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 40456
    },
    {
      "epoch": 9.463571512103847,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 40457
    },
    {
      "epoch": 9.463805402876856,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6106,
      "step": 40458
    },
    {
      "epoch": 9.464039293649865,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 40459
    },
    {
      "epoch": 9.464273184422874,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 40460
    },
    {
      "epoch": 9.464507075195883,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 40461
    },
    {
      "epoch": 9.464740965968893,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6463,
      "step": 40462
    },
    {
      "epoch": 9.464974856741902,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8549,
      "step": 40463
    },
    {
      "epoch": 9.465208747514911,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7079,
      "step": 40464
    },
    {
      "epoch": 9.46544263828792,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 40465
    },
    {
      "epoch": 9.465676529060929,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 40466
    },
    {
      "epoch": 9.465910419833937,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 40467
    },
    {
      "epoch": 9.466144310606946,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5037,
      "step": 40468
    },
    {
      "epoch": 9.466378201379955,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7064,
      "step": 40469
    },
    {
      "epoch": 9.466612092152964,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 40470
    },
    {
      "epoch": 9.466845982925973,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 40471
    },
    {
      "epoch": 9.467079873698983,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8456,
      "step": 40472
    },
    {
      "epoch": 9.467313764471992,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6255,
      "step": 40473
    },
    {
      "epoch": 9.467547655245001,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 40474
    },
    {
      "epoch": 9.46778154601801,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 40475
    },
    {
      "epoch": 9.468015436791019,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7002,
      "step": 40476
    },
    {
      "epoch": 9.468249327564028,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7634,
      "step": 40477
    },
    {
      "epoch": 9.468483218337036,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6608,
      "step": 40478
    },
    {
      "epoch": 9.468717109110045,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6435,
      "step": 40479
    },
    {
      "epoch": 9.468950999883054,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 40480
    },
    {
      "epoch": 9.469184890656063,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 40481
    },
    {
      "epoch": 9.469418781429072,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7311,
      "step": 40482
    },
    {
      "epoch": 9.469652672202082,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5494,
      "step": 40483
    },
    {
      "epoch": 9.469886562975091,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6629,
      "step": 40484
    },
    {
      "epoch": 9.4701204537481,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 40485
    },
    {
      "epoch": 9.470354344521109,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 40486
    },
    {
      "epoch": 9.470588235294118,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 40487
    },
    {
      "epoch": 9.470822126067127,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 40488
    },
    {
      "epoch": 9.471056016840135,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.506,
      "step": 40489
    },
    {
      "epoch": 9.471289907613144,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.4286,
      "step": 40490
    },
    {
      "epoch": 9.471523798386153,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 40491
    },
    {
      "epoch": 9.471757689159162,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.3663,
      "step": 40492
    },
    {
      "epoch": 9.471991579932173,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 40493
    },
    {
      "epoch": 9.472225470705181,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.341,
      "step": 40494
    },
    {
      "epoch": 9.47245936147819,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6447,
      "step": 40495
    },
    {
      "epoch": 9.472693252251199,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6765,
      "step": 40496
    },
    {
      "epoch": 9.472927143024208,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 40497
    },
    {
      "epoch": 9.473161033797217,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 40498
    },
    {
      "epoch": 9.473394924570226,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 40499
    },
    {
      "epoch": 9.473628815343234,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 40500
    },
    {
      "epoch": 9.473628815343234,
      "eval_runtime": 4.6173,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 40500
    },
    {
      "epoch": 9.473862706116243,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 40501
    },
    {
      "epoch": 9.474096596889252,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.4684,
      "step": 40502
    },
    {
      "epoch": 9.474330487662261,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6026,
      "step": 40503
    },
    {
      "epoch": 9.474564378435272,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6571,
      "step": 40504
    },
    {
      "epoch": 9.47479826920828,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6453,
      "step": 40505
    },
    {
      "epoch": 9.47503215998129,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4557,
      "step": 40506
    },
    {
      "epoch": 9.475266050754298,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 40507
    },
    {
      "epoch": 9.475499941527307,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7275,
      "step": 40508
    },
    {
      "epoch": 9.475733832300316,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 40509
    },
    {
      "epoch": 9.475967723073325,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 40510
    },
    {
      "epoch": 9.476201613846333,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 40511
    },
    {
      "epoch": 9.476435504619342,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 40512
    },
    {
      "epoch": 9.476669395392351,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 40513
    },
    {
      "epoch": 9.47690328616536,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 40514
    },
    {
      "epoch": 9.47713717693837,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 40515
    },
    {
      "epoch": 9.47737106771138,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6194,
      "step": 40516
    },
    {
      "epoch": 9.477604958484388,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 40517
    },
    {
      "epoch": 9.477838849257397,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 40518
    },
    {
      "epoch": 9.478072740030406,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6115,
      "step": 40519
    },
    {
      "epoch": 9.478306630803415,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6262,
      "step": 40520
    },
    {
      "epoch": 9.478540521576424,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.731,
      "step": 40521
    },
    {
      "epoch": 9.478774412349432,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 40522
    },
    {
      "epoch": 9.479008303122441,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 40523
    },
    {
      "epoch": 9.47924219389545,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 40524
    },
    {
      "epoch": 9.479476084668459,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6761,
      "step": 40525
    },
    {
      "epoch": 9.47970997544147,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 40526
    },
    {
      "epoch": 9.479943866214478,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7653,
      "step": 40527
    },
    {
      "epoch": 9.480177756987487,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5733,
      "step": 40528
    },
    {
      "epoch": 9.480411647760496,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 40529
    },
    {
      "epoch": 9.480645538533505,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 40530
    },
    {
      "epoch": 9.480879429306514,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6294,
      "step": 40531
    },
    {
      "epoch": 9.481113320079523,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6362,
      "step": 40532
    },
    {
      "epoch": 9.481347210852531,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 40533
    },
    {
      "epoch": 9.48158110162554,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8554,
      "step": 40534
    },
    {
      "epoch": 9.481814992398549,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4356,
      "step": 40535
    },
    {
      "epoch": 9.48204888317156,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7781,
      "step": 40536
    },
    {
      "epoch": 9.482282773944569,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5833,
      "step": 40537
    },
    {
      "epoch": 9.482516664717577,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6178,
      "step": 40538
    },
    {
      "epoch": 9.482750555490586,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5863,
      "step": 40539
    },
    {
      "epoch": 9.482984446263595,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.2862,
      "step": 40540
    },
    {
      "epoch": 9.483218337036604,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5225,
      "step": 40541
    },
    {
      "epoch": 9.483452227809613,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 40542
    },
    {
      "epoch": 9.483686118582622,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5414,
      "step": 40543
    },
    {
      "epoch": 9.48392000935563,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7231,
      "step": 40544
    },
    {
      "epoch": 9.48415390012864,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 40545
    },
    {
      "epoch": 9.484387790901648,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 40546
    },
    {
      "epoch": 9.484621681674659,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5955,
      "step": 40547
    },
    {
      "epoch": 9.484855572447668,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7116,
      "step": 40548
    },
    {
      "epoch": 9.485089463220676,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 40549
    },
    {
      "epoch": 9.485323353993685,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 40550
    },
    {
      "epoch": 9.485557244766694,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5467,
      "step": 40551
    },
    {
      "epoch": 9.485791135539703,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 40552
    },
    {
      "epoch": 9.486025026312712,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6355,
      "step": 40553
    },
    {
      "epoch": 9.48625891708572,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.636,
      "step": 40554
    },
    {
      "epoch": 9.48649280785873,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 40555
    },
    {
      "epoch": 9.486726698631738,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7081,
      "step": 40556
    },
    {
      "epoch": 9.486960589404749,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 40557
    },
    {
      "epoch": 9.487194480177758,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8549,
      "step": 40558
    },
    {
      "epoch": 9.487428370950767,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 40559
    },
    {
      "epoch": 9.487662261723775,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5347,
      "step": 40560
    },
    {
      "epoch": 9.487896152496784,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 40561
    },
    {
      "epoch": 9.488130043269793,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6783,
      "step": 40562
    },
    {
      "epoch": 9.488363934042802,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6029,
      "step": 40563
    },
    {
      "epoch": 9.48859782481581,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 40564
    },
    {
      "epoch": 9.48883171558882,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6854,
      "step": 40565
    },
    {
      "epoch": 9.489065606361828,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 40566
    },
    {
      "epoch": 9.489299497134837,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6327,
      "step": 40567
    },
    {
      "epoch": 9.489533387907848,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 40568
    },
    {
      "epoch": 9.489767278680857,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 40569
    },
    {
      "epoch": 9.490001169453866,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 40570
    },
    {
      "epoch": 9.490235060226874,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 40571
    },
    {
      "epoch": 9.490468950999883,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5808,
      "step": 40572
    },
    {
      "epoch": 9.490702841772892,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5272,
      "step": 40573
    },
    {
      "epoch": 9.4909367325459,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5854,
      "step": 40574
    },
    {
      "epoch": 9.49117062331891,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 40575
    },
    {
      "epoch": 9.491404514091919,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 40576
    },
    {
      "epoch": 9.491638404864927,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 40577
    },
    {
      "epoch": 9.491872295637936,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7236,
      "step": 40578
    },
    {
      "epoch": 9.492106186410947,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6491,
      "step": 40579
    },
    {
      "epoch": 9.492340077183956,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6117,
      "step": 40580
    },
    {
      "epoch": 9.492573967956965,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 40581
    },
    {
      "epoch": 9.492807858729973,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 40582
    },
    {
      "epoch": 9.493041749502982,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6408,
      "step": 40583
    },
    {
      "epoch": 9.493275640275991,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6498,
      "step": 40584
    },
    {
      "epoch": 9.493509531049,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.6034,
      "step": 40585
    },
    {
      "epoch": 9.493743421822009,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 40586
    },
    {
      "epoch": 9.493977312595018,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5932,
      "step": 40587
    },
    {
      "epoch": 9.494211203368026,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 40588
    },
    {
      "epoch": 9.494445094141037,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7069,
      "step": 40589
    },
    {
      "epoch": 9.494678984914046,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7257,
      "step": 40590
    },
    {
      "epoch": 9.494912875687055,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5348,
      "step": 40591
    },
    {
      "epoch": 9.495146766460064,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 40592
    },
    {
      "epoch": 9.495380657233072,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.7943,
      "step": 40593
    },
    {
      "epoch": 9.495614548006081,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 40594
    },
    {
      "epoch": 9.49584843877909,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 40595
    },
    {
      "epoch": 9.496082329552099,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.6972,
      "step": 40596
    },
    {
      "epoch": 9.496316220325108,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 40597
    },
    {
      "epoch": 9.496550111098117,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5815,
      "step": 40598
    },
    {
      "epoch": 9.496784001871125,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 40599
    },
    {
      "epoch": 9.497017892644136,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7573,
      "step": 40600
    },
    {
      "epoch": 9.497017892644136,
      "eval_runtime": 4.6061,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 40600
    },
    {
      "epoch": 9.497251783417145,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 40601
    },
    {
      "epoch": 9.497485674190154,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 40602
    },
    {
      "epoch": 9.497719564963162,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 40603
    },
    {
      "epoch": 9.497953455736171,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.4718,
      "step": 40604
    },
    {
      "epoch": 9.49818734650918,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 40605
    },
    {
      "epoch": 9.498421237282189,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4913,
      "step": 40606
    },
    {
      "epoch": 9.498655128055198,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5367,
      "step": 40607
    },
    {
      "epoch": 9.498889018828207,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 40608
    },
    {
      "epoch": 9.499122909601216,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 40609
    },
    {
      "epoch": 9.499356800374226,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 40610
    },
    {
      "epoch": 9.499590691147235,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7838,
      "step": 40611
    },
    {
      "epoch": 9.499824581920244,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 40612
    },
    {
      "epoch": 9.500058472693253,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5484,
      "step": 40613
    },
    {
      "epoch": 9.500292363466261,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 40614
    },
    {
      "epoch": 9.50052625423927,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6146,
      "step": 40615
    },
    {
      "epoch": 9.50076014501228,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7141,
      "step": 40616
    },
    {
      "epoch": 9.500994035785288,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 40617
    },
    {
      "epoch": 9.501227926558297,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 40618
    },
    {
      "epoch": 9.501461817331306,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 40619
    },
    {
      "epoch": 9.501695708104315,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7548,
      "step": 40620
    },
    {
      "epoch": 9.501929598877325,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 40621
    },
    {
      "epoch": 9.502163489650334,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 40622
    },
    {
      "epoch": 9.502397380423343,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 40623
    },
    {
      "epoch": 9.502631271196352,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 40624
    },
    {
      "epoch": 9.50286516196936,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 40625
    },
    {
      "epoch": 9.50309905274237,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6782,
      "step": 40626
    },
    {
      "epoch": 9.503332943515378,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 40627
    },
    {
      "epoch": 9.503566834288387,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 40628
    },
    {
      "epoch": 9.503800725061396,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.6993,
      "step": 40629
    },
    {
      "epoch": 9.504034615834405,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6827,
      "step": 40630
    },
    {
      "epoch": 9.504268506607414,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 40631
    },
    {
      "epoch": 9.504502397380424,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 40632
    },
    {
      "epoch": 9.504736288153433,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 40633
    },
    {
      "epoch": 9.504970178926442,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8297,
      "step": 40634
    },
    {
      "epoch": 9.50520406969945,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 40635
    },
    {
      "epoch": 9.50543796047246,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.733,
      "step": 40636
    },
    {
      "epoch": 9.505671851245468,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5518,
      "step": 40637
    },
    {
      "epoch": 9.505905742018477,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.5948,
      "step": 40638
    },
    {
      "epoch": 9.506139632791486,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7657,
      "step": 40639
    },
    {
      "epoch": 9.506373523564495,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 40640
    },
    {
      "epoch": 9.506607414337504,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5863,
      "step": 40641
    },
    {
      "epoch": 9.506841305110513,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7094,
      "step": 40642
    },
    {
      "epoch": 9.507075195883523,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 40643
    },
    {
      "epoch": 9.507309086656532,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 40644
    },
    {
      "epoch": 9.50754297742954,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1204,
      "step": 40645
    },
    {
      "epoch": 9.50777686820255,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.5132,
      "step": 40646
    },
    {
      "epoch": 9.508010758975558,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7094,
      "step": 40647
    },
    {
      "epoch": 9.508244649748567,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5773,
      "step": 40648
    },
    {
      "epoch": 9.508478540521576,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5543,
      "step": 40649
    },
    {
      "epoch": 9.508712431294585,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 40650
    },
    {
      "epoch": 9.508946322067594,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 40651
    },
    {
      "epoch": 9.509180212840603,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4662,
      "step": 40652
    },
    {
      "epoch": 9.509414103613612,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4602,
      "step": 40653
    },
    {
      "epoch": 9.509647994386622,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 40654
    },
    {
      "epoch": 9.509881885159631,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6476,
      "step": 40655
    },
    {
      "epoch": 9.51011577593264,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 40656
    },
    {
      "epoch": 9.510349666705649,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7178,
      "step": 40657
    },
    {
      "epoch": 9.510583557478657,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 40658
    },
    {
      "epoch": 9.510817448251666,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7979,
      "step": 40659
    },
    {
      "epoch": 9.511051339024675,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5944,
      "step": 40660
    },
    {
      "epoch": 9.511285229797684,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 40661
    },
    {
      "epoch": 9.511519120570693,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 40662
    },
    {
      "epoch": 9.511753011343703,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6946,
      "step": 40663
    },
    {
      "epoch": 9.511986902116712,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5378,
      "step": 40664
    },
    {
      "epoch": 9.512220792889721,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5333,
      "step": 40665
    },
    {
      "epoch": 9.51245468366273,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 40666
    },
    {
      "epoch": 9.512688574435739,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5023,
      "step": 40667
    },
    {
      "epoch": 9.512922465208748,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.582,
      "step": 40668
    },
    {
      "epoch": 9.513156355981756,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7212,
      "step": 40669
    },
    {
      "epoch": 9.513390246754765,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.769,
      "step": 40670
    },
    {
      "epoch": 9.513624137527774,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6795,
      "step": 40671
    },
    {
      "epoch": 9.513858028300783,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.3986,
      "step": 40672
    },
    {
      "epoch": 9.514091919073792,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 40673
    },
    {
      "epoch": 9.514325809846802,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 40674
    },
    {
      "epoch": 9.514559700619811,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 40675
    },
    {
      "epoch": 9.51479359139282,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7592,
      "step": 40676
    },
    {
      "epoch": 9.515027482165829,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 40677
    },
    {
      "epoch": 9.515261372938838,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 40678
    },
    {
      "epoch": 9.515495263711847,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 40679
    },
    {
      "epoch": 9.515729154484855,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.29,
      "step": 40680
    },
    {
      "epoch": 9.515963045257864,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 40681
    },
    {
      "epoch": 9.516196936030873,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 40682
    },
    {
      "epoch": 9.516430826803882,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 40683
    },
    {
      "epoch": 9.51666471757689,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7511,
      "step": 40684
    },
    {
      "epoch": 9.516898608349901,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 40685
    },
    {
      "epoch": 9.51713249912291,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.6306,
      "step": 40686
    },
    {
      "epoch": 9.517366389895919,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5849,
      "step": 40687
    },
    {
      "epoch": 9.517600280668928,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 40688
    },
    {
      "epoch": 9.517834171441937,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4987,
      "step": 40689
    },
    {
      "epoch": 9.518068062214946,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 40690
    },
    {
      "epoch": 9.518301952987954,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5072,
      "step": 40691
    },
    {
      "epoch": 9.518535843760963,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 40692
    },
    {
      "epoch": 9.518769734533972,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 40693
    },
    {
      "epoch": 9.519003625306981,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 40694
    },
    {
      "epoch": 9.51923751607999,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7258,
      "step": 40695
    },
    {
      "epoch": 9.519471406853,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7708,
      "step": 40696
    },
    {
      "epoch": 9.51970529762601,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6493,
      "step": 40697
    },
    {
      "epoch": 9.519939188399018,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6816,
      "step": 40698
    },
    {
      "epoch": 9.520173079172027,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 40699
    },
    {
      "epoch": 9.520406969945036,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6613,
      "step": 40700
    },
    {
      "epoch": 9.520406969945036,
      "eval_runtime": 4.6156,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 40700
    },
    {
      "epoch": 9.520640860718045,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6551,
      "step": 40701
    },
    {
      "epoch": 9.520874751491053,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 40702
    },
    {
      "epoch": 9.521108642264062,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.6361,
      "step": 40703
    },
    {
      "epoch": 9.521342533037071,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5874,
      "step": 40704
    },
    {
      "epoch": 9.52157642381008,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 40705
    },
    {
      "epoch": 9.521810314583089,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 40706
    },
    {
      "epoch": 9.5220442053561,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6564,
      "step": 40707
    },
    {
      "epoch": 9.522278096129108,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 40708
    },
    {
      "epoch": 9.522511986902117,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 40709
    },
    {
      "epoch": 9.522745877675126,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6133,
      "step": 40710
    },
    {
      "epoch": 9.522979768448135,
      "grad_norm": 10.25,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 40711
    },
    {
      "epoch": 9.523213659221144,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6613,
      "step": 40712
    },
    {
      "epoch": 9.523447549994152,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 40713
    },
    {
      "epoch": 9.523681440767161,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6498,
      "step": 40714
    },
    {
      "epoch": 9.52391533154017,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6201,
      "step": 40715
    },
    {
      "epoch": 9.524149222313179,
      "grad_norm": 2.53125,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 40716
    },
    {
      "epoch": 9.52438311308619,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 40717
    },
    {
      "epoch": 9.524617003859198,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 40718
    },
    {
      "epoch": 9.524850894632207,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5654,
      "step": 40719
    },
    {
      "epoch": 9.525084785405216,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7202,
      "step": 40720
    },
    {
      "epoch": 9.525318676178225,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7928,
      "step": 40721
    },
    {
      "epoch": 9.525552566951234,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5374,
      "step": 40722
    },
    {
      "epoch": 9.525786457724243,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 40723
    },
    {
      "epoch": 9.526020348497251,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 40724
    },
    {
      "epoch": 9.52625423927026,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 40725
    },
    {
      "epoch": 9.526488130043269,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 40726
    },
    {
      "epoch": 9.52672202081628,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 40727
    },
    {
      "epoch": 9.526955911589289,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 40728
    },
    {
      "epoch": 9.527189802362297,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5927,
      "step": 40729
    },
    {
      "epoch": 9.527423693135306,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9241,
      "step": 40730
    },
    {
      "epoch": 9.527657583908315,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 40731
    },
    {
      "epoch": 9.527891474681324,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.708,
      "step": 40732
    },
    {
      "epoch": 9.528125365454333,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 40733
    },
    {
      "epoch": 9.528359256227342,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6894,
      "step": 40734
    },
    {
      "epoch": 9.52859314700035,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7086,
      "step": 40735
    },
    {
      "epoch": 9.52882703777336,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 40736
    },
    {
      "epoch": 9.529060928546368,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 40737
    },
    {
      "epoch": 9.529294819319379,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 40738
    },
    {
      "epoch": 9.529528710092388,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 40739
    },
    {
      "epoch": 9.529762600865396,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7192,
      "step": 40740
    },
    {
      "epoch": 9.529996491638405,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 40741
    },
    {
      "epoch": 9.530230382411414,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 40742
    },
    {
      "epoch": 9.530464273184423,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7616,
      "step": 40743
    },
    {
      "epoch": 9.530698163957432,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 40744
    },
    {
      "epoch": 9.53093205473044,
      "grad_norm": 8.8125,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 40745
    },
    {
      "epoch": 9.53116594550345,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.614,
      "step": 40746
    },
    {
      "epoch": 9.531399836276458,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 40747
    },
    {
      "epoch": 9.531633727049467,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5354,
      "step": 40748
    },
    {
      "epoch": 9.531867617822478,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 40749
    },
    {
      "epoch": 9.532101508595487,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7131,
      "step": 40750
    },
    {
      "epoch": 9.532335399368495,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 40751
    },
    {
      "epoch": 9.532569290141504,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.584,
      "step": 40752
    },
    {
      "epoch": 9.532803180914513,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 40753
    },
    {
      "epoch": 9.533037071687522,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 40754
    },
    {
      "epoch": 9.53327096246053,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 40755
    },
    {
      "epoch": 9.53350485323354,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.1485,
      "step": 40756
    },
    {
      "epoch": 9.533738744006548,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 40757
    },
    {
      "epoch": 9.533972634779557,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 40758
    },
    {
      "epoch": 9.534206525552566,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6887,
      "step": 40759
    },
    {
      "epoch": 9.534440416325577,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 40760
    },
    {
      "epoch": 9.534674307098586,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 40761
    },
    {
      "epoch": 9.534908197871594,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 40762
    },
    {
      "epoch": 9.535142088644603,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7893,
      "step": 40763
    },
    {
      "epoch": 9.535375979417612,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.699,
      "step": 40764
    },
    {
      "epoch": 9.535609870190621,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7024,
      "step": 40765
    },
    {
      "epoch": 9.53584376096363,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7418,
      "step": 40766
    },
    {
      "epoch": 9.536077651736639,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7527,
      "step": 40767
    },
    {
      "epoch": 9.536311542509647,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 40768
    },
    {
      "epoch": 9.536545433282656,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 40769
    },
    {
      "epoch": 9.536779324055665,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6151,
      "step": 40770
    },
    {
      "epoch": 9.537013214828676,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 40771
    },
    {
      "epoch": 9.537247105601685,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 40772
    },
    {
      "epoch": 9.537480996374693,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5155,
      "step": 40773
    },
    {
      "epoch": 9.537714887147702,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 40774
    },
    {
      "epoch": 9.537948777920711,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.522,
      "step": 40775
    },
    {
      "epoch": 9.53818266869372,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 40776
    },
    {
      "epoch": 9.538416559466729,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7051,
      "step": 40777
    },
    {
      "epoch": 9.538650450239738,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5932,
      "step": 40778
    },
    {
      "epoch": 9.538884341012746,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7337,
      "step": 40779
    },
    {
      "epoch": 9.539118231785755,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 40780
    },
    {
      "epoch": 9.539352122558766,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6314,
      "step": 40781
    },
    {
      "epoch": 9.539586013331775,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5985,
      "step": 40782
    },
    {
      "epoch": 9.539819904104784,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 40783
    },
    {
      "epoch": 9.540053794877792,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6546,
      "step": 40784
    },
    {
      "epoch": 9.540287685650801,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 40785
    },
    {
      "epoch": 9.54052157642381,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5669,
      "step": 40786
    },
    {
      "epoch": 9.540755467196819,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 40787
    },
    {
      "epoch": 9.540989357969828,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6442,
      "step": 40788
    },
    {
      "epoch": 9.541223248742837,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6347,
      "step": 40789
    },
    {
      "epoch": 9.541457139515845,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6893,
      "step": 40790
    },
    {
      "epoch": 9.541691030288856,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 40791
    },
    {
      "epoch": 9.541924921061865,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 40792
    },
    {
      "epoch": 9.542158811834874,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7295,
      "step": 40793
    },
    {
      "epoch": 9.542392702607883,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8183,
      "step": 40794
    },
    {
      "epoch": 9.542626593380891,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 40795
    },
    {
      "epoch": 9.5428604841539,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 40796
    },
    {
      "epoch": 9.543094374926909,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5667,
      "step": 40797
    },
    {
      "epoch": 9.543328265699918,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 40798
    },
    {
      "epoch": 9.543562156472927,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 40799
    },
    {
      "epoch": 9.543796047245936,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 40800
    },
    {
      "epoch": 9.543796047245936,
      "eval_runtime": 4.6349,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 40800
    },
    {
      "epoch": 9.544029938018944,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6677,
      "step": 40801
    },
    {
      "epoch": 9.544263828791955,
      "grad_norm": 8.1875,
      "learning_rate": 3e-05,
      "loss": 1.5464,
      "step": 40802
    },
    {
      "epoch": 9.544497719564964,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 40803
    },
    {
      "epoch": 9.544731610337973,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 40804
    },
    {
      "epoch": 9.544965501110982,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 40805
    },
    {
      "epoch": 9.54519939188399,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7048,
      "step": 40806
    },
    {
      "epoch": 9.545433282657,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 40807
    },
    {
      "epoch": 9.545667173430008,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2205,
      "step": 40808
    },
    {
      "epoch": 9.545901064203017,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7361,
      "step": 40809
    },
    {
      "epoch": 9.546134954976026,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.715,
      "step": 40810
    },
    {
      "epoch": 9.546368845749035,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 40811
    },
    {
      "epoch": 9.546602736522043,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 40812
    },
    {
      "epoch": 9.546836627295054,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.3025,
      "step": 40813
    },
    {
      "epoch": 9.547070518068063,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 40814
    },
    {
      "epoch": 9.547304408841072,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 40815
    },
    {
      "epoch": 9.54753829961408,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7519,
      "step": 40816
    },
    {
      "epoch": 9.54777219038709,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 40817
    },
    {
      "epoch": 9.548006081160098,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 40818
    },
    {
      "epoch": 9.548239971933107,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 40819
    },
    {
      "epoch": 9.548473862706116,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.3463,
      "step": 40820
    },
    {
      "epoch": 9.548707753479125,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 1.5201,
      "step": 40821
    },
    {
      "epoch": 9.548941644252134,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 40822
    },
    {
      "epoch": 9.549175535025142,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6777,
      "step": 40823
    },
    {
      "epoch": 9.549409425798153,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 40824
    },
    {
      "epoch": 9.549643316571162,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6002,
      "step": 40825
    },
    {
      "epoch": 9.54987720734417,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.684,
      "step": 40826
    },
    {
      "epoch": 9.55011109811718,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6252,
      "step": 40827
    },
    {
      "epoch": 9.550344988890188,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5147,
      "step": 40828
    },
    {
      "epoch": 9.550578879663197,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 40829
    },
    {
      "epoch": 9.550812770436206,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7365,
      "step": 40830
    },
    {
      "epoch": 9.551046661209215,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 40831
    },
    {
      "epoch": 9.551280551982224,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5869,
      "step": 40832
    },
    {
      "epoch": 9.551514442755233,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 40833
    },
    {
      "epoch": 9.551748333528241,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 40834
    },
    {
      "epoch": 9.551982224301252,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 40835
    },
    {
      "epoch": 9.55221611507426,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5593,
      "step": 40836
    },
    {
      "epoch": 9.55245000584727,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5097,
      "step": 40837
    },
    {
      "epoch": 9.552683896620279,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 40838
    },
    {
      "epoch": 9.552917787393287,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.6262,
      "step": 40839
    },
    {
      "epoch": 9.553151678166296,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 40840
    },
    {
      "epoch": 9.553385568939305,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.3928,
      "step": 40841
    },
    {
      "epoch": 9.553619459712314,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.5348,
      "step": 40842
    },
    {
      "epoch": 9.553853350485323,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 40843
    },
    {
      "epoch": 9.554087241258332,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.5445,
      "step": 40844
    },
    {
      "epoch": 9.554321132031342,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 40845
    },
    {
      "epoch": 9.554555022804351,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 40846
    },
    {
      "epoch": 9.55478891357736,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 40847
    },
    {
      "epoch": 9.555022804350369,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 40848
    },
    {
      "epoch": 9.555256695123377,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 40849
    },
    {
      "epoch": 9.555490585896386,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 40850
    },
    {
      "epoch": 9.555724476669395,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 40851
    },
    {
      "epoch": 9.555958367442404,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 40852
    },
    {
      "epoch": 9.556192258215413,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 40853
    },
    {
      "epoch": 9.556426148988422,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6302,
      "step": 40854
    },
    {
      "epoch": 9.556660039761432,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 40855
    },
    {
      "epoch": 9.556893930534441,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 40856
    },
    {
      "epoch": 9.55712782130745,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5926,
      "step": 40857
    },
    {
      "epoch": 9.557361712080459,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6191,
      "step": 40858
    },
    {
      "epoch": 9.557595602853468,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 40859
    },
    {
      "epoch": 9.557829493626476,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4422,
      "step": 40860
    },
    {
      "epoch": 9.558063384399485,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6864,
      "step": 40861
    },
    {
      "epoch": 9.558297275172494,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 40862
    },
    {
      "epoch": 9.558531165945503,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6095,
      "step": 40863
    },
    {
      "epoch": 9.558765056718512,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5839,
      "step": 40864
    },
    {
      "epoch": 9.55899894749152,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6032,
      "step": 40865
    },
    {
      "epoch": 9.559232838264531,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.605,
      "step": 40866
    },
    {
      "epoch": 9.55946672903754,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5224,
      "step": 40867
    },
    {
      "epoch": 9.559700619810549,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6818,
      "step": 40868
    },
    {
      "epoch": 9.559934510583558,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8135,
      "step": 40869
    },
    {
      "epoch": 9.560168401356567,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6381,
      "step": 40870
    },
    {
      "epoch": 9.560402292129575,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4435,
      "step": 40871
    },
    {
      "epoch": 9.560636182902584,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4969,
      "step": 40872
    },
    {
      "epoch": 9.560870073675593,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 40873
    },
    {
      "epoch": 9.561103964448602,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 40874
    },
    {
      "epoch": 9.56133785522161,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.4522,
      "step": 40875
    },
    {
      "epoch": 9.56157174599462,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7305,
      "step": 40876
    },
    {
      "epoch": 9.56180563676763,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.3562,
      "step": 40877
    },
    {
      "epoch": 9.56203952754064,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 40878
    },
    {
      "epoch": 9.562273418313648,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7387,
      "step": 40879
    },
    {
      "epoch": 9.562507309086657,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6433,
      "step": 40880
    },
    {
      "epoch": 9.562741199859666,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 40881
    },
    {
      "epoch": 9.562975090632674,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 40882
    },
    {
      "epoch": 9.563208981405683,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 40883
    },
    {
      "epoch": 9.563442872178692,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6866,
      "step": 40884
    },
    {
      "epoch": 9.563676762951701,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 40885
    },
    {
      "epoch": 9.56391065372471,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 40886
    },
    {
      "epoch": 9.564144544497719,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 40887
    },
    {
      "epoch": 9.56437843527073,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 40888
    },
    {
      "epoch": 9.564612326043738,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 40889
    },
    {
      "epoch": 9.564846216816747,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 40890
    },
    {
      "epoch": 9.565080107589756,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 40891
    },
    {
      "epoch": 9.565313998362765,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6748,
      "step": 40892
    },
    {
      "epoch": 9.565547889135773,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 40893
    },
    {
      "epoch": 9.565781779908782,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 40894
    },
    {
      "epoch": 9.566015670681791,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 40895
    },
    {
      "epoch": 9.5662495614548,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 40896
    },
    {
      "epoch": 9.566483452227809,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8625,
      "step": 40897
    },
    {
      "epoch": 9.56671734300082,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.622,
      "step": 40898
    },
    {
      "epoch": 9.566951233773828,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.73,
      "step": 40899
    },
    {
      "epoch": 9.567185124546837,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5169,
      "step": 40900
    },
    {
      "epoch": 9.567185124546837,
      "eval_runtime": 4.6078,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 40900
    },
    {
      "epoch": 9.567419015319846,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6011,
      "step": 40901
    },
    {
      "epoch": 9.567652906092855,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6901,
      "step": 40902
    },
    {
      "epoch": 9.567886796865864,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 40903
    },
    {
      "epoch": 9.568120687638872,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 40904
    },
    {
      "epoch": 9.568354578411881,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 40905
    },
    {
      "epoch": 9.56858846918489,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8223,
      "step": 40906
    },
    {
      "epoch": 9.568822359957899,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7468,
      "step": 40907
    },
    {
      "epoch": 9.56905625073091,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 40908
    },
    {
      "epoch": 9.569290141503918,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6829,
      "step": 40909
    },
    {
      "epoch": 9.569524032276927,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 40910
    },
    {
      "epoch": 9.569757923049936,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5656,
      "step": 40911
    },
    {
      "epoch": 9.569991813822945,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 40912
    },
    {
      "epoch": 9.570225704595954,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 40913
    },
    {
      "epoch": 9.570459595368963,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 40914
    },
    {
      "epoch": 9.570693486141971,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 40915
    },
    {
      "epoch": 9.57092737691498,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.3235,
      "step": 40916
    },
    {
      "epoch": 9.57116126768799,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 40917
    },
    {
      "epoch": 9.571395158460998,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 40918
    },
    {
      "epoch": 9.571629049234009,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 40919
    },
    {
      "epoch": 9.571862940007017,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 40920
    },
    {
      "epoch": 9.572096830780026,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 40921
    },
    {
      "epoch": 9.572330721553035,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7168,
      "step": 40922
    },
    {
      "epoch": 9.572564612326044,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.5467,
      "step": 40923
    },
    {
      "epoch": 9.572798503099053,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6432,
      "step": 40924
    },
    {
      "epoch": 9.573032393872062,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6342,
      "step": 40925
    },
    {
      "epoch": 9.57326628464507,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 40926
    },
    {
      "epoch": 9.57350017541808,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 40927
    },
    {
      "epoch": 9.573734066191088,
      "grad_norm": 8.375,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 40928
    },
    {
      "epoch": 9.573967956964097,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7959,
      "step": 40929
    },
    {
      "epoch": 9.574201847737108,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5982,
      "step": 40930
    },
    {
      "epoch": 9.574435738510116,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7037,
      "step": 40931
    },
    {
      "epoch": 9.574669629283125,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4846,
      "step": 40932
    },
    {
      "epoch": 9.574903520056134,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 40933
    },
    {
      "epoch": 9.575137410829143,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 40934
    },
    {
      "epoch": 9.575371301602152,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 40935
    },
    {
      "epoch": 9.57560519237516,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 40936
    },
    {
      "epoch": 9.57583908314817,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 40937
    },
    {
      "epoch": 9.576072973921178,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 40938
    },
    {
      "epoch": 9.576306864694187,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 40939
    },
    {
      "epoch": 9.576540755467196,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 1.6423,
      "step": 40940
    },
    {
      "epoch": 9.576774646240207,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 40941
    },
    {
      "epoch": 9.577008537013215,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 40942
    },
    {
      "epoch": 9.577242427786224,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 40943
    },
    {
      "epoch": 9.577476318559233,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 40944
    },
    {
      "epoch": 9.577710209332242,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6222,
      "step": 40945
    },
    {
      "epoch": 9.57794410010525,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 40946
    },
    {
      "epoch": 9.57817799087826,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5937,
      "step": 40947
    },
    {
      "epoch": 9.578411881651268,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 40948
    },
    {
      "epoch": 9.578645772424277,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4253,
      "step": 40949
    },
    {
      "epoch": 9.578879663197286,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 40950
    },
    {
      "epoch": 9.579113553970295,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 40951
    },
    {
      "epoch": 9.579347444743306,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.532,
      "step": 40952
    },
    {
      "epoch": 9.579581335516314,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 40953
    },
    {
      "epoch": 9.579815226289323,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 40954
    },
    {
      "epoch": 9.580049117062332,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 40955
    },
    {
      "epoch": 9.580283007835341,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 40956
    },
    {
      "epoch": 9.58051689860835,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 40957
    },
    {
      "epoch": 9.580750789381359,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5668,
      "step": 40958
    },
    {
      "epoch": 9.580984680154367,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 40959
    },
    {
      "epoch": 9.581218570927376,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6038,
      "step": 40960
    },
    {
      "epoch": 9.581452461700385,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7437,
      "step": 40961
    },
    {
      "epoch": 9.581686352473396,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6631,
      "step": 40962
    },
    {
      "epoch": 9.581920243246405,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 40963
    },
    {
      "epoch": 9.582154134019413,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 40964
    },
    {
      "epoch": 9.582388024792422,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 40965
    },
    {
      "epoch": 9.582621915565431,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 40966
    },
    {
      "epoch": 9.58285580633844,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 40967
    },
    {
      "epoch": 9.583089697111449,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7834,
      "step": 40968
    },
    {
      "epoch": 9.583323587884458,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4187,
      "step": 40969
    },
    {
      "epoch": 9.583557478657466,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6418,
      "step": 40970
    },
    {
      "epoch": 9.583791369430475,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5819,
      "step": 40971
    },
    {
      "epoch": 9.584025260203486,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 40972
    },
    {
      "epoch": 9.584259150976495,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 40973
    },
    {
      "epoch": 9.584493041749504,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 40974
    },
    {
      "epoch": 9.584726932522512,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7275,
      "step": 40975
    },
    {
      "epoch": 9.584960823295521,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 40976
    },
    {
      "epoch": 9.58519471406853,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.623,
      "step": 40977
    },
    {
      "epoch": 9.585428604841539,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9733,
      "step": 40978
    },
    {
      "epoch": 9.585662495614548,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 40979
    },
    {
      "epoch": 9.585896386387557,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 40980
    },
    {
      "epoch": 9.586130277160565,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 40981
    },
    {
      "epoch": 9.586364167933574,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 40982
    },
    {
      "epoch": 9.586598058706585,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 40983
    },
    {
      "epoch": 9.586831949479594,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 40984
    },
    {
      "epoch": 9.587065840252603,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 40985
    },
    {
      "epoch": 9.587299731025611,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 40986
    },
    {
      "epoch": 9.58753362179862,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7403,
      "step": 40987
    },
    {
      "epoch": 9.587767512571629,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 40988
    },
    {
      "epoch": 9.588001403344638,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 40989
    },
    {
      "epoch": 9.588235294117647,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7051,
      "step": 40990
    },
    {
      "epoch": 9.588469184890656,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 40991
    },
    {
      "epoch": 9.588703075663664,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 40992
    },
    {
      "epoch": 9.588936966436673,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6216,
      "step": 40993
    },
    {
      "epoch": 9.589170857209684,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8203,
      "step": 40994
    },
    {
      "epoch": 9.589404747982693,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 40995
    },
    {
      "epoch": 9.589638638755702,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7403,
      "step": 40996
    },
    {
      "epoch": 9.58987252952871,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 40997
    },
    {
      "epoch": 9.59010642030172,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6039,
      "step": 40998
    },
    {
      "epoch": 9.590340311074728,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.5558,
      "step": 40999
    },
    {
      "epoch": 9.590574201847737,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6706,
      "step": 41000
    },
    {
      "epoch": 9.590574201847737,
      "eval_runtime": 4.601,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 41000
    },
    {
      "epoch": 9.590808092620746,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 41001
    },
    {
      "epoch": 9.591041983393755,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 41002
    },
    {
      "epoch": 9.591275874166763,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 41003
    },
    {
      "epoch": 9.591509764939772,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 41004
    },
    {
      "epoch": 9.591743655712783,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5768,
      "step": 41005
    },
    {
      "epoch": 9.591977546485792,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7548,
      "step": 41006
    },
    {
      "epoch": 9.5922114372588,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 41007
    },
    {
      "epoch": 9.59244532803181,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 41008
    },
    {
      "epoch": 9.592679218804818,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5078,
      "step": 41009
    },
    {
      "epoch": 9.592913109577827,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 41010
    },
    {
      "epoch": 9.593147000350836,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 41011
    },
    {
      "epoch": 9.593380891123845,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.515,
      "step": 41012
    },
    {
      "epoch": 9.593614781896854,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 41013
    },
    {
      "epoch": 9.593848672669862,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6042,
      "step": 41014
    },
    {
      "epoch": 9.594082563442871,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7569,
      "step": 41015
    },
    {
      "epoch": 9.594316454215882,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 41016
    },
    {
      "epoch": 9.59455034498889,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 41017
    },
    {
      "epoch": 9.5947842357619,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7549,
      "step": 41018
    },
    {
      "epoch": 9.595018126534908,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 41019
    },
    {
      "epoch": 9.595252017307917,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4996,
      "step": 41020
    },
    {
      "epoch": 9.595485908080926,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6683,
      "step": 41021
    },
    {
      "epoch": 9.595719798853935,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5844,
      "step": 41022
    },
    {
      "epoch": 9.595953689626944,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7073,
      "step": 41023
    },
    {
      "epoch": 9.596187580399953,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.4763,
      "step": 41024
    },
    {
      "epoch": 9.596421471172961,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 41025
    },
    {
      "epoch": 9.596655361945972,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 41026
    },
    {
      "epoch": 9.59688925271898,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7812,
      "step": 41027
    },
    {
      "epoch": 9.59712314349199,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.731,
      "step": 41028
    },
    {
      "epoch": 9.597357034264999,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 41029
    },
    {
      "epoch": 9.597590925038007,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 41030
    },
    {
      "epoch": 9.597824815811016,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 41031
    },
    {
      "epoch": 9.598058706584025,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6583,
      "step": 41032
    },
    {
      "epoch": 9.598292597357034,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7099,
      "step": 41033
    },
    {
      "epoch": 9.598526488130043,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 41034
    },
    {
      "epoch": 9.598760378903052,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.3921,
      "step": 41035
    },
    {
      "epoch": 9.598994269676062,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6703,
      "step": 41036
    },
    {
      "epoch": 9.599228160449071,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 41037
    },
    {
      "epoch": 9.59946205122208,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 41038
    },
    {
      "epoch": 9.599695941995089,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 41039
    },
    {
      "epoch": 9.599929832768098,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8182,
      "step": 41040
    },
    {
      "epoch": 9.600163723541106,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6333,
      "step": 41041
    },
    {
      "epoch": 9.600397614314115,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7706,
      "step": 41042
    },
    {
      "epoch": 9.600631505087124,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5381,
      "step": 41043
    },
    {
      "epoch": 9.600865395860133,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6393,
      "step": 41044
    },
    {
      "epoch": 9.601099286633142,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 41045
    },
    {
      "epoch": 9.60133317740615,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.575,
      "step": 41046
    },
    {
      "epoch": 9.601567068179161,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7864,
      "step": 41047
    },
    {
      "epoch": 9.60180095895217,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 41048
    },
    {
      "epoch": 9.602034849725179,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6416,
      "step": 41049
    },
    {
      "epoch": 9.602268740498188,
      "grad_norm": 15.625,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 41050
    },
    {
      "epoch": 9.602502631271197,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 41051
    },
    {
      "epoch": 9.602736522044205,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 41052
    },
    {
      "epoch": 9.602970412817214,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9136,
      "step": 41053
    },
    {
      "epoch": 9.603204303590223,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 41054
    },
    {
      "epoch": 9.603438194363232,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7459,
      "step": 41055
    },
    {
      "epoch": 9.60367208513624,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 41056
    },
    {
      "epoch": 9.60390597590925,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6581,
      "step": 41057
    },
    {
      "epoch": 9.60413986668226,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6655,
      "step": 41058
    },
    {
      "epoch": 9.604373757455269,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.4688,
      "step": 41059
    },
    {
      "epoch": 9.604607648228278,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.633,
      "step": 41060
    },
    {
      "epoch": 9.604841539001287,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 41061
    },
    {
      "epoch": 9.605075429774296,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7152,
      "step": 41062
    },
    {
      "epoch": 9.605309320547304,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 41063
    },
    {
      "epoch": 9.605543211320313,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 41064
    },
    {
      "epoch": 9.605777102093322,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 41065
    },
    {
      "epoch": 9.60601099286633,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 41066
    },
    {
      "epoch": 9.60624488363934,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7528,
      "step": 41067
    },
    {
      "epoch": 9.606478774412349,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.5216,
      "step": 41068
    },
    {
      "epoch": 9.60671266518536,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 41069
    },
    {
      "epoch": 9.606946555958368,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6544,
      "step": 41070
    },
    {
      "epoch": 9.607180446731377,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 41071
    },
    {
      "epoch": 9.607414337504386,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.5575,
      "step": 41072
    },
    {
      "epoch": 9.607648228277395,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 41073
    },
    {
      "epoch": 9.607882119050403,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7855,
      "step": 41074
    },
    {
      "epoch": 9.608116009823412,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 41075
    },
    {
      "epoch": 9.608349900596421,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 41076
    },
    {
      "epoch": 9.60858379136943,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.722,
      "step": 41077
    },
    {
      "epoch": 9.608817682142439,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 41078
    },
    {
      "epoch": 9.60905157291545,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 41079
    },
    {
      "epoch": 9.609285463688458,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 41080
    },
    {
      "epoch": 9.609519354461467,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7692,
      "step": 41081
    },
    {
      "epoch": 9.609753245234476,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 41082
    },
    {
      "epoch": 9.609987136007485,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6284,
      "step": 41083
    },
    {
      "epoch": 9.610221026780494,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 41084
    },
    {
      "epoch": 9.610454917553502,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7468,
      "step": 41085
    },
    {
      "epoch": 9.610688808326511,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 41086
    },
    {
      "epoch": 9.61092269909952,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7859,
      "step": 41087
    },
    {
      "epoch": 9.611156589872529,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 41088
    },
    {
      "epoch": 9.61139048064554,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 41089
    },
    {
      "epoch": 9.611624371418548,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.732,
      "step": 41090
    },
    {
      "epoch": 9.611858262191557,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.4361,
      "step": 41091
    },
    {
      "epoch": 9.612092152964566,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7103,
      "step": 41092
    },
    {
      "epoch": 9.612326043737575,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.671,
      "step": 41093
    },
    {
      "epoch": 9.612559934510584,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 41094
    },
    {
      "epoch": 9.612793825283592,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5709,
      "step": 41095
    },
    {
      "epoch": 9.613027716056601,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 41096
    },
    {
      "epoch": 9.61326160682961,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6324,
      "step": 41097
    },
    {
      "epoch": 9.613495497602619,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 41098
    },
    {
      "epoch": 9.613729388375628,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 41099
    },
    {
      "epoch": 9.613963279148638,
      "grad_norm": 9.125,
      "learning_rate": 3e-05,
      "loss": 1.6834,
      "step": 41100
    },
    {
      "epoch": 9.613963279148638,
      "eval_runtime": 4.673,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 41100
    },
    {
      "epoch": 9.614197169921647,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 41101
    },
    {
      "epoch": 9.614431060694656,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 41102
    },
    {
      "epoch": 9.614664951467665,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5149,
      "step": 41103
    },
    {
      "epoch": 9.614898842240674,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4662,
      "step": 41104
    },
    {
      "epoch": 9.615132733013683,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 41105
    },
    {
      "epoch": 9.615366623786691,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 41106
    },
    {
      "epoch": 9.6156005145597,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 41107
    },
    {
      "epoch": 9.61583440533271,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 41108
    },
    {
      "epoch": 9.616068296105718,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 41109
    },
    {
      "epoch": 9.616302186878727,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 41110
    },
    {
      "epoch": 9.616536077651737,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7395,
      "step": 41111
    },
    {
      "epoch": 9.616769968424746,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5096,
      "step": 41112
    },
    {
      "epoch": 9.617003859197755,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4876,
      "step": 41113
    },
    {
      "epoch": 9.617237749970764,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7172,
      "step": 41114
    },
    {
      "epoch": 9.617471640743773,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 41115
    },
    {
      "epoch": 9.617705531516782,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8276,
      "step": 41116
    },
    {
      "epoch": 9.61793942228979,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 41117
    },
    {
      "epoch": 9.6181733130628,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5235,
      "step": 41118
    },
    {
      "epoch": 9.618407203835808,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 41119
    },
    {
      "epoch": 9.618641094608817,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.4929,
      "step": 41120
    },
    {
      "epoch": 9.618874985381826,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 41121
    },
    {
      "epoch": 9.619108876154836,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 41122
    },
    {
      "epoch": 9.619342766927845,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6382,
      "step": 41123
    },
    {
      "epoch": 9.619576657700854,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 41124
    },
    {
      "epoch": 9.619810548473863,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5133,
      "step": 41125
    },
    {
      "epoch": 9.620044439246872,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7295,
      "step": 41126
    },
    {
      "epoch": 9.62027833001988,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 41127
    },
    {
      "epoch": 9.62051222079289,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4085,
      "step": 41128
    },
    {
      "epoch": 9.620746111565898,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6472,
      "step": 41129
    },
    {
      "epoch": 9.620980002338907,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 41130
    },
    {
      "epoch": 9.621213893111916,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 41131
    },
    {
      "epoch": 9.621447783884925,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.499,
      "step": 41132
    },
    {
      "epoch": 9.621681674657935,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5574,
      "step": 41133
    },
    {
      "epoch": 9.621915565430944,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.6742,
      "step": 41134
    },
    {
      "epoch": 9.622149456203953,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 41135
    },
    {
      "epoch": 9.622383346976962,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 41136
    },
    {
      "epoch": 9.62261723774997,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7184,
      "step": 41137
    },
    {
      "epoch": 9.62285112852298,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 41138
    },
    {
      "epoch": 9.623085019295988,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 41139
    },
    {
      "epoch": 9.623318910068997,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5831,
      "step": 41140
    },
    {
      "epoch": 9.623552800842006,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6908,
      "step": 41141
    },
    {
      "epoch": 9.623786691615015,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5931,
      "step": 41142
    },
    {
      "epoch": 9.624020582388026,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 41143
    },
    {
      "epoch": 9.624254473161034,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 41144
    },
    {
      "epoch": 9.624488363934043,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 41145
    },
    {
      "epoch": 9.624722254707052,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 41146
    },
    {
      "epoch": 9.624956145480061,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7213,
      "step": 41147
    },
    {
      "epoch": 9.62519003625307,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 41148
    },
    {
      "epoch": 9.625423927026079,
      "grad_norm": 2.453125,
      "learning_rate": 3e-05,
      "loss": 1.4714,
      "step": 41149
    },
    {
      "epoch": 9.625657817799087,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 41150
    },
    {
      "epoch": 9.625891708572096,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6294,
      "step": 41151
    },
    {
      "epoch": 9.626125599345105,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 41152
    },
    {
      "epoch": 9.626359490118116,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 41153
    },
    {
      "epoch": 9.626593380891125,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 41154
    },
    {
      "epoch": 9.626827271664133,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7145,
      "step": 41155
    },
    {
      "epoch": 9.627061162437142,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 41156
    },
    {
      "epoch": 9.627295053210151,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 41157
    },
    {
      "epoch": 9.62752894398316,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 41158
    },
    {
      "epoch": 9.627762834756169,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6857,
      "step": 41159
    },
    {
      "epoch": 9.627996725529178,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 41160
    },
    {
      "epoch": 9.628230616302186,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7448,
      "step": 41161
    },
    {
      "epoch": 9.628464507075195,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4596,
      "step": 41162
    },
    {
      "epoch": 9.628698397848204,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 41163
    },
    {
      "epoch": 9.628932288621215,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 41164
    },
    {
      "epoch": 9.629166179394224,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 41165
    },
    {
      "epoch": 9.629400070167232,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7489,
      "step": 41166
    },
    {
      "epoch": 9.629633960940241,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 41167
    },
    {
      "epoch": 9.62986785171325,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 41168
    },
    {
      "epoch": 9.630101742486259,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 41169
    },
    {
      "epoch": 9.630335633259268,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.777,
      "step": 41170
    },
    {
      "epoch": 9.630569524032277,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 41171
    },
    {
      "epoch": 9.630803414805285,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2851,
      "step": 41172
    },
    {
      "epoch": 9.631037305578294,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 41173
    },
    {
      "epoch": 9.631271196351303,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.486,
      "step": 41174
    },
    {
      "epoch": 9.631505087124314,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6391,
      "step": 41175
    },
    {
      "epoch": 9.631738977897323,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 41176
    },
    {
      "epoch": 9.631972868670331,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 41177
    },
    {
      "epoch": 9.63220675944334,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6414,
      "step": 41178
    },
    {
      "epoch": 9.632440650216349,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 41179
    },
    {
      "epoch": 9.632674540989358,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 41180
    },
    {
      "epoch": 9.632908431762367,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 41181
    },
    {
      "epoch": 9.633142322535376,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 41182
    },
    {
      "epoch": 9.633376213308384,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7114,
      "step": 41183
    },
    {
      "epoch": 9.633610104081393,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 41184
    },
    {
      "epoch": 9.633843994854402,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 41185
    },
    {
      "epoch": 9.634077885627413,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 41186
    },
    {
      "epoch": 9.634311776400422,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 41187
    },
    {
      "epoch": 9.63454566717343,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4623,
      "step": 41188
    },
    {
      "epoch": 9.63477955794644,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 41189
    },
    {
      "epoch": 9.635013448719448,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 41190
    },
    {
      "epoch": 9.635247339492457,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 41191
    },
    {
      "epoch": 9.635481230265466,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7319,
      "step": 41192
    },
    {
      "epoch": 9.635715121038475,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 41193
    },
    {
      "epoch": 9.635949011811483,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 41194
    },
    {
      "epoch": 9.636182902584492,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5831,
      "step": 41195
    },
    {
      "epoch": 9.636416793357501,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 41196
    },
    {
      "epoch": 9.636650684130512,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6842,
      "step": 41197
    },
    {
      "epoch": 9.63688457490352,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 41198
    },
    {
      "epoch": 9.63711846567653,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7001,
      "step": 41199
    },
    {
      "epoch": 9.637352356449538,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 41200
    },
    {
      "epoch": 9.637352356449538,
      "eval_runtime": 4.6178,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 41200
    },
    {
      "epoch": 9.637586247222547,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6398,
      "step": 41201
    },
    {
      "epoch": 9.637820137995556,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.664,
      "step": 41202
    },
    {
      "epoch": 9.638054028768565,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8068,
      "step": 41203
    },
    {
      "epoch": 9.638287919541574,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5183,
      "step": 41204
    },
    {
      "epoch": 9.638521810314582,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6144,
      "step": 41205
    },
    {
      "epoch": 9.638755701087591,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.5879,
      "step": 41206
    },
    {
      "epoch": 9.638989591860602,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 41207
    },
    {
      "epoch": 9.63922348263361,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 41208
    },
    {
      "epoch": 9.63945737340662,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 41209
    },
    {
      "epoch": 9.639691264179628,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7131,
      "step": 41210
    },
    {
      "epoch": 9.639925154952637,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 41211
    },
    {
      "epoch": 9.640159045725646,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 41212
    },
    {
      "epoch": 9.640392936498655,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.5013,
      "step": 41213
    },
    {
      "epoch": 9.640626827271664,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 41214
    },
    {
      "epoch": 9.640860718044673,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 41215
    },
    {
      "epoch": 9.641094608817681,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6308,
      "step": 41216
    },
    {
      "epoch": 9.641328499590692,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6605,
      "step": 41217
    },
    {
      "epoch": 9.6415623903637,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 41218
    },
    {
      "epoch": 9.64179628113671,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 41219
    },
    {
      "epoch": 9.642030171909719,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8043,
      "step": 41220
    },
    {
      "epoch": 9.642264062682727,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 41221
    },
    {
      "epoch": 9.642497953455736,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 41222
    },
    {
      "epoch": 9.642731844228745,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 41223
    },
    {
      "epoch": 9.642965735001754,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8161,
      "step": 41224
    },
    {
      "epoch": 9.643199625774763,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2809,
      "step": 41225
    },
    {
      "epoch": 9.643433516547772,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 41226
    },
    {
      "epoch": 9.64366740732078,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 41227
    },
    {
      "epoch": 9.643901298093791,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7609,
      "step": 41228
    },
    {
      "epoch": 9.6441351888668,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6317,
      "step": 41229
    },
    {
      "epoch": 9.644369079639809,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 41230
    },
    {
      "epoch": 9.644602970412818,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6951,
      "step": 41231
    },
    {
      "epoch": 9.644836861185826,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 41232
    },
    {
      "epoch": 9.645070751958835,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 41233
    },
    {
      "epoch": 9.645304642731844,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7526,
      "step": 41234
    },
    {
      "epoch": 9.645538533504853,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5892,
      "step": 41235
    },
    {
      "epoch": 9.645772424277862,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7673,
      "step": 41236
    },
    {
      "epoch": 9.64600631505087,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7088,
      "step": 41237
    },
    {
      "epoch": 9.64624020582388,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6116,
      "step": 41238
    },
    {
      "epoch": 9.64647409659689,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6769,
      "step": 41239
    },
    {
      "epoch": 9.646707987369899,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 41240
    },
    {
      "epoch": 9.646941878142908,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8359,
      "step": 41241
    },
    {
      "epoch": 9.647175768915917,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 41242
    },
    {
      "epoch": 9.647409659688925,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6007,
      "step": 41243
    },
    {
      "epoch": 9.647643550461934,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7254,
      "step": 41244
    },
    {
      "epoch": 9.647877441234943,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7396,
      "step": 41245
    },
    {
      "epoch": 9.648111332007952,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.74,
      "step": 41246
    },
    {
      "epoch": 9.64834522278096,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4733,
      "step": 41247
    },
    {
      "epoch": 9.64857911355397,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 41248
    },
    {
      "epoch": 9.648813004326978,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4245,
      "step": 41249
    },
    {
      "epoch": 9.649046895099989,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 41250
    },
    {
      "epoch": 9.649280785872998,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 41251
    },
    {
      "epoch": 9.649514676646007,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 41252
    },
    {
      "epoch": 9.649748567419016,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 41253
    },
    {
      "epoch": 9.649982458192024,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5578,
      "step": 41254
    },
    {
      "epoch": 9.650216348965033,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 41255
    },
    {
      "epoch": 9.650450239738042,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 41256
    },
    {
      "epoch": 9.650684130511051,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 41257
    },
    {
      "epoch": 9.65091802128406,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7172,
      "step": 41258
    },
    {
      "epoch": 9.651151912057069,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5777,
      "step": 41259
    },
    {
      "epoch": 9.65138580283008,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 41260
    },
    {
      "epoch": 9.651619693603088,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5707,
      "step": 41261
    },
    {
      "epoch": 9.651853584376097,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 41262
    },
    {
      "epoch": 9.652087475149106,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7354,
      "step": 41263
    },
    {
      "epoch": 9.652321365922115,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 41264
    },
    {
      "epoch": 9.652555256695123,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7035,
      "step": 41265
    },
    {
      "epoch": 9.652789147468132,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4936,
      "step": 41266
    },
    {
      "epoch": 9.653023038241141,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5568,
      "step": 41267
    },
    {
      "epoch": 9.65325692901415,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 41268
    },
    {
      "epoch": 9.653490819787159,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7218,
      "step": 41269
    },
    {
      "epoch": 9.65372471056017,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6624,
      "step": 41270
    },
    {
      "epoch": 9.653958601333178,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5388,
      "step": 41271
    },
    {
      "epoch": 9.654192492106187,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 41272
    },
    {
      "epoch": 9.654426382879196,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6725,
      "step": 41273
    },
    {
      "epoch": 9.654660273652205,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 41274
    },
    {
      "epoch": 9.654894164425214,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6912,
      "step": 41275
    },
    {
      "epoch": 9.655128055198222,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 41276
    },
    {
      "epoch": 9.655361945971231,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 41277
    },
    {
      "epoch": 9.65559583674424,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 41278
    },
    {
      "epoch": 9.655829727517249,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 41279
    },
    {
      "epoch": 9.656063618290258,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.619,
      "step": 41280
    },
    {
      "epoch": 9.656297509063268,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4931,
      "step": 41281
    },
    {
      "epoch": 9.656531399836277,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 41282
    },
    {
      "epoch": 9.656765290609286,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 41283
    },
    {
      "epoch": 9.656999181382295,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.4396,
      "step": 41284
    },
    {
      "epoch": 9.657233072155304,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 1.3044,
      "step": 41285
    },
    {
      "epoch": 9.657466962928313,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 41286
    },
    {
      "epoch": 9.657700853701321,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.5231,
      "step": 41287
    },
    {
      "epoch": 9.65793474447433,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 41288
    },
    {
      "epoch": 9.658168635247339,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7561,
      "step": 41289
    },
    {
      "epoch": 9.658402526020348,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5682,
      "step": 41290
    },
    {
      "epoch": 9.658636416793357,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.4024,
      "step": 41291
    },
    {
      "epoch": 9.658870307566367,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 41292
    },
    {
      "epoch": 9.659104198339376,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5935,
      "step": 41293
    },
    {
      "epoch": 9.659338089112385,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 41294
    },
    {
      "epoch": 9.659571979885394,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 41295
    },
    {
      "epoch": 9.659805870658403,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6638,
      "step": 41296
    },
    {
      "epoch": 9.660039761431412,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6525,
      "step": 41297
    },
    {
      "epoch": 9.66027365220442,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4875,
      "step": 41298
    },
    {
      "epoch": 9.66050754297743,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 41299
    },
    {
      "epoch": 9.660741433750438,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.5831,
      "step": 41300
    },
    {
      "epoch": 9.660741433750438,
      "eval_runtime": 4.6608,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 41300
    },
    {
      "epoch": 9.660975324523447,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7089,
      "step": 41301
    },
    {
      "epoch": 9.661209215296456,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 41302
    },
    {
      "epoch": 9.661443106069466,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 41303
    },
    {
      "epoch": 9.661676996842475,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 41304
    },
    {
      "epoch": 9.661910887615484,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4931,
      "step": 41305
    },
    {
      "epoch": 9.662144778388493,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 41306
    },
    {
      "epoch": 9.662378669161502,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4665,
      "step": 41307
    },
    {
      "epoch": 9.66261255993451,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 41308
    },
    {
      "epoch": 9.66284645070752,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.3621,
      "step": 41309
    },
    {
      "epoch": 9.663080341480528,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 41310
    },
    {
      "epoch": 9.663314232253537,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 41311
    },
    {
      "epoch": 9.663548123026546,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7641,
      "step": 41312
    },
    {
      "epoch": 9.663782013799555,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 41313
    },
    {
      "epoch": 9.664015904572565,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 41314
    },
    {
      "epoch": 9.664249795345574,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 41315
    },
    {
      "epoch": 9.664483686118583,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 41316
    },
    {
      "epoch": 9.664717576891592,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 41317
    },
    {
      "epoch": 9.6649514676646,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 41318
    },
    {
      "epoch": 9.66518535843761,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 41319
    },
    {
      "epoch": 9.665419249210618,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 41320
    },
    {
      "epoch": 9.665653139983627,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.716,
      "step": 41321
    },
    {
      "epoch": 9.665887030756636,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 41322
    },
    {
      "epoch": 9.666120921529645,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 41323
    },
    {
      "epoch": 9.666354812302655,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6596,
      "step": 41324
    },
    {
      "epoch": 9.666588703075664,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 41325
    },
    {
      "epoch": 9.666822593848673,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5995,
      "step": 41326
    },
    {
      "epoch": 9.667056484621682,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 1.7526,
      "step": 41327
    },
    {
      "epoch": 9.66729037539469,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 41328
    },
    {
      "epoch": 9.6675242661677,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6386,
      "step": 41329
    },
    {
      "epoch": 9.667758156940709,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 41330
    },
    {
      "epoch": 9.667992047713717,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7965,
      "step": 41331
    },
    {
      "epoch": 9.668225938486726,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 41332
    },
    {
      "epoch": 9.668459829259735,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6309,
      "step": 41333
    },
    {
      "epoch": 9.668693720032746,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.788,
      "step": 41334
    },
    {
      "epoch": 9.668927610805754,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 41335
    },
    {
      "epoch": 9.669161501578763,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 41336
    },
    {
      "epoch": 9.669395392351772,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7341,
      "step": 41337
    },
    {
      "epoch": 9.669629283124781,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6517,
      "step": 41338
    },
    {
      "epoch": 9.66986317389779,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 41339
    },
    {
      "epoch": 9.670097064670799,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 41340
    },
    {
      "epoch": 9.670330955443807,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6863,
      "step": 41341
    },
    {
      "epoch": 9.670564846216816,
      "grad_norm": 7.75,
      "learning_rate": 3e-05,
      "loss": 1.6327,
      "step": 41342
    },
    {
      "epoch": 9.670798736989825,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 41343
    },
    {
      "epoch": 9.671032627762834,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.677,
      "step": 41344
    },
    {
      "epoch": 9.671266518535845,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6144,
      "step": 41345
    },
    {
      "epoch": 9.671500409308853,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 41346
    },
    {
      "epoch": 9.671734300081862,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7403,
      "step": 41347
    },
    {
      "epoch": 9.671968190854871,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5774,
      "step": 41348
    },
    {
      "epoch": 9.67220208162788,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 41349
    },
    {
      "epoch": 9.672435972400889,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6392,
      "step": 41350
    },
    {
      "epoch": 9.672669863173898,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.5853,
      "step": 41351
    },
    {
      "epoch": 9.672903753946906,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5989,
      "step": 41352
    },
    {
      "epoch": 9.673137644719915,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 41353
    },
    {
      "epoch": 9.673371535492924,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9787,
      "step": 41354
    },
    {
      "epoch": 9.673605426265933,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 41355
    },
    {
      "epoch": 9.673839317038944,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6318,
      "step": 41356
    },
    {
      "epoch": 9.674073207811952,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 41357
    },
    {
      "epoch": 9.674307098584961,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8165,
      "step": 41358
    },
    {
      "epoch": 9.67454098935797,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5174,
      "step": 41359
    },
    {
      "epoch": 9.674774880130979,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 41360
    },
    {
      "epoch": 9.675008770903988,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 41361
    },
    {
      "epoch": 9.675242661676997,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6497,
      "step": 41362
    },
    {
      "epoch": 9.675476552450005,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5302,
      "step": 41363
    },
    {
      "epoch": 9.675710443223014,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5436,
      "step": 41364
    },
    {
      "epoch": 9.675944333996023,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 41365
    },
    {
      "epoch": 9.676178224769032,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7287,
      "step": 41366
    },
    {
      "epoch": 9.676412115542043,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6253,
      "step": 41367
    },
    {
      "epoch": 9.676646006315051,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6698,
      "step": 41368
    },
    {
      "epoch": 9.67687989708806,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 41369
    },
    {
      "epoch": 9.67711378786107,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.694,
      "step": 41370
    },
    {
      "epoch": 9.677347678634078,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.5399,
      "step": 41371
    },
    {
      "epoch": 9.677581569407087,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 41372
    },
    {
      "epoch": 9.677815460180096,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 1.64,
      "step": 41373
    },
    {
      "epoch": 9.678049350953104,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 41374
    },
    {
      "epoch": 9.678283241726113,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.676,
      "step": 41375
    },
    {
      "epoch": 9.678517132499122,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5358,
      "step": 41376
    },
    {
      "epoch": 9.678751023272131,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.4531,
      "step": 41377
    },
    {
      "epoch": 9.678984914045142,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 41378
    },
    {
      "epoch": 9.67921880481815,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 41379
    },
    {
      "epoch": 9.67945269559116,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 41380
    },
    {
      "epoch": 9.679686586364168,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7553,
      "step": 41381
    },
    {
      "epoch": 9.679920477137177,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 41382
    },
    {
      "epoch": 9.680154367910186,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7741,
      "step": 41383
    },
    {
      "epoch": 9.680388258683195,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 41384
    },
    {
      "epoch": 9.680622149456203,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 41385
    },
    {
      "epoch": 9.680856040229212,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5718,
      "step": 41386
    },
    {
      "epoch": 9.681089931002221,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8175,
      "step": 41387
    },
    {
      "epoch": 9.681323821775232,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6068,
      "step": 41388
    },
    {
      "epoch": 9.68155771254824,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 41389
    },
    {
      "epoch": 9.68179160332125,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7135,
      "step": 41390
    },
    {
      "epoch": 9.682025494094258,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 41391
    },
    {
      "epoch": 9.682259384867267,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6807,
      "step": 41392
    },
    {
      "epoch": 9.682493275640276,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6333,
      "step": 41393
    },
    {
      "epoch": 9.682727166413285,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 41394
    },
    {
      "epoch": 9.682961057186294,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6278,
      "step": 41395
    },
    {
      "epoch": 9.683194947959302,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 41396
    },
    {
      "epoch": 9.683428838732311,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 41397
    },
    {
      "epoch": 9.683662729505322,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 41398
    },
    {
      "epoch": 9.68389662027833,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7021,
      "step": 41399
    },
    {
      "epoch": 9.68413051105134,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 41400
    },
    {
      "epoch": 9.68413051105134,
      "eval_runtime": 4.6382,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 41400
    },
    {
      "epoch": 9.684364401824348,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6632,
      "step": 41401
    },
    {
      "epoch": 9.684598292597357,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7219,
      "step": 41402
    },
    {
      "epoch": 9.684832183370366,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 41403
    },
    {
      "epoch": 9.685066074143375,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.4795,
      "step": 41404
    },
    {
      "epoch": 9.685299964916384,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.691,
      "step": 41405
    },
    {
      "epoch": 9.685533855689393,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5923,
      "step": 41406
    },
    {
      "epoch": 9.685767746462401,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 41407
    },
    {
      "epoch": 9.68600163723541,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6653,
      "step": 41408
    },
    {
      "epoch": 9.686235528008421,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6329,
      "step": 41409
    },
    {
      "epoch": 9.68646941878143,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6292,
      "step": 41410
    },
    {
      "epoch": 9.686703309554439,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.606,
      "step": 41411
    },
    {
      "epoch": 9.686937200327447,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6638,
      "step": 41412
    },
    {
      "epoch": 9.687171091100456,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.3902,
      "step": 41413
    },
    {
      "epoch": 9.687404981873465,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 41414
    },
    {
      "epoch": 9.687638872646474,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5031,
      "step": 41415
    },
    {
      "epoch": 9.687872763419483,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.6539,
      "step": 41416
    },
    {
      "epoch": 9.688106654192492,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6582,
      "step": 41417
    },
    {
      "epoch": 9.6883405449655,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 41418
    },
    {
      "epoch": 9.68857443573851,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 41419
    },
    {
      "epoch": 9.68880832651152,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 41420
    },
    {
      "epoch": 9.689042217284529,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 41421
    },
    {
      "epoch": 9.689276108057538,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 41422
    },
    {
      "epoch": 9.689509998830546,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 41423
    },
    {
      "epoch": 9.689743889603555,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 41424
    },
    {
      "epoch": 9.689977780376564,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 41425
    },
    {
      "epoch": 9.690211671149573,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6163,
      "step": 41426
    },
    {
      "epoch": 9.690445561922582,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6639,
      "step": 41427
    },
    {
      "epoch": 9.69067945269559,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8038,
      "step": 41428
    },
    {
      "epoch": 9.6909133434686,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8372,
      "step": 41429
    },
    {
      "epoch": 9.691147234241608,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 41430
    },
    {
      "epoch": 9.691381125014619,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6309,
      "step": 41431
    },
    {
      "epoch": 9.691615015787628,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6952,
      "step": 41432
    },
    {
      "epoch": 9.691848906560637,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 41433
    },
    {
      "epoch": 9.692082797333645,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 41434
    },
    {
      "epoch": 9.692316688106654,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 41435
    },
    {
      "epoch": 9.692550578879663,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 41436
    },
    {
      "epoch": 9.692784469652672,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 41437
    },
    {
      "epoch": 9.69301836042568,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.574,
      "step": 41438
    },
    {
      "epoch": 9.69325225119869,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 41439
    },
    {
      "epoch": 9.693486141971698,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2988,
      "step": 41440
    },
    {
      "epoch": 9.693720032744707,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 41441
    },
    {
      "epoch": 9.693953923517718,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7571,
      "step": 41442
    },
    {
      "epoch": 9.694187814290727,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6452,
      "step": 41443
    },
    {
      "epoch": 9.694421705063736,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6882,
      "step": 41444
    },
    {
      "epoch": 9.694655595836744,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 41445
    },
    {
      "epoch": 9.694889486609753,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 41446
    },
    {
      "epoch": 9.695123377382762,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4195,
      "step": 41447
    },
    {
      "epoch": 9.695357268155771,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 41448
    },
    {
      "epoch": 9.69559115892878,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 41449
    },
    {
      "epoch": 9.695825049701789,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5505,
      "step": 41450
    },
    {
      "epoch": 9.696058940474797,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9061,
      "step": 41451
    },
    {
      "epoch": 9.696292831247808,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5428,
      "step": 41452
    },
    {
      "epoch": 9.696526722020817,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 41453
    },
    {
      "epoch": 9.696760612793826,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 41454
    },
    {
      "epoch": 9.696994503566835,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5432,
      "step": 41455
    },
    {
      "epoch": 9.697228394339843,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 41456
    },
    {
      "epoch": 9.697462285112852,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5277,
      "step": 41457
    },
    {
      "epoch": 9.697696175885861,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 41458
    },
    {
      "epoch": 9.69793006665887,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5571,
      "step": 41459
    },
    {
      "epoch": 9.698163957431879,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 41460
    },
    {
      "epoch": 9.698397848204888,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.6168,
      "step": 41461
    },
    {
      "epoch": 9.698631738977898,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 41462
    },
    {
      "epoch": 9.698865629750907,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 41463
    },
    {
      "epoch": 9.699099520523916,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 41464
    },
    {
      "epoch": 9.699333411296925,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7296,
      "step": 41465
    },
    {
      "epoch": 9.699567302069934,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 41466
    },
    {
      "epoch": 9.699801192842942,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 41467
    },
    {
      "epoch": 9.700035083615951,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6099,
      "step": 41468
    },
    {
      "epoch": 9.70026897438896,
      "grad_norm": 2.546875,
      "learning_rate": 3e-05,
      "loss": 1.4665,
      "step": 41469
    },
    {
      "epoch": 9.700502865161969,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3949,
      "step": 41470
    },
    {
      "epoch": 9.700736755934978,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 41471
    },
    {
      "epoch": 9.700970646707987,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 41472
    },
    {
      "epoch": 9.701204537480997,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 41473
    },
    {
      "epoch": 9.701438428254006,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 41474
    },
    {
      "epoch": 9.701672319027015,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6539,
      "step": 41475
    },
    {
      "epoch": 9.701906209800024,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 41476
    },
    {
      "epoch": 9.702140100573033,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 41477
    },
    {
      "epoch": 9.702373991346041,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6591,
      "step": 41478
    },
    {
      "epoch": 9.70260788211905,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6031,
      "step": 41479
    },
    {
      "epoch": 9.702841772892059,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.4249,
      "step": 41480
    },
    {
      "epoch": 9.703075663665068,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4765,
      "step": 41481
    },
    {
      "epoch": 9.703309554438077,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 41482
    },
    {
      "epoch": 9.703543445211086,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 41483
    },
    {
      "epoch": 9.703777335984096,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.5873,
      "step": 41484
    },
    {
      "epoch": 9.704011226757105,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 41485
    },
    {
      "epoch": 9.704245117530114,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4966,
      "step": 41486
    },
    {
      "epoch": 9.704479008303123,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 41487
    },
    {
      "epoch": 9.704712899076132,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.761,
      "step": 41488
    },
    {
      "epoch": 9.70494678984914,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 41489
    },
    {
      "epoch": 9.70518068062215,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4142,
      "step": 41490
    },
    {
      "epoch": 9.705414571395158,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4545,
      "step": 41491
    },
    {
      "epoch": 9.705648462168167,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.572,
      "step": 41492
    },
    {
      "epoch": 9.705882352941176,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6533,
      "step": 41493
    },
    {
      "epoch": 9.706116243714185,
      "grad_norm": 11.375,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 41494
    },
    {
      "epoch": 9.706350134487195,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5122,
      "step": 41495
    },
    {
      "epoch": 9.706584025260204,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6325,
      "step": 41496
    },
    {
      "epoch": 9.706817916033213,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6307,
      "step": 41497
    },
    {
      "epoch": 9.707051806806222,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 41498
    },
    {
      "epoch": 9.70728569757923,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 41499
    },
    {
      "epoch": 9.70751958835224,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 41500
    },
    {
      "epoch": 9.70751958835224,
      "eval_runtime": 4.6345,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 41500
    },
    {
      "epoch": 9.707753479125248,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 41501
    },
    {
      "epoch": 9.707987369898257,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 41502
    },
    {
      "epoch": 9.708221260671266,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 41503
    },
    {
      "epoch": 9.708455151444275,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 41504
    },
    {
      "epoch": 9.708689042217285,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 41505
    },
    {
      "epoch": 9.708922932990294,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 41506
    },
    {
      "epoch": 9.709156823763303,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7237,
      "step": 41507
    },
    {
      "epoch": 9.709390714536312,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 41508
    },
    {
      "epoch": 9.70962460530932,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6841,
      "step": 41509
    },
    {
      "epoch": 9.70985849608233,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 41510
    },
    {
      "epoch": 9.710092386855338,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 41511
    },
    {
      "epoch": 9.710326277628347,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6384,
      "step": 41512
    },
    {
      "epoch": 9.710560168401356,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 41513
    },
    {
      "epoch": 9.710794059174365,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4036,
      "step": 41514
    },
    {
      "epoch": 9.711027949947376,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 41515
    },
    {
      "epoch": 9.711261840720384,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 41516
    },
    {
      "epoch": 9.711495731493393,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 41517
    },
    {
      "epoch": 9.711729622266402,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 41518
    },
    {
      "epoch": 9.71196351303941,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6479,
      "step": 41519
    },
    {
      "epoch": 9.71219740381242,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6731,
      "step": 41520
    },
    {
      "epoch": 9.712431294585429,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 41521
    },
    {
      "epoch": 9.712665185358437,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 41522
    },
    {
      "epoch": 9.712899076131446,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 41523
    },
    {
      "epoch": 9.713132966904455,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 41524
    },
    {
      "epoch": 9.713366857677464,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 41525
    },
    {
      "epoch": 9.713600748450474,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.629,
      "step": 41526
    },
    {
      "epoch": 9.713834639223483,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.532,
      "step": 41527
    },
    {
      "epoch": 9.714068529996492,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7089,
      "step": 41528
    },
    {
      "epoch": 9.714302420769501,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 41529
    },
    {
      "epoch": 9.71453631154251,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 41530
    },
    {
      "epoch": 9.714770202315519,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 41531
    },
    {
      "epoch": 9.715004093088528,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 41532
    },
    {
      "epoch": 9.715237983861536,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 41533
    },
    {
      "epoch": 9.715471874634545,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4759,
      "step": 41534
    },
    {
      "epoch": 9.715705765407554,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 41535
    },
    {
      "epoch": 9.715939656180563,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 41536
    },
    {
      "epoch": 9.716173546953573,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6866,
      "step": 41537
    },
    {
      "epoch": 9.716407437726582,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5581,
      "step": 41538
    },
    {
      "epoch": 9.716641328499591,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 41539
    },
    {
      "epoch": 9.7168752192726,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6276,
      "step": 41540
    },
    {
      "epoch": 9.717109110045609,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7941,
      "step": 41541
    },
    {
      "epoch": 9.717343000818618,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5157,
      "step": 41542
    },
    {
      "epoch": 9.717576891591627,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 41543
    },
    {
      "epoch": 9.717810782364635,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 41544
    },
    {
      "epoch": 9.718044673137644,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2047,
      "step": 41545
    },
    {
      "epoch": 9.718278563910653,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 41546
    },
    {
      "epoch": 9.718512454683662,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6282,
      "step": 41547
    },
    {
      "epoch": 9.718746345456672,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.819,
      "step": 41548
    },
    {
      "epoch": 9.718980236229681,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 41549
    },
    {
      "epoch": 9.71921412700269,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.6019,
      "step": 41550
    },
    {
      "epoch": 9.719448017775699,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 41551
    },
    {
      "epoch": 9.719681908548708,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 41552
    },
    {
      "epoch": 9.719915799321717,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7491,
      "step": 41553
    },
    {
      "epoch": 9.720149690094726,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5393,
      "step": 41554
    },
    {
      "epoch": 9.720383580867734,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 41555
    },
    {
      "epoch": 9.720617471640743,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 41556
    },
    {
      "epoch": 9.720851362413752,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 41557
    },
    {
      "epoch": 9.72108525318676,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 41558
    },
    {
      "epoch": 9.721319143959771,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 41559
    },
    {
      "epoch": 9.72155303473278,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 41560
    },
    {
      "epoch": 9.72178692550579,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6813,
      "step": 41561
    },
    {
      "epoch": 9.722020816278798,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6986,
      "step": 41562
    },
    {
      "epoch": 9.722254707051807,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.812,
      "step": 41563
    },
    {
      "epoch": 9.722488597824816,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6749,
      "step": 41564
    },
    {
      "epoch": 9.722722488597825,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 41565
    },
    {
      "epoch": 9.722956379370833,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6876,
      "step": 41566
    },
    {
      "epoch": 9.723190270143842,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4975,
      "step": 41567
    },
    {
      "epoch": 9.723424160916851,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.58,
      "step": 41568
    },
    {
      "epoch": 9.723658051689862,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 41569
    },
    {
      "epoch": 9.72389194246287,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6923,
      "step": 41570
    },
    {
      "epoch": 9.72412583323588,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 41571
    },
    {
      "epoch": 9.724359724008888,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6278,
      "step": 41572
    },
    {
      "epoch": 9.724593614781897,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.3927,
      "step": 41573
    },
    {
      "epoch": 9.724827505554906,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 41574
    },
    {
      "epoch": 9.725061396327915,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 41575
    },
    {
      "epoch": 9.725295287100924,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 41576
    },
    {
      "epoch": 9.725529177873932,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5769,
      "step": 41577
    },
    {
      "epoch": 9.725763068646941,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.404,
      "step": 41578
    },
    {
      "epoch": 9.725996959419952,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7203,
      "step": 41579
    },
    {
      "epoch": 9.72623085019296,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4698,
      "step": 41580
    },
    {
      "epoch": 9.72646474096597,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6729,
      "step": 41581
    },
    {
      "epoch": 9.726698631738978,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.4993,
      "step": 41582
    },
    {
      "epoch": 9.726932522511987,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 41583
    },
    {
      "epoch": 9.727166413284996,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 41584
    },
    {
      "epoch": 9.727400304058005,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5538,
      "step": 41585
    },
    {
      "epoch": 9.727634194831014,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 41586
    },
    {
      "epoch": 9.727868085604022,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6211,
      "step": 41587
    },
    {
      "epoch": 9.728101976377031,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 41588
    },
    {
      "epoch": 9.72833586715004,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 41589
    },
    {
      "epoch": 9.72856975792305,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 41590
    },
    {
      "epoch": 9.72880364869606,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7541,
      "step": 41591
    },
    {
      "epoch": 9.729037539469068,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 41592
    },
    {
      "epoch": 9.729271430242077,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 41593
    },
    {
      "epoch": 9.729505321015086,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 41594
    },
    {
      "epoch": 9.729739211788095,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8332,
      "step": 41595
    },
    {
      "epoch": 9.729973102561104,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.8006,
      "step": 41596
    },
    {
      "epoch": 9.730206993334113,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8324,
      "step": 41597
    },
    {
      "epoch": 9.730440884107121,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.4261,
      "step": 41598
    },
    {
      "epoch": 9.73067477488013,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8988,
      "step": 41599
    },
    {
      "epoch": 9.73090866565314,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.5741,
      "step": 41600
    },
    {
      "epoch": 9.73090866565314,
      "eval_runtime": 4.6204,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 41600
    },
    {
      "epoch": 9.73114255642615,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 41601
    },
    {
      "epoch": 9.731376447199159,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.4325,
      "step": 41602
    },
    {
      "epoch": 9.731610337972167,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 41603
    },
    {
      "epoch": 9.731844228745176,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.691,
      "step": 41604
    },
    {
      "epoch": 9.732078119518185,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 41605
    },
    {
      "epoch": 9.732312010291194,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6566,
      "step": 41606
    },
    {
      "epoch": 9.732545901064203,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7047,
      "step": 41607
    },
    {
      "epoch": 9.732779791837212,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 41608
    },
    {
      "epoch": 9.73301368261022,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 41609
    },
    {
      "epoch": 9.73324757338323,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6127,
      "step": 41610
    },
    {
      "epoch": 9.733481464156238,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6443,
      "step": 41611
    },
    {
      "epoch": 9.733715354929249,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5439,
      "step": 41612
    },
    {
      "epoch": 9.733949245702258,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6576,
      "step": 41613
    },
    {
      "epoch": 9.734183136475266,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4178,
      "step": 41614
    },
    {
      "epoch": 9.734417027248275,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6772,
      "step": 41615
    },
    {
      "epoch": 9.734650918021284,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8239,
      "step": 41616
    },
    {
      "epoch": 9.734884808794293,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 41617
    },
    {
      "epoch": 9.735118699567302,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.5416,
      "step": 41618
    },
    {
      "epoch": 9.73535259034031,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 41619
    },
    {
      "epoch": 9.73558648111332,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 41620
    },
    {
      "epoch": 9.735820371886328,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 41621
    },
    {
      "epoch": 9.736054262659337,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 41622
    },
    {
      "epoch": 9.736288153432348,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6762,
      "step": 41623
    },
    {
      "epoch": 9.736522044205357,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 41624
    },
    {
      "epoch": 9.736755934978365,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 41625
    },
    {
      "epoch": 9.736989825751374,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 41626
    },
    {
      "epoch": 9.737223716524383,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 41627
    },
    {
      "epoch": 9.737457607297392,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4617,
      "step": 41628
    },
    {
      "epoch": 9.7376914980704,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7572,
      "step": 41629
    },
    {
      "epoch": 9.73792538884341,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 41630
    },
    {
      "epoch": 9.738159279616418,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 41631
    },
    {
      "epoch": 9.738393170389427,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.3042,
      "step": 41632
    },
    {
      "epoch": 9.738627061162438,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 41633
    },
    {
      "epoch": 9.738860951935447,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.635,
      "step": 41634
    },
    {
      "epoch": 9.739094842708456,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7186,
      "step": 41635
    },
    {
      "epoch": 9.739328733481464,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.2499,
      "step": 41636
    },
    {
      "epoch": 9.739562624254473,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 41637
    },
    {
      "epoch": 9.739796515027482,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 41638
    },
    {
      "epoch": 9.740030405800491,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.5995,
      "step": 41639
    },
    {
      "epoch": 9.7402642965735,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.6932,
      "step": 41640
    },
    {
      "epoch": 9.740498187346509,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7822,
      "step": 41641
    },
    {
      "epoch": 9.740732078119517,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.58,
      "step": 41642
    },
    {
      "epoch": 9.740965968892528,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 41643
    },
    {
      "epoch": 9.741199859665537,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6575,
      "step": 41644
    },
    {
      "epoch": 9.741433750438546,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.4807,
      "step": 41645
    },
    {
      "epoch": 9.741667641211555,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 41646
    },
    {
      "epoch": 9.741901531984563,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7191,
      "step": 41647
    },
    {
      "epoch": 9.742135422757572,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7241,
      "step": 41648
    },
    {
      "epoch": 9.742369313530581,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.67,
      "step": 41649
    },
    {
      "epoch": 9.74260320430359,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 41650
    },
    {
      "epoch": 9.742837095076599,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 41651
    },
    {
      "epoch": 9.743070985849608,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 41652
    },
    {
      "epoch": 9.743304876622616,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 41653
    },
    {
      "epoch": 9.743538767395627,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6559,
      "step": 41654
    },
    {
      "epoch": 9.743772658168636,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7365,
      "step": 41655
    },
    {
      "epoch": 9.744006548941645,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.3789,
      "step": 41656
    },
    {
      "epoch": 9.744240439714654,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5775,
      "step": 41657
    },
    {
      "epoch": 9.744474330487662,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.4548,
      "step": 41658
    },
    {
      "epoch": 9.744708221260671,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6602,
      "step": 41659
    },
    {
      "epoch": 9.74494211203368,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 41660
    },
    {
      "epoch": 9.745176002806689,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 41661
    },
    {
      "epoch": 9.745409893579698,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 41662
    },
    {
      "epoch": 9.745643784352707,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4506,
      "step": 41663
    },
    {
      "epoch": 9.745877675125715,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 41664
    },
    {
      "epoch": 9.746111565898726,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.54,
      "step": 41665
    },
    {
      "epoch": 9.746345456671735,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6087,
      "step": 41666
    },
    {
      "epoch": 9.746579347444744,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7403,
      "step": 41667
    },
    {
      "epoch": 9.746813238217753,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 41668
    },
    {
      "epoch": 9.747047128990761,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7552,
      "step": 41669
    },
    {
      "epoch": 9.74728101976377,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5977,
      "step": 41670
    },
    {
      "epoch": 9.747514910536779,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5744,
      "step": 41671
    },
    {
      "epoch": 9.747748801309788,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 41672
    },
    {
      "epoch": 9.747982692082797,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 41673
    },
    {
      "epoch": 9.748216582855806,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4921,
      "step": 41674
    },
    {
      "epoch": 9.748450473628814,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 41675
    },
    {
      "epoch": 9.748684364401825,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5776,
      "step": 41676
    },
    {
      "epoch": 9.748918255174834,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5636,
      "step": 41677
    },
    {
      "epoch": 9.749152145947843,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 41678
    },
    {
      "epoch": 9.749386036720852,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 41679
    },
    {
      "epoch": 9.74961992749386,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 41680
    },
    {
      "epoch": 9.74985381826687,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 41681
    },
    {
      "epoch": 9.750087709039878,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 41682
    },
    {
      "epoch": 9.750321599812887,
      "grad_norm": 21.75,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 41683
    },
    {
      "epoch": 9.750555490585896,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 41684
    },
    {
      "epoch": 9.750789381358905,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5032,
      "step": 41685
    },
    {
      "epoch": 9.751023272131915,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 41686
    },
    {
      "epoch": 9.751257162904924,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.5533,
      "step": 41687
    },
    {
      "epoch": 9.751491053677933,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.6297,
      "step": 41688
    },
    {
      "epoch": 9.751724944450942,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 41689
    },
    {
      "epoch": 9.75195883522395,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 41690
    },
    {
      "epoch": 9.75219272599696,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 41691
    },
    {
      "epoch": 9.752426616769968,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7463,
      "step": 41692
    },
    {
      "epoch": 9.752660507542977,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.6209,
      "step": 41693
    },
    {
      "epoch": 9.752894398315986,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 41694
    },
    {
      "epoch": 9.753128289088995,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 41695
    },
    {
      "epoch": 9.753362179862005,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6843,
      "step": 41696
    },
    {
      "epoch": 9.753596070635014,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7288,
      "step": 41697
    },
    {
      "epoch": 9.753829961408023,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 41698
    },
    {
      "epoch": 9.754063852181032,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 41699
    },
    {
      "epoch": 9.75429774295404,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6479,
      "step": 41700
    },
    {
      "epoch": 9.75429774295404,
      "eval_runtime": 4.6024,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 41700
    },
    {
      "epoch": 9.75453163372705,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8199,
      "step": 41701
    },
    {
      "epoch": 9.754765524500058,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 41702
    },
    {
      "epoch": 9.754999415273067,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 41703
    },
    {
      "epoch": 9.755233306046076,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.5613,
      "step": 41704
    },
    {
      "epoch": 9.755467196819085,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.8306,
      "step": 41705
    },
    {
      "epoch": 9.755701087592094,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 41706
    },
    {
      "epoch": 9.755934978365104,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5665,
      "step": 41707
    },
    {
      "epoch": 9.756168869138113,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 41708
    },
    {
      "epoch": 9.756402759911122,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6861,
      "step": 41709
    },
    {
      "epoch": 9.75663665068413,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7245,
      "step": 41710
    },
    {
      "epoch": 9.75687054145714,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4091,
      "step": 41711
    },
    {
      "epoch": 9.757104432230149,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6726,
      "step": 41712
    },
    {
      "epoch": 9.757338323003157,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5541,
      "step": 41713
    },
    {
      "epoch": 9.757572213776166,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.744,
      "step": 41714
    },
    {
      "epoch": 9.757806104549175,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 41715
    },
    {
      "epoch": 9.758039995322184,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 41716
    },
    {
      "epoch": 9.758273886095193,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 41717
    },
    {
      "epoch": 9.758507776868203,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7576,
      "step": 41718
    },
    {
      "epoch": 9.758741667641212,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 41719
    },
    {
      "epoch": 9.758975558414221,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 41720
    },
    {
      "epoch": 9.75920944918723,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6567,
      "step": 41721
    },
    {
      "epoch": 9.759443339960239,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 41722
    },
    {
      "epoch": 9.759677230733248,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.3997,
      "step": 41723
    },
    {
      "epoch": 9.759911121506256,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6754,
      "step": 41724
    },
    {
      "epoch": 9.760145012279265,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7018,
      "step": 41725
    },
    {
      "epoch": 9.760378903052274,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5709,
      "step": 41726
    },
    {
      "epoch": 9.760612793825283,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.5828,
      "step": 41727
    },
    {
      "epoch": 9.760846684598292,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6738,
      "step": 41728
    },
    {
      "epoch": 9.761080575371302,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6401,
      "step": 41729
    },
    {
      "epoch": 9.761314466144311,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 41730
    },
    {
      "epoch": 9.76154835691732,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6488,
      "step": 41731
    },
    {
      "epoch": 9.761782247690329,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 41732
    },
    {
      "epoch": 9.762016138463338,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.5677,
      "step": 41733
    },
    {
      "epoch": 9.762250029236347,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 41734
    },
    {
      "epoch": 9.762483920009355,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.59,
      "step": 41735
    },
    {
      "epoch": 9.762717810782364,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 41736
    },
    {
      "epoch": 9.762951701555373,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 41737
    },
    {
      "epoch": 9.763185592328382,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 41738
    },
    {
      "epoch": 9.76341948310139,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 41739
    },
    {
      "epoch": 9.763653373874401,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7144,
      "step": 41740
    },
    {
      "epoch": 9.76388726464741,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5602,
      "step": 41741
    },
    {
      "epoch": 9.764121155420419,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 41742
    },
    {
      "epoch": 9.764355046193428,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.3766,
      "step": 41743
    },
    {
      "epoch": 9.764588936966437,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 41744
    },
    {
      "epoch": 9.764822827739446,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 41745
    },
    {
      "epoch": 9.765056718512454,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6753,
      "step": 41746
    },
    {
      "epoch": 9.765290609285463,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5516,
      "step": 41747
    },
    {
      "epoch": 9.765524500058472,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 41748
    },
    {
      "epoch": 9.765758390831481,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 41749
    },
    {
      "epoch": 9.765992281604492,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 41750
    },
    {
      "epoch": 9.7662261723775,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9811,
      "step": 41751
    },
    {
      "epoch": 9.76646006315051,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 41752
    },
    {
      "epoch": 9.766693953923518,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.5292,
      "step": 41753
    },
    {
      "epoch": 9.766927844696527,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 41754
    },
    {
      "epoch": 9.767161735469536,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.6411,
      "step": 41755
    },
    {
      "epoch": 9.767395626242545,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 41756
    },
    {
      "epoch": 9.767629517015553,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 41757
    },
    {
      "epoch": 9.767863407788562,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 41758
    },
    {
      "epoch": 9.768097298561571,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.7079,
      "step": 41759
    },
    {
      "epoch": 9.768331189334582,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.612,
      "step": 41760
    },
    {
      "epoch": 9.76856508010759,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 41761
    },
    {
      "epoch": 9.7687989708806,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 41762
    },
    {
      "epoch": 9.769032861653608,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6733,
      "step": 41763
    },
    {
      "epoch": 9.769266752426617,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 41764
    },
    {
      "epoch": 9.769500643199626,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6682,
      "step": 41765
    },
    {
      "epoch": 9.769734533972635,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 41766
    },
    {
      "epoch": 9.769968424745644,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7135,
      "step": 41767
    },
    {
      "epoch": 9.770202315518652,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6031,
      "step": 41768
    },
    {
      "epoch": 9.770436206291661,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 41769
    },
    {
      "epoch": 9.77067009706467,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6667,
      "step": 41770
    },
    {
      "epoch": 9.77090398783768,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.5988,
      "step": 41771
    },
    {
      "epoch": 9.77113787861069,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7284,
      "step": 41772
    },
    {
      "epoch": 9.771371769383698,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5998,
      "step": 41773
    },
    {
      "epoch": 9.771605660156707,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7994,
      "step": 41774
    },
    {
      "epoch": 9.771839550929716,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8107,
      "step": 41775
    },
    {
      "epoch": 9.772073441702725,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 41776
    },
    {
      "epoch": 9.772307332475734,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5968,
      "step": 41777
    },
    {
      "epoch": 9.772541223248743,
      "grad_norm": 7.46875,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 41778
    },
    {
      "epoch": 9.772775114021751,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 41779
    },
    {
      "epoch": 9.77300900479476,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7643,
      "step": 41780
    },
    {
      "epoch": 9.773242895567769,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 41781
    },
    {
      "epoch": 9.77347678634078,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 41782
    },
    {
      "epoch": 9.773710677113788,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 41783
    },
    {
      "epoch": 9.773944567886797,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.4778,
      "step": 41784
    },
    {
      "epoch": 9.774178458659806,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5155,
      "step": 41785
    },
    {
      "epoch": 9.774412349432815,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7306,
      "step": 41786
    },
    {
      "epoch": 9.774646240205824,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.823,
      "step": 41787
    },
    {
      "epoch": 9.774880130978833,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.4415,
      "step": 41788
    },
    {
      "epoch": 9.775114021751842,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6977,
      "step": 41789
    },
    {
      "epoch": 9.77534791252485,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7177,
      "step": 41790
    },
    {
      "epoch": 9.77558180329786,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 41791
    },
    {
      "epoch": 9.775815694070868,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6465,
      "step": 41792
    },
    {
      "epoch": 9.776049584843879,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 41793
    },
    {
      "epoch": 9.776283475616887,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7863,
      "step": 41794
    },
    {
      "epoch": 9.776517366389896,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 41795
    },
    {
      "epoch": 9.776751257162905,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 41796
    },
    {
      "epoch": 9.776985147935914,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.548,
      "step": 41797
    },
    {
      "epoch": 9.777219038708923,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 41798
    },
    {
      "epoch": 9.777452929481932,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7364,
      "step": 41799
    },
    {
      "epoch": 9.77768682025494,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8657,
      "step": 41800
    },
    {
      "epoch": 9.77768682025494,
      "eval_runtime": 4.6033,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 41800
    },
    {
      "epoch": 9.77792071102795,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 41801
    },
    {
      "epoch": 9.778154601800958,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 41802
    },
    {
      "epoch": 9.778388492573967,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 41803
    },
    {
      "epoch": 9.778622383346978,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.4621,
      "step": 41804
    },
    {
      "epoch": 9.778856274119986,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 41805
    },
    {
      "epoch": 9.779090164892995,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6737,
      "step": 41806
    },
    {
      "epoch": 9.779324055666004,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 41807
    },
    {
      "epoch": 9.779557946439013,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7088,
      "step": 41808
    },
    {
      "epoch": 9.779791837212022,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6617,
      "step": 41809
    },
    {
      "epoch": 9.78002572798503,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7112,
      "step": 41810
    },
    {
      "epoch": 9.78025961875804,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6927,
      "step": 41811
    },
    {
      "epoch": 9.780493509531048,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.61,
      "step": 41812
    },
    {
      "epoch": 9.780727400304057,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.8141,
      "step": 41813
    },
    {
      "epoch": 9.780961291077068,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7271,
      "step": 41814
    },
    {
      "epoch": 9.781195181850077,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 41815
    },
    {
      "epoch": 9.781429072623085,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 41816
    },
    {
      "epoch": 9.781662963396094,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 41817
    },
    {
      "epoch": 9.781896854169103,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 1.5369,
      "step": 41818
    },
    {
      "epoch": 9.782130744942112,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7734,
      "step": 41819
    },
    {
      "epoch": 9.78236463571512,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 41820
    },
    {
      "epoch": 9.78259852648813,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6149,
      "step": 41821
    },
    {
      "epoch": 9.782832417261139,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7752,
      "step": 41822
    },
    {
      "epoch": 9.783066308034147,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6195,
      "step": 41823
    },
    {
      "epoch": 9.783300198807158,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6016,
      "step": 41824
    },
    {
      "epoch": 9.783534089580167,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7203,
      "step": 41825
    },
    {
      "epoch": 9.783767980353176,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 41826
    },
    {
      "epoch": 9.784001871126184,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 41827
    },
    {
      "epoch": 9.784235761899193,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 41828
    },
    {
      "epoch": 9.784469652672202,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 41829
    },
    {
      "epoch": 9.784703543445211,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 41830
    },
    {
      "epoch": 9.78493743421822,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 41831
    },
    {
      "epoch": 9.785171324991229,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.5282,
      "step": 41832
    },
    {
      "epoch": 9.785405215764237,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 41833
    },
    {
      "epoch": 9.785639106537246,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7137,
      "step": 41834
    },
    {
      "epoch": 9.785872997310257,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.4973,
      "step": 41835
    },
    {
      "epoch": 9.786106888083266,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 41836
    },
    {
      "epoch": 9.786340778856275,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6943,
      "step": 41837
    },
    {
      "epoch": 9.786574669629283,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8308,
      "step": 41838
    },
    {
      "epoch": 9.786808560402292,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7579,
      "step": 41839
    },
    {
      "epoch": 9.787042451175301,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6578,
      "step": 41840
    },
    {
      "epoch": 9.78727634194831,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 41841
    },
    {
      "epoch": 9.787510232721319,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7546,
      "step": 41842
    },
    {
      "epoch": 9.787744123494328,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 41843
    },
    {
      "epoch": 9.787978014267336,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6921,
      "step": 41844
    },
    {
      "epoch": 9.788211905040345,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5971,
      "step": 41845
    },
    {
      "epoch": 9.788445795813356,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 41846
    },
    {
      "epoch": 9.788679686586365,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 41847
    },
    {
      "epoch": 9.788913577359374,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7375,
      "step": 41848
    },
    {
      "epoch": 9.789147468132382,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 41849
    },
    {
      "epoch": 9.789381358905391,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.721,
      "step": 41850
    },
    {
      "epoch": 9.7896152496784,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 41851
    },
    {
      "epoch": 9.789849140451409,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7124,
      "step": 41852
    },
    {
      "epoch": 9.790083031224418,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 41853
    },
    {
      "epoch": 9.790316921997427,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 41854
    },
    {
      "epoch": 9.790550812770435,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.4802,
      "step": 41855
    },
    {
      "epoch": 9.790784703543444,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7722,
      "step": 41856
    },
    {
      "epoch": 9.791018594316455,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 41857
    },
    {
      "epoch": 9.791252485089464,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7714,
      "step": 41858
    },
    {
      "epoch": 9.791486375862473,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 41859
    },
    {
      "epoch": 9.791720266635481,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 41860
    },
    {
      "epoch": 9.79195415740849,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 41861
    },
    {
      "epoch": 9.7921880481815,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.714,
      "step": 41862
    },
    {
      "epoch": 9.792421938954508,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 41863
    },
    {
      "epoch": 9.792655829727517,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 41864
    },
    {
      "epoch": 9.792889720500526,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6405,
      "step": 41865
    },
    {
      "epoch": 9.793123611273534,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 41866
    },
    {
      "epoch": 9.793357502046545,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 41867
    },
    {
      "epoch": 9.793591392819554,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 41868
    },
    {
      "epoch": 9.793825283592563,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.6449,
      "step": 41869
    },
    {
      "epoch": 9.794059174365572,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 1.7803,
      "step": 41870
    },
    {
      "epoch": 9.79429306513858,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.4764,
      "step": 41871
    },
    {
      "epoch": 9.79452695591159,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5748,
      "step": 41872
    },
    {
      "epoch": 9.794760846684598,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.713,
      "step": 41873
    },
    {
      "epoch": 9.794994737457607,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7595,
      "step": 41874
    },
    {
      "epoch": 9.795228628230616,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7404,
      "step": 41875
    },
    {
      "epoch": 9.795462519003625,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 41876
    },
    {
      "epoch": 9.795696409776635,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 41877
    },
    {
      "epoch": 9.795930300549644,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 41878
    },
    {
      "epoch": 9.796164191322653,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 41879
    },
    {
      "epoch": 9.796398082095662,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 41880
    },
    {
      "epoch": 9.79663197286867,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6225,
      "step": 41881
    },
    {
      "epoch": 9.79686586364168,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7883,
      "step": 41882
    },
    {
      "epoch": 9.797099754414688,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6448,
      "step": 41883
    },
    {
      "epoch": 9.797333645187697,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6806,
      "step": 41884
    },
    {
      "epoch": 9.797567535960706,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 41885
    },
    {
      "epoch": 9.797801426733715,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8034,
      "step": 41886
    },
    {
      "epoch": 9.798035317506724,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 41887
    },
    {
      "epoch": 9.798269208279734,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 41888
    },
    {
      "epoch": 9.798503099052743,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7448,
      "step": 41889
    },
    {
      "epoch": 9.798736989825752,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7377,
      "step": 41890
    },
    {
      "epoch": 9.79897088059876,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 41891
    },
    {
      "epoch": 9.79920477137177,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5689,
      "step": 41892
    },
    {
      "epoch": 9.799438662144778,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6357,
      "step": 41893
    },
    {
      "epoch": 9.799672552917787,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 41894
    },
    {
      "epoch": 9.799906443690796,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.613,
      "step": 41895
    },
    {
      "epoch": 9.800140334463805,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 41896
    },
    {
      "epoch": 9.800374225236814,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 41897
    },
    {
      "epoch": 9.800608116009823,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 41898
    },
    {
      "epoch": 9.800842006782833,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7088,
      "step": 41899
    },
    {
      "epoch": 9.801075897555842,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 41900
    },
    {
      "epoch": 9.801075897555842,
      "eval_runtime": 4.6395,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 41900
    },
    {
      "epoch": 9.801309788328851,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6268,
      "step": 41901
    },
    {
      "epoch": 9.80154367910186,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 41902
    },
    {
      "epoch": 9.801777569874869,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 41903
    },
    {
      "epoch": 9.802011460647877,
      "grad_norm": 7.3125,
      "learning_rate": 3e-05,
      "loss": 1.5783,
      "step": 41904
    },
    {
      "epoch": 9.802245351420886,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5973,
      "step": 41905
    },
    {
      "epoch": 9.802479242193895,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7838,
      "step": 41906
    },
    {
      "epoch": 9.802713132966904,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7106,
      "step": 41907
    },
    {
      "epoch": 9.802947023739913,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6427,
      "step": 41908
    },
    {
      "epoch": 9.803180914512922,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 41909
    },
    {
      "epoch": 9.803414805285932,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 41910
    },
    {
      "epoch": 9.803648696058941,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 41911
    },
    {
      "epoch": 9.80388258683195,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 41912
    },
    {
      "epoch": 9.804116477604959,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7907,
      "step": 41913
    },
    {
      "epoch": 9.804350368377968,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.493,
      "step": 41914
    },
    {
      "epoch": 9.804584259150976,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 41915
    },
    {
      "epoch": 9.804818149923985,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4869,
      "step": 41916
    },
    {
      "epoch": 9.805052040696994,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.6588,
      "step": 41917
    },
    {
      "epoch": 9.805285931470003,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 41918
    },
    {
      "epoch": 9.805519822243012,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4772,
      "step": 41919
    },
    {
      "epoch": 9.80575371301602,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 41920
    },
    {
      "epoch": 9.805987603789031,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 41921
    },
    {
      "epoch": 9.80622149456204,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7723,
      "step": 41922
    },
    {
      "epoch": 9.806455385335049,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 41923
    },
    {
      "epoch": 9.806689276108058,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7296,
      "step": 41924
    },
    {
      "epoch": 9.806923166881067,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.5833,
      "step": 41925
    },
    {
      "epoch": 9.807157057654075,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7937,
      "step": 41926
    },
    {
      "epoch": 9.807390948427084,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 41927
    },
    {
      "epoch": 9.807624839200093,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5315,
      "step": 41928
    },
    {
      "epoch": 9.807858729973102,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 41929
    },
    {
      "epoch": 9.80809262074611,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.553,
      "step": 41930
    },
    {
      "epoch": 9.808326511519121,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.6548,
      "step": 41931
    },
    {
      "epoch": 9.80856040229213,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 41932
    },
    {
      "epoch": 9.808794293065139,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.739,
      "step": 41933
    },
    {
      "epoch": 9.809028183838148,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.5974,
      "step": 41934
    },
    {
      "epoch": 9.809262074611157,
      "grad_norm": 11.8125,
      "learning_rate": 3e-05,
      "loss": 1.5458,
      "step": 41935
    },
    {
      "epoch": 9.809495965384166,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 41936
    },
    {
      "epoch": 9.809729856157174,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 41937
    },
    {
      "epoch": 9.809963746930183,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 41938
    },
    {
      "epoch": 9.810197637703192,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 41939
    },
    {
      "epoch": 9.810431528476201,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 41940
    },
    {
      "epoch": 9.810665419249212,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.659,
      "step": 41941
    },
    {
      "epoch": 9.81089931002222,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 41942
    },
    {
      "epoch": 9.81113320079523,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7359,
      "step": 41943
    },
    {
      "epoch": 9.811367091568238,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 41944
    },
    {
      "epoch": 9.811600982341247,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.4288,
      "step": 41945
    },
    {
      "epoch": 9.811834873114256,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 41946
    },
    {
      "epoch": 9.812068763887265,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.6174,
      "step": 41947
    },
    {
      "epoch": 9.812302654660273,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7478,
      "step": 41948
    },
    {
      "epoch": 9.812536545433282,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.5712,
      "step": 41949
    },
    {
      "epoch": 9.812770436206291,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 41950
    },
    {
      "epoch": 9.8130043269793,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6237,
      "step": 41951
    },
    {
      "epoch": 9.81323821775231,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.694,
      "step": 41952
    },
    {
      "epoch": 9.81347210852532,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.6795,
      "step": 41953
    },
    {
      "epoch": 9.813705999298328,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 41954
    },
    {
      "epoch": 9.813939890071337,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 41955
    },
    {
      "epoch": 9.814173780844346,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8669,
      "step": 41956
    },
    {
      "epoch": 9.814407671617355,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5378,
      "step": 41957
    },
    {
      "epoch": 9.814641562390364,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 41958
    },
    {
      "epoch": 9.814875453163372,
      "grad_norm": 11.4375,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 41959
    },
    {
      "epoch": 9.815109343936381,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 41960
    },
    {
      "epoch": 9.81534323470939,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7021,
      "step": 41961
    },
    {
      "epoch": 9.815577125482399,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 41962
    },
    {
      "epoch": 9.81581101625541,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 41963
    },
    {
      "epoch": 9.816044907028418,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 41964
    },
    {
      "epoch": 9.816278797801427,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 41965
    },
    {
      "epoch": 9.816512688574436,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 41966
    },
    {
      "epoch": 9.816746579347445,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 41967
    },
    {
      "epoch": 9.816980470120454,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6432,
      "step": 41968
    },
    {
      "epoch": 9.817214360893463,
      "grad_norm": 7.03125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 41969
    },
    {
      "epoch": 9.817448251666471,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7924,
      "step": 41970
    },
    {
      "epoch": 9.81768214243948,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6913,
      "step": 41971
    },
    {
      "epoch": 9.817916033212489,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7023,
      "step": 41972
    },
    {
      "epoch": 9.818149923985498,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 41973
    },
    {
      "epoch": 9.818383814758509,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4641,
      "step": 41974
    },
    {
      "epoch": 9.818617705531517,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 41975
    },
    {
      "epoch": 9.818851596304526,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.4781,
      "step": 41976
    },
    {
      "epoch": 9.819085487077535,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 41977
    },
    {
      "epoch": 9.819319377850544,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 41978
    },
    {
      "epoch": 9.819553268623553,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 41979
    },
    {
      "epoch": 9.819787159396562,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7633,
      "step": 41980
    },
    {
      "epoch": 9.82002105016957,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6793,
      "step": 41981
    },
    {
      "epoch": 9.82025494094258,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 41982
    },
    {
      "epoch": 9.820488831715588,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 41983
    },
    {
      "epoch": 9.820722722488597,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5549,
      "step": 41984
    },
    {
      "epoch": 9.820956613261608,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 41985
    },
    {
      "epoch": 9.821190504034616,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 41986
    },
    {
      "epoch": 9.821424394807625,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 41987
    },
    {
      "epoch": 9.821658285580634,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 41988
    },
    {
      "epoch": 9.821892176353643,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6072,
      "step": 41989
    },
    {
      "epoch": 9.822126067126652,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.5748,
      "step": 41990
    },
    {
      "epoch": 9.82235995789966,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 41991
    },
    {
      "epoch": 9.82259384867267,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 41992
    },
    {
      "epoch": 9.822827739445678,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 41993
    },
    {
      "epoch": 9.823061630218687,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5796,
      "step": 41994
    },
    {
      "epoch": 9.823295520991698,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6623,
      "step": 41995
    },
    {
      "epoch": 9.823529411764707,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 41996
    },
    {
      "epoch": 9.823763302537715,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6006,
      "step": 41997
    },
    {
      "epoch": 9.823997193310724,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7277,
      "step": 41998
    },
    {
      "epoch": 9.824231084083733,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 41999
    },
    {
      "epoch": 9.824464974856742,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 42000
    },
    {
      "epoch": 9.824464974856742,
      "eval_runtime": 4.6396,
      "eval_samples_per_second": 0.216,
      "eval_steps_per_second": 0.216,
      "step": 42000
    },
    {
      "epoch": 9.82469886562975,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6499,
      "step": 42001
    },
    {
      "epoch": 9.82493275640276,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.5608,
      "step": 42002
    },
    {
      "epoch": 9.825166647175768,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 42003
    },
    {
      "epoch": 9.825400537948777,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.5492,
      "step": 42004
    },
    {
      "epoch": 9.825634428721788,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4685,
      "step": 42005
    },
    {
      "epoch": 9.825868319494797,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7939,
      "step": 42006
    },
    {
      "epoch": 9.826102210267806,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 42007
    },
    {
      "epoch": 9.826336101040814,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 42008
    },
    {
      "epoch": 9.826569991813823,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 42009
    },
    {
      "epoch": 9.826803882586832,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 42010
    },
    {
      "epoch": 9.82703777335984,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.698,
      "step": 42011
    },
    {
      "epoch": 9.82727166413285,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 42012
    },
    {
      "epoch": 9.827505554905859,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6041,
      "step": 42013
    },
    {
      "epoch": 9.827739445678867,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 42014
    },
    {
      "epoch": 9.827973336451876,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 42015
    },
    {
      "epoch": 9.828207227224887,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 42016
    },
    {
      "epoch": 9.828441117997896,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 42017
    },
    {
      "epoch": 9.828675008770904,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7093,
      "step": 42018
    },
    {
      "epoch": 9.828908899543913,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.531,
      "step": 42019
    },
    {
      "epoch": 9.829142790316922,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7311,
      "step": 42020
    },
    {
      "epoch": 9.829376681089931,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7649,
      "step": 42021
    },
    {
      "epoch": 9.82961057186294,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.63,
      "step": 42022
    },
    {
      "epoch": 9.829844462635949,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 42023
    },
    {
      "epoch": 9.830078353408958,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 42024
    },
    {
      "epoch": 9.830312244181966,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.6839,
      "step": 42025
    },
    {
      "epoch": 9.830546134954975,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 42026
    },
    {
      "epoch": 9.830780025727986,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5473,
      "step": 42027
    },
    {
      "epoch": 9.831013916500995,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6696,
      "step": 42028
    },
    {
      "epoch": 9.831247807274003,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5194,
      "step": 42029
    },
    {
      "epoch": 9.831481698047012,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6859,
      "step": 42030
    },
    {
      "epoch": 9.831715588820021,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 42031
    },
    {
      "epoch": 9.83194947959303,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 42032
    },
    {
      "epoch": 9.832183370366039,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6399,
      "step": 42033
    },
    {
      "epoch": 9.832417261139048,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 42034
    },
    {
      "epoch": 9.832651151912057,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 42035
    },
    {
      "epoch": 9.832885042685065,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6812,
      "step": 42036
    },
    {
      "epoch": 9.833118933458074,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6806,
      "step": 42037
    },
    {
      "epoch": 9.833352824231085,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 42038
    },
    {
      "epoch": 9.833586715004094,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 42039
    },
    {
      "epoch": 9.833820605777102,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8223,
      "step": 42040
    },
    {
      "epoch": 9.834054496550111,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4944,
      "step": 42041
    },
    {
      "epoch": 9.83428838732312,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 42042
    },
    {
      "epoch": 9.834522278096129,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 42043
    },
    {
      "epoch": 9.834756168869138,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1168,
      "step": 42044
    },
    {
      "epoch": 9.834990059642147,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8124,
      "step": 42045
    },
    {
      "epoch": 9.835223950415156,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7195,
      "step": 42046
    },
    {
      "epoch": 9.835457841188164,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 42047
    },
    {
      "epoch": 9.835691731961173,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 42048
    },
    {
      "epoch": 9.835925622734184,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7167,
      "step": 42049
    },
    {
      "epoch": 9.836159513507193,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.7723,
      "step": 42050
    },
    {
      "epoch": 9.836393404280201,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 42051
    },
    {
      "epoch": 9.83662729505321,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 42052
    },
    {
      "epoch": 9.83686118582622,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5692,
      "step": 42053
    },
    {
      "epoch": 9.837095076599228,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 42054
    },
    {
      "epoch": 9.837328967372237,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 42055
    },
    {
      "epoch": 9.837562858145246,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 42056
    },
    {
      "epoch": 9.837796748918255,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 42057
    },
    {
      "epoch": 9.838030639691265,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 42058
    },
    {
      "epoch": 9.838264530464274,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5676,
      "step": 42059
    },
    {
      "epoch": 9.838498421237283,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.715,
      "step": 42060
    },
    {
      "epoch": 9.838732312010292,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 42061
    },
    {
      "epoch": 9.8389662027833,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.6921,
      "step": 42062
    },
    {
      "epoch": 9.83920009355631,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8365,
      "step": 42063
    },
    {
      "epoch": 9.839433984329318,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 42064
    },
    {
      "epoch": 9.839667875102327,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 42065
    },
    {
      "epoch": 9.839901765875336,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 42066
    },
    {
      "epoch": 9.840135656648345,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8434,
      "step": 42067
    },
    {
      "epoch": 9.840369547421354,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 42068
    },
    {
      "epoch": 9.840603438194364,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 42069
    },
    {
      "epoch": 9.840837328967373,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.3589,
      "step": 42070
    },
    {
      "epoch": 9.841071219740382,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 42071
    },
    {
      "epoch": 9.84130511051339,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7058,
      "step": 42072
    },
    {
      "epoch": 9.8415390012864,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6743,
      "step": 42073
    },
    {
      "epoch": 9.841772892059408,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6224,
      "step": 42074
    },
    {
      "epoch": 9.842006782832417,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 42075
    },
    {
      "epoch": 9.842240673605426,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 42076
    },
    {
      "epoch": 9.842474564378435,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.6283,
      "step": 42077
    },
    {
      "epoch": 9.842708455151444,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 42078
    },
    {
      "epoch": 9.842942345924452,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.4874,
      "step": 42079
    },
    {
      "epoch": 9.843176236697463,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6645,
      "step": 42080
    },
    {
      "epoch": 9.843410127470472,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7072,
      "step": 42081
    },
    {
      "epoch": 9.84364401824348,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.5473,
      "step": 42082
    },
    {
      "epoch": 9.84387790901649,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4211,
      "step": 42083
    },
    {
      "epoch": 9.844111799789498,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5267,
      "step": 42084
    },
    {
      "epoch": 9.844345690562507,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 42085
    },
    {
      "epoch": 9.844579581335516,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6365,
      "step": 42086
    },
    {
      "epoch": 9.844813472108525,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 42087
    },
    {
      "epoch": 9.845047362881534,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7357,
      "step": 42088
    },
    {
      "epoch": 9.845281253654543,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 42089
    },
    {
      "epoch": 9.845515144427551,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7747,
      "step": 42090
    },
    {
      "epoch": 9.845749035200562,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6393,
      "step": 42091
    },
    {
      "epoch": 9.845982925973571,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 42092
    },
    {
      "epoch": 9.84621681674658,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4103,
      "step": 42093
    },
    {
      "epoch": 9.846450707519589,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 42094
    },
    {
      "epoch": 9.846684598292597,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.5646,
      "step": 42095
    },
    {
      "epoch": 9.846918489065606,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 42096
    },
    {
      "epoch": 9.847152379838615,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 42097
    },
    {
      "epoch": 9.847386270611624,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.5122,
      "step": 42098
    },
    {
      "epoch": 9.847620161384633,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 42099
    },
    {
      "epoch": 9.847854052157642,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 42100
    },
    {
      "epoch": 9.847854052157642,
      "eval_runtime": 4.613,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 42100
    },
    {
      "epoch": 9.84808794293065,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7493,
      "step": 42101
    },
    {
      "epoch": 9.848321833703661,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 42102
    },
    {
      "epoch": 9.84855572447667,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7374,
      "step": 42103
    },
    {
      "epoch": 9.848789615249679,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 42104
    },
    {
      "epoch": 9.849023506022688,
      "grad_norm": 10.0625,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 42105
    },
    {
      "epoch": 9.849257396795696,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 42106
    },
    {
      "epoch": 9.849491287568705,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 42107
    },
    {
      "epoch": 9.849725178341714,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 42108
    },
    {
      "epoch": 9.849959069114723,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 42109
    },
    {
      "epoch": 9.850192959887732,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 42110
    },
    {
      "epoch": 9.85042685066074,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6826,
      "step": 42111
    },
    {
      "epoch": 9.850660741433751,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 42112
    },
    {
      "epoch": 9.85089463220676,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6654,
      "step": 42113
    },
    {
      "epoch": 9.851128522979769,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 42114
    },
    {
      "epoch": 9.851362413752778,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 42115
    },
    {
      "epoch": 9.851596304525787,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5523,
      "step": 42116
    },
    {
      "epoch": 9.851830195298795,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6688,
      "step": 42117
    },
    {
      "epoch": 9.852064086071804,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 42118
    },
    {
      "epoch": 9.852297976844813,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7371,
      "step": 42119
    },
    {
      "epoch": 9.852531867617822,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7458,
      "step": 42120
    },
    {
      "epoch": 9.85276575839083,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 42121
    },
    {
      "epoch": 9.852999649163841,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7532,
      "step": 42122
    },
    {
      "epoch": 9.85323353993685,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 42123
    },
    {
      "epoch": 9.853467430709859,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6371,
      "step": 42124
    },
    {
      "epoch": 9.853701321482868,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 42125
    },
    {
      "epoch": 9.853935212255877,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4597,
      "step": 42126
    },
    {
      "epoch": 9.854169103028886,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 42127
    },
    {
      "epoch": 9.854402993801894,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.6427,
      "step": 42128
    },
    {
      "epoch": 9.854636884574903,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.769,
      "step": 42129
    },
    {
      "epoch": 9.854870775347912,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.634,
      "step": 42130
    },
    {
      "epoch": 9.855104666120921,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 42131
    },
    {
      "epoch": 9.85533855689393,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5314,
      "step": 42132
    },
    {
      "epoch": 9.85557244766694,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 42133
    },
    {
      "epoch": 9.85580633843995,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.5293,
      "step": 42134
    },
    {
      "epoch": 9.856040229212958,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 42135
    },
    {
      "epoch": 9.856274119985967,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 42136
    },
    {
      "epoch": 9.856508010758976,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7725,
      "step": 42137
    },
    {
      "epoch": 9.856741901531985,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9831,
      "step": 42138
    },
    {
      "epoch": 9.856975792304993,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.4218,
      "step": 42139
    },
    {
      "epoch": 9.857209683078002,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5802,
      "step": 42140
    },
    {
      "epoch": 9.857443573851011,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.7396,
      "step": 42141
    },
    {
      "epoch": 9.85767746462402,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6525,
      "step": 42142
    },
    {
      "epoch": 9.857911355397029,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 42143
    },
    {
      "epoch": 9.85814524617004,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6454,
      "step": 42144
    },
    {
      "epoch": 9.858379136943048,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.5095,
      "step": 42145
    },
    {
      "epoch": 9.858613027716057,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 42146
    },
    {
      "epoch": 9.858846918489066,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 42147
    },
    {
      "epoch": 9.859080809262075,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 42148
    },
    {
      "epoch": 9.859314700035084,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 42149
    },
    {
      "epoch": 9.859548590808092,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.5153,
      "step": 42150
    },
    {
      "epoch": 9.859782481581101,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 42151
    },
    {
      "epoch": 9.86001637235411,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 42152
    },
    {
      "epoch": 9.860250263127119,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 42153
    },
    {
      "epoch": 9.860484153900128,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6834,
      "step": 42154
    },
    {
      "epoch": 9.860718044673138,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 42155
    },
    {
      "epoch": 9.860951935446147,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7858,
      "step": 42156
    },
    {
      "epoch": 9.861185826219156,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5762,
      "step": 42157
    },
    {
      "epoch": 9.861419716992165,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5994,
      "step": 42158
    },
    {
      "epoch": 9.861653607765174,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 42159
    },
    {
      "epoch": 9.861887498538183,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7935,
      "step": 42160
    },
    {
      "epoch": 9.862121389311191,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6023,
      "step": 42161
    },
    {
      "epoch": 9.8623552800842,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7758,
      "step": 42162
    },
    {
      "epoch": 9.862589170857209,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 42163
    },
    {
      "epoch": 9.862823061630218,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 42164
    },
    {
      "epoch": 9.863056952403227,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7171,
      "step": 42165
    },
    {
      "epoch": 9.863290843176237,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5075,
      "step": 42166
    },
    {
      "epoch": 9.863524733949246,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6545,
      "step": 42167
    },
    {
      "epoch": 9.863758624722255,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.5178,
      "step": 42168
    },
    {
      "epoch": 9.863992515495264,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 42169
    },
    {
      "epoch": 9.864226406268273,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 42170
    },
    {
      "epoch": 9.864460297041282,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6818,
      "step": 42171
    },
    {
      "epoch": 9.86469418781429,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 42172
    },
    {
      "epoch": 9.8649280785873,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6557,
      "step": 42173
    },
    {
      "epoch": 9.865161969360308,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.711,
      "step": 42174
    },
    {
      "epoch": 9.865395860133317,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6875,
      "step": 42175
    },
    {
      "epoch": 9.865629750906328,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4077,
      "step": 42176
    },
    {
      "epoch": 9.865863641679336,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8684,
      "step": 42177
    },
    {
      "epoch": 9.866097532452345,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.612,
      "step": 42178
    },
    {
      "epoch": 9.866331423225354,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.549,
      "step": 42179
    },
    {
      "epoch": 9.866565313998363,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.3805,
      "step": 42180
    },
    {
      "epoch": 9.866799204771372,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 42181
    },
    {
      "epoch": 9.86703309554438,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 42182
    },
    {
      "epoch": 9.86726698631739,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6771,
      "step": 42183
    },
    {
      "epoch": 9.867500877090398,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 42184
    },
    {
      "epoch": 9.867734767863407,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6992,
      "step": 42185
    },
    {
      "epoch": 9.867968658636418,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8075,
      "step": 42186
    },
    {
      "epoch": 9.868202549409427,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 42187
    },
    {
      "epoch": 9.868436440182435,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6989,
      "step": 42188
    },
    {
      "epoch": 9.868670330955444,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 42189
    },
    {
      "epoch": 9.868904221728453,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 42190
    },
    {
      "epoch": 9.869138112501462,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 42191
    },
    {
      "epoch": 9.86937200327447,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.3046,
      "step": 42192
    },
    {
      "epoch": 9.86960589404748,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 42193
    },
    {
      "epoch": 9.869839784820488,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.4789,
      "step": 42194
    },
    {
      "epoch": 9.870073675593497,
      "grad_norm": 11.875,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 42195
    },
    {
      "epoch": 9.870307566366506,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 42196
    },
    {
      "epoch": 9.870541457139517,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.4024,
      "step": 42197
    },
    {
      "epoch": 9.870775347912526,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 42198
    },
    {
      "epoch": 9.871009238685534,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 42199
    },
    {
      "epoch": 9.871243129458543,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 42200
    },
    {
      "epoch": 9.871243129458543,
      "eval_runtime": 4.6152,
      "eval_samples_per_second": 0.217,
      "eval_steps_per_second": 0.217,
      "step": 42200
    },
    {
      "epoch": 9.871477020231552,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5167,
      "step": 42201
    },
    {
      "epoch": 9.87171091100456,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5067,
      "step": 42202
    },
    {
      "epoch": 9.87194480177757,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7543,
      "step": 42203
    },
    {
      "epoch": 9.872178692550579,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7432,
      "step": 42204
    },
    {
      "epoch": 9.872412583323587,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 42205
    },
    {
      "epoch": 9.872646474096596,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4567,
      "step": 42206
    },
    {
      "epoch": 9.872880364869605,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 42207
    },
    {
      "epoch": 9.873114255642616,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 42208
    },
    {
      "epoch": 9.873348146415625,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 42209
    },
    {
      "epoch": 9.873582037188633,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 42210
    },
    {
      "epoch": 9.873815927961642,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.649,
      "step": 42211
    },
    {
      "epoch": 9.874049818734651,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6796,
      "step": 42212
    },
    {
      "epoch": 9.87428370950766,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 42213
    },
    {
      "epoch": 9.874517600280669,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 42214
    },
    {
      "epoch": 9.874751491053678,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 42215
    },
    {
      "epoch": 9.874985381826686,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 42216
    },
    {
      "epoch": 9.875219272599695,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6957,
      "step": 42217
    },
    {
      "epoch": 9.875453163372704,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 42218
    },
    {
      "epoch": 9.875687054145715,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 42219
    },
    {
      "epoch": 9.875920944918724,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5031,
      "step": 42220
    },
    {
      "epoch": 9.876154835691732,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6835,
      "step": 42221
    },
    {
      "epoch": 9.876388726464741,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7059,
      "step": 42222
    },
    {
      "epoch": 9.87662261723775,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 42223
    },
    {
      "epoch": 9.876856508010759,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6674,
      "step": 42224
    },
    {
      "epoch": 9.877090398783768,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 42225
    },
    {
      "epoch": 9.877324289556777,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 42226
    },
    {
      "epoch": 9.877558180329785,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.7631,
      "step": 42227
    },
    {
      "epoch": 9.877792071102794,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 42228
    },
    {
      "epoch": 9.878025961875803,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 42229
    },
    {
      "epoch": 9.878259852648814,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.3953,
      "step": 42230
    },
    {
      "epoch": 9.878493743421823,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.6641,
      "step": 42231
    },
    {
      "epoch": 9.878727634194831,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7482,
      "step": 42232
    },
    {
      "epoch": 9.87896152496784,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.639,
      "step": 42233
    },
    {
      "epoch": 9.879195415740849,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5982,
      "step": 42234
    },
    {
      "epoch": 9.879429306513858,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 42235
    },
    {
      "epoch": 9.879663197286867,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 42236
    },
    {
      "epoch": 9.879897088059876,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6572,
      "step": 42237
    },
    {
      "epoch": 9.880130978832884,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5398,
      "step": 42238
    },
    {
      "epoch": 9.880364869605893,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 42239
    },
    {
      "epoch": 9.880598760378904,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7835,
      "step": 42240
    },
    {
      "epoch": 9.880832651151913,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4171,
      "step": 42241
    },
    {
      "epoch": 9.881066541924922,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.691,
      "step": 42242
    },
    {
      "epoch": 9.88130043269793,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8075,
      "step": 42243
    },
    {
      "epoch": 9.88153432347094,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8216,
      "step": 42244
    },
    {
      "epoch": 9.881768214243948,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6819,
      "step": 42245
    },
    {
      "epoch": 9.882002105016957,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 42246
    },
    {
      "epoch": 9.882235995789966,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 42247
    },
    {
      "epoch": 9.882469886562975,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 42248
    },
    {
      "epoch": 9.882703777335983,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 42249
    },
    {
      "epoch": 9.882937668108994,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.5719,
      "step": 42250
    },
    {
      "epoch": 9.883171558882003,
      "grad_norm": 2.59375,
      "learning_rate": 3e-05,
      "loss": 1.5447,
      "step": 42251
    },
    {
      "epoch": 9.883405449655012,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 42252
    },
    {
      "epoch": 9.88363934042802,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 42253
    },
    {
      "epoch": 9.88387323120103,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 42254
    },
    {
      "epoch": 9.884107121974038,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5691,
      "step": 42255
    },
    {
      "epoch": 9.884341012747047,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5496,
      "step": 42256
    },
    {
      "epoch": 9.884574903520056,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5607,
      "step": 42257
    },
    {
      "epoch": 9.884808794293065,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 42258
    },
    {
      "epoch": 9.885042685066074,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.4865,
      "step": 42259
    },
    {
      "epoch": 9.885276575839082,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 42260
    },
    {
      "epoch": 9.885510466612093,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7027,
      "step": 42261
    },
    {
      "epoch": 9.885744357385102,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 42262
    },
    {
      "epoch": 9.88597824815811,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7807,
      "step": 42263
    },
    {
      "epoch": 9.88621213893112,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.569,
      "step": 42264
    },
    {
      "epoch": 9.886446029704128,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7317,
      "step": 42265
    },
    {
      "epoch": 9.886679920477137,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.5739,
      "step": 42266
    },
    {
      "epoch": 9.886913811250146,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 42267
    },
    {
      "epoch": 9.887147702023155,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 42268
    },
    {
      "epoch": 9.887381592796164,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 42269
    },
    {
      "epoch": 9.887615483569173,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6157,
      "step": 42270
    },
    {
      "epoch": 9.887849374342181,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.496,
      "step": 42271
    },
    {
      "epoch": 9.888083265115192,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6419,
      "step": 42272
    },
    {
      "epoch": 9.8883171558882,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6183,
      "step": 42273
    },
    {
      "epoch": 9.88855104666121,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 42274
    },
    {
      "epoch": 9.888784937434218,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 42275
    },
    {
      "epoch": 9.889018828207227,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.6204,
      "step": 42276
    },
    {
      "epoch": 9.889252718980236,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 42277
    },
    {
      "epoch": 9.889486609753245,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 42278
    },
    {
      "epoch": 9.889720500526254,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.569,
      "step": 42279
    },
    {
      "epoch": 9.889954391299263,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 42280
    },
    {
      "epoch": 9.890188282072272,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7435,
      "step": 42281
    },
    {
      "epoch": 9.89042217284528,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7008,
      "step": 42282
    },
    {
      "epoch": 9.890656063618291,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8411,
      "step": 42283
    },
    {
      "epoch": 9.8908899543913,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8416,
      "step": 42284
    },
    {
      "epoch": 9.891123845164309,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.4933,
      "step": 42285
    },
    {
      "epoch": 9.891357735937317,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5058,
      "step": 42286
    },
    {
      "epoch": 9.891591626710326,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6855,
      "step": 42287
    },
    {
      "epoch": 9.891825517483335,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 42288
    },
    {
      "epoch": 9.892059408256344,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 42289
    },
    {
      "epoch": 9.892293299029353,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 42290
    },
    {
      "epoch": 9.892527189802362,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 42291
    },
    {
      "epoch": 9.89276108057537,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 42292
    },
    {
      "epoch": 9.892994971348381,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.4029,
      "step": 42293
    },
    {
      "epoch": 9.89322886212139,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7674,
      "step": 42294
    },
    {
      "epoch": 9.893462752894399,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 42295
    },
    {
      "epoch": 9.893696643667408,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 42296
    },
    {
      "epoch": 9.893930534440416,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 42297
    },
    {
      "epoch": 9.894164425213425,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6858,
      "step": 42298
    },
    {
      "epoch": 9.894398315986434,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 42299
    },
    {
      "epoch": 9.894632206759443,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7096,
      "step": 42300
    },
    {
      "epoch": 9.894632206759443,
      "eval_runtime": 4.6662,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 42300
    },
    {
      "epoch": 9.894866097532452,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6656,
      "step": 42301
    },
    {
      "epoch": 9.89509998830546,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7297,
      "step": 42302
    },
    {
      "epoch": 9.895333879078471,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.75,
      "step": 42303
    },
    {
      "epoch": 9.89556776985148,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6942,
      "step": 42304
    },
    {
      "epoch": 9.895801660624489,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 42305
    },
    {
      "epoch": 9.896035551397498,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 42306
    },
    {
      "epoch": 9.896269442170507,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.5183,
      "step": 42307
    },
    {
      "epoch": 9.896503332943515,
      "grad_norm": 7.3125,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 42308
    },
    {
      "epoch": 9.896737223716524,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6066,
      "step": 42309
    },
    {
      "epoch": 9.896971114489533,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 42310
    },
    {
      "epoch": 9.897205005262542,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6546,
      "step": 42311
    },
    {
      "epoch": 9.89743889603555,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7472,
      "step": 42312
    },
    {
      "epoch": 9.89767278680856,
      "grad_norm": 7.65625,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 42313
    },
    {
      "epoch": 9.89790667758157,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7292,
      "step": 42314
    },
    {
      "epoch": 9.89814056835458,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 42315
    },
    {
      "epoch": 9.898374459127588,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7575,
      "step": 42316
    },
    {
      "epoch": 9.898608349900597,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6592,
      "step": 42317
    },
    {
      "epoch": 9.898842240673606,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 42318
    },
    {
      "epoch": 9.899076131446614,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 42319
    },
    {
      "epoch": 9.899310022219623,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5482,
      "step": 42320
    },
    {
      "epoch": 9.899543912992632,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6892,
      "step": 42321
    },
    {
      "epoch": 9.899777803765641,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 42322
    },
    {
      "epoch": 9.90001169453865,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 42323
    },
    {
      "epoch": 9.900245585311659,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 42324
    },
    {
      "epoch": 9.90047947608467,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 42325
    },
    {
      "epoch": 9.900713366857678,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.4116,
      "step": 42326
    },
    {
      "epoch": 9.900947257630687,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 42327
    },
    {
      "epoch": 9.901181148403696,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7129,
      "step": 42328
    },
    {
      "epoch": 9.901415039176705,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 42329
    },
    {
      "epoch": 9.901648929949713,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 42330
    },
    {
      "epoch": 9.901882820722722,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 42331
    },
    {
      "epoch": 9.902116711495731,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.5111,
      "step": 42332
    },
    {
      "epoch": 9.90235060226874,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 42333
    },
    {
      "epoch": 9.902584493041749,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 42334
    },
    {
      "epoch": 9.902818383814758,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 42335
    },
    {
      "epoch": 9.903052274587768,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.5603,
      "step": 42336
    },
    {
      "epoch": 9.903286165360777,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 42337
    },
    {
      "epoch": 9.903520056133786,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 42338
    },
    {
      "epoch": 9.903753946906795,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 42339
    },
    {
      "epoch": 9.903987837679804,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.6234,
      "step": 42340
    },
    {
      "epoch": 9.904221728452812,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 42341
    },
    {
      "epoch": 9.904455619225821,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7229,
      "step": 42342
    },
    {
      "epoch": 9.90468950999883,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 42343
    },
    {
      "epoch": 9.904923400771839,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 42344
    },
    {
      "epoch": 9.905157291544848,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5846,
      "step": 42345
    },
    {
      "epoch": 9.905391182317857,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 42346
    },
    {
      "epoch": 9.905625073090867,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 42347
    },
    {
      "epoch": 9.905858963863876,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.6335,
      "step": 42348
    },
    {
      "epoch": 9.906092854636885,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9506,
      "step": 42349
    },
    {
      "epoch": 9.906326745409894,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.622,
      "step": 42350
    },
    {
      "epoch": 9.906560636182903,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 42351
    },
    {
      "epoch": 9.906794526955911,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 42352
    },
    {
      "epoch": 9.90702841772892,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 42353
    },
    {
      "epoch": 9.90726230850193,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 42354
    },
    {
      "epoch": 9.907496199274938,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.747,
      "step": 42355
    },
    {
      "epoch": 9.907730090047947,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.5832,
      "step": 42356
    },
    {
      "epoch": 9.907963980820957,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7433,
      "step": 42357
    },
    {
      "epoch": 9.908197871593966,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 42358
    },
    {
      "epoch": 9.908431762366975,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6929,
      "step": 42359
    },
    {
      "epoch": 9.908665653139984,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6804,
      "step": 42360
    },
    {
      "epoch": 9.908899543912993,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 42361
    },
    {
      "epoch": 9.909133434686002,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 42362
    },
    {
      "epoch": 9.90936732545901,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 42363
    },
    {
      "epoch": 9.90960121623202,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 42364
    },
    {
      "epoch": 9.909835107005028,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.5982,
      "step": 42365
    },
    {
      "epoch": 9.910068997778037,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 42366
    },
    {
      "epoch": 9.910302888551048,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 42367
    },
    {
      "epoch": 9.910536779324056,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 42368
    },
    {
      "epoch": 9.910770670097065,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 42369
    },
    {
      "epoch": 9.911004560870074,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1595,
      "step": 42370
    },
    {
      "epoch": 9.911238451643083,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6229,
      "step": 42371
    },
    {
      "epoch": 9.911472342416092,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 42372
    },
    {
      "epoch": 9.9117062331891,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6046,
      "step": 42373
    },
    {
      "epoch": 9.91194012396211,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 42374
    },
    {
      "epoch": 9.912174014735118,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 42375
    },
    {
      "epoch": 9.912407905508127,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 42376
    },
    {
      "epoch": 9.912641796281136,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6821,
      "step": 42377
    },
    {
      "epoch": 9.912875687054147,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7223,
      "step": 42378
    },
    {
      "epoch": 9.913109577827155,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7902,
      "step": 42379
    },
    {
      "epoch": 9.913343468600164,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 42380
    },
    {
      "epoch": 9.913577359373173,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 42381
    },
    {
      "epoch": 9.913811250146182,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.4471,
      "step": 42382
    },
    {
      "epoch": 9.91404514091919,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 42383
    },
    {
      "epoch": 9.9142790316922,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7617,
      "step": 42384
    },
    {
      "epoch": 9.914512922465208,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8045,
      "step": 42385
    },
    {
      "epoch": 9.914746813238217,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 42386
    },
    {
      "epoch": 9.914980704011226,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 42387
    },
    {
      "epoch": 9.915214594784235,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 42388
    },
    {
      "epoch": 9.915448485557246,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.573,
      "step": 42389
    },
    {
      "epoch": 9.915682376330254,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.5097,
      "step": 42390
    },
    {
      "epoch": 9.915916267103263,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.4484,
      "step": 42391
    },
    {
      "epoch": 9.916150157876272,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6507,
      "step": 42392
    },
    {
      "epoch": 9.916384048649281,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 42393
    },
    {
      "epoch": 9.91661793942229,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7825,
      "step": 42394
    },
    {
      "epoch": 9.916851830195299,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.4814,
      "step": 42395
    },
    {
      "epoch": 9.917085720968307,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.5814,
      "step": 42396
    },
    {
      "epoch": 9.917319611741316,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 42397
    },
    {
      "epoch": 9.917553502514325,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.4162,
      "step": 42398
    },
    {
      "epoch": 9.917787393287334,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.4951,
      "step": 42399
    },
    {
      "epoch": 9.918021284060345,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.6637,
      "step": 42400
    },
    {
      "epoch": 9.918021284060345,
      "eval_runtime": 4.6702,
      "eval_samples_per_second": 0.214,
      "eval_steps_per_second": 0.214,
      "step": 42400
    },
    {
      "epoch": 9.918255174833353,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.7948,
      "step": 42401
    },
    {
      "epoch": 9.918489065606362,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 42402
    },
    {
      "epoch": 9.918722956379371,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.5017,
      "step": 42403
    },
    {
      "epoch": 9.91895684715238,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5061,
      "step": 42404
    },
    {
      "epoch": 9.919190737925389,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7428,
      "step": 42405
    },
    {
      "epoch": 9.919424628698398,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 42406
    },
    {
      "epoch": 9.919658519471406,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7749,
      "step": 42407
    },
    {
      "epoch": 9.919892410244415,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 42408
    },
    {
      "epoch": 9.920126301017424,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 42409
    },
    {
      "epoch": 9.920360191790433,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.638,
      "step": 42410
    },
    {
      "epoch": 9.920594082563444,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 42411
    },
    {
      "epoch": 9.920827973336452,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 42412
    },
    {
      "epoch": 9.921061864109461,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 42413
    },
    {
      "epoch": 9.92129575488247,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7763,
      "step": 42414
    },
    {
      "epoch": 9.921529645655479,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6512,
      "step": 42415
    },
    {
      "epoch": 9.921763536428488,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.724,
      "step": 42416
    },
    {
      "epoch": 9.921997427201497,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.4894,
      "step": 42417
    },
    {
      "epoch": 9.922231317974505,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7172,
      "step": 42418
    },
    {
      "epoch": 9.922465208747514,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.6899,
      "step": 42419
    },
    {
      "epoch": 9.922699099520523,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 42420
    },
    {
      "epoch": 9.922932990293534,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9244,
      "step": 42421
    },
    {
      "epoch": 9.923166881066543,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 42422
    },
    {
      "epoch": 9.923400771839551,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 42423
    },
    {
      "epoch": 9.92363466261256,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.6016,
      "step": 42424
    },
    {
      "epoch": 9.923868553385569,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7378,
      "step": 42425
    },
    {
      "epoch": 9.924102444158578,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.5151,
      "step": 42426
    },
    {
      "epoch": 9.924336334931587,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7598,
      "step": 42427
    },
    {
      "epoch": 9.924570225704596,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.5811,
      "step": 42428
    },
    {
      "epoch": 9.924804116477604,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 42429
    },
    {
      "epoch": 9.925038007250613,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 42430
    },
    {
      "epoch": 9.925271898023624,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 42431
    },
    {
      "epoch": 9.925505788796633,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.56,
      "step": 42432
    },
    {
      "epoch": 9.925739679569642,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.632,
      "step": 42433
    },
    {
      "epoch": 9.92597357034265,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 42434
    },
    {
      "epoch": 9.92620746111566,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 42435
    },
    {
      "epoch": 9.926441351888668,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.324,
      "step": 42436
    },
    {
      "epoch": 9.926675242661677,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6574,
      "step": 42437
    },
    {
      "epoch": 9.926909133434686,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 42438
    },
    {
      "epoch": 9.927143024207695,
      "grad_norm": 14.4375,
      "learning_rate": 3e-05,
      "loss": 1.8369,
      "step": 42439
    },
    {
      "epoch": 9.927376914980703,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.5617,
      "step": 42440
    },
    {
      "epoch": 9.927610805753712,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 42441
    },
    {
      "epoch": 9.927844696526723,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 42442
    },
    {
      "epoch": 9.928078587299732,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 42443
    },
    {
      "epoch": 9.92831247807274,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.4334,
      "step": 42444
    },
    {
      "epoch": 9.92854636884575,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 42445
    },
    {
      "epoch": 9.928780259618758,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 42446
    },
    {
      "epoch": 9.929014150391767,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.3839,
      "step": 42447
    },
    {
      "epoch": 9.929248041164776,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7238,
      "step": 42448
    },
    {
      "epoch": 9.929481931937785,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 42449
    },
    {
      "epoch": 9.929715822710794,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.694,
      "step": 42450
    },
    {
      "epoch": 9.929949713483802,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 42451
    },
    {
      "epoch": 9.930183604256811,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.5293,
      "step": 42452
    },
    {
      "epoch": 9.930417495029822,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 42453
    },
    {
      "epoch": 9.93065138580283,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6752,
      "step": 42454
    },
    {
      "epoch": 9.93088527657584,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 42455
    },
    {
      "epoch": 9.931119167348848,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8688,
      "step": 42456
    },
    {
      "epoch": 9.931353058121857,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7029,
      "step": 42457
    },
    {
      "epoch": 9.931586948894866,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 42458
    },
    {
      "epoch": 9.931820839667875,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6351,
      "step": 42459
    },
    {
      "epoch": 9.932054730440884,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 42460
    },
    {
      "epoch": 9.932288621213893,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.6108,
      "step": 42461
    },
    {
      "epoch": 9.932522511986901,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6185,
      "step": 42462
    },
    {
      "epoch": 9.93275640275991,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.762,
      "step": 42463
    },
    {
      "epoch": 9.93299029353292,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.4186,
      "step": 42464
    },
    {
      "epoch": 9.93322418430593,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6306,
      "step": 42465
    },
    {
      "epoch": 9.933458075078939,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 42466
    },
    {
      "epoch": 9.933691965851947,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 42467
    },
    {
      "epoch": 9.933925856624956,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 42468
    },
    {
      "epoch": 9.934159747397965,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.5856,
      "step": 42469
    },
    {
      "epoch": 9.934393638170974,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7308,
      "step": 42470
    },
    {
      "epoch": 9.934627528943983,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 42471
    },
    {
      "epoch": 9.934861419716992,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 42472
    },
    {
      "epoch": 9.93509531049,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5296,
      "step": 42473
    },
    {
      "epoch": 9.935329201263011,
      "grad_norm": 7.96875,
      "learning_rate": 3e-05,
      "loss": 1.7041,
      "step": 42474
    },
    {
      "epoch": 9.93556309203602,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7019,
      "step": 42475
    },
    {
      "epoch": 9.935796982809029,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 42476
    },
    {
      "epoch": 9.936030873582038,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8601,
      "step": 42477
    },
    {
      "epoch": 9.936264764355046,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6352,
      "step": 42478
    },
    {
      "epoch": 9.936498655128055,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4542,
      "step": 42479
    },
    {
      "epoch": 9.936732545901064,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 42480
    },
    {
      "epoch": 9.936966436674073,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.685,
      "step": 42481
    },
    {
      "epoch": 9.937200327447082,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 42482
    },
    {
      "epoch": 9.93743421822009,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7733,
      "step": 42483
    },
    {
      "epoch": 9.937668108993101,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 42484
    },
    {
      "epoch": 9.93790199976611,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7267,
      "step": 42485
    },
    {
      "epoch": 9.938135890539119,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.5781,
      "step": 42486
    },
    {
      "epoch": 9.938369781312128,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 42487
    },
    {
      "epoch": 9.938603672085137,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6166,
      "step": 42488
    },
    {
      "epoch": 9.938837562858145,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 42489
    },
    {
      "epoch": 9.939071453631154,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6214,
      "step": 42490
    },
    {
      "epoch": 9.939305344404163,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7115,
      "step": 42491
    },
    {
      "epoch": 9.939539235177172,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 42492
    },
    {
      "epoch": 9.93977312595018,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 42493
    },
    {
      "epoch": 9.94000701672319,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6675,
      "step": 42494
    },
    {
      "epoch": 9.9402409074962,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 42495
    },
    {
      "epoch": 9.940474798269209,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7097,
      "step": 42496
    },
    {
      "epoch": 9.940708689042218,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7544,
      "step": 42497
    },
    {
      "epoch": 9.940942579815227,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.768,
      "step": 42498
    },
    {
      "epoch": 9.941176470588236,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.7589,
      "step": 42499
    },
    {
      "epoch": 9.941410361361244,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.886,
      "step": 42500
    },
    {
      "epoch": 9.941410361361244,
      "eval_runtime": 4.6603,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 42500
    },
    {
      "epoch": 9.941644252134253,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 42501
    },
    {
      "epoch": 9.941878142907262,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 42502
    },
    {
      "epoch": 9.94211203368027,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 42503
    },
    {
      "epoch": 9.94234592445328,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7558,
      "step": 42504
    },
    {
      "epoch": 9.942579815226289,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7108,
      "step": 42505
    },
    {
      "epoch": 9.9428137059993,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7178,
      "step": 42506
    },
    {
      "epoch": 9.943047596772308,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 42507
    },
    {
      "epoch": 9.943281487545317,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.5484,
      "step": 42508
    },
    {
      "epoch": 9.943515378318326,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.565,
      "step": 42509
    },
    {
      "epoch": 9.943749269091334,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.4448,
      "step": 42510
    },
    {
      "epoch": 9.943983159864343,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.6663,
      "step": 42511
    },
    {
      "epoch": 9.944217050637352,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 42512
    },
    {
      "epoch": 9.944450941410361,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8051,
      "step": 42513
    },
    {
      "epoch": 9.94468483218337,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 42514
    },
    {
      "epoch": 9.944918722956379,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 42515
    },
    {
      "epoch": 9.945152613729388,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 42516
    },
    {
      "epoch": 9.945386504502398,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 42517
    },
    {
      "epoch": 9.945620395275407,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5577,
      "step": 42518
    },
    {
      "epoch": 9.945854286048416,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.7187,
      "step": 42519
    },
    {
      "epoch": 9.946088176821425,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.707,
      "step": 42520
    },
    {
      "epoch": 9.946322067594433,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.4738,
      "step": 42521
    },
    {
      "epoch": 9.946555958367442,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.782,
      "step": 42522
    },
    {
      "epoch": 9.946789849140451,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 42523
    },
    {
      "epoch": 9.94702373991346,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 42524
    },
    {
      "epoch": 9.947257630686469,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.6198,
      "step": 42525
    },
    {
      "epoch": 9.947491521459478,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.5606,
      "step": 42526
    },
    {
      "epoch": 9.947725412232487,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 42527
    },
    {
      "epoch": 9.947959303005497,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5697,
      "step": 42528
    },
    {
      "epoch": 9.948193193778506,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 42529
    },
    {
      "epoch": 9.948427084551515,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 42530
    },
    {
      "epoch": 9.948660975324524,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 42531
    },
    {
      "epoch": 9.948894866097532,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.5348,
      "step": 42532
    },
    {
      "epoch": 9.949128756870541,
      "grad_norm": 10.1875,
      "learning_rate": 3e-05,
      "loss": 2.2241,
      "step": 42533
    },
    {
      "epoch": 9.94936264764355,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5311,
      "step": 42534
    },
    {
      "epoch": 9.949596538416559,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 42535
    },
    {
      "epoch": 9.949830429189568,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 42536
    },
    {
      "epoch": 9.950064319962577,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.4736,
      "step": 42537
    },
    {
      "epoch": 9.950298210735587,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6652,
      "step": 42538
    },
    {
      "epoch": 9.950532101508596,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7629,
      "step": 42539
    },
    {
      "epoch": 9.950765992281605,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 42540
    },
    {
      "epoch": 9.950999883054614,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6246,
      "step": 42541
    },
    {
      "epoch": 9.951233773827623,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8462,
      "step": 42542
    },
    {
      "epoch": 9.951467664600631,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 42543
    },
    {
      "epoch": 9.95170155537364,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 42544
    },
    {
      "epoch": 9.95193544614665,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6685,
      "step": 42545
    },
    {
      "epoch": 9.952169336919658,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6367,
      "step": 42546
    },
    {
      "epoch": 9.952403227692667,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6207,
      "step": 42547
    },
    {
      "epoch": 9.952637118465677,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 42548
    },
    {
      "epoch": 9.952871009238686,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.6734,
      "step": 42549
    },
    {
      "epoch": 9.953104900011695,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 42550
    },
    {
      "epoch": 9.953338790784704,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7325,
      "step": 42551
    },
    {
      "epoch": 9.953572681557713,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.4019,
      "step": 42552
    },
    {
      "epoch": 9.953806572330722,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 42553
    },
    {
      "epoch": 9.95404046310373,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 42554
    },
    {
      "epoch": 9.95427435387674,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6914,
      "step": 42555
    },
    {
      "epoch": 9.954508244649748,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 42556
    },
    {
      "epoch": 9.954742135422757,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8449,
      "step": 42557
    },
    {
      "epoch": 9.954976026195766,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.5608,
      "step": 42558
    },
    {
      "epoch": 9.955209916968776,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 42559
    },
    {
      "epoch": 9.955443807741785,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 42560
    },
    {
      "epoch": 9.955677698514794,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6809,
      "step": 42561
    },
    {
      "epoch": 9.955911589287803,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 42562
    },
    {
      "epoch": 9.956145480060812,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4277,
      "step": 42563
    },
    {
      "epoch": 9.95637937083382,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.6853,
      "step": 42564
    },
    {
      "epoch": 9.95661326160683,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7804,
      "step": 42565
    },
    {
      "epoch": 9.956847152379838,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.5009,
      "step": 42566
    },
    {
      "epoch": 9.957081043152847,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6874,
      "step": 42567
    },
    {
      "epoch": 9.957314933925856,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.5804,
      "step": 42568
    },
    {
      "epoch": 9.957548824698865,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6827,
      "step": 42569
    },
    {
      "epoch": 9.957782715471875,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.5413,
      "step": 42570
    },
    {
      "epoch": 9.958016606244884,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 42571
    },
    {
      "epoch": 9.958250497017893,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 42572
    },
    {
      "epoch": 9.958484387790902,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 42573
    },
    {
      "epoch": 9.95871827856391,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7496,
      "step": 42574
    },
    {
      "epoch": 9.95895216933692,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.6546,
      "step": 42575
    },
    {
      "epoch": 9.959186060109928,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 42576
    },
    {
      "epoch": 9.959419950882937,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.6465,
      "step": 42577
    },
    {
      "epoch": 9.959653841655946,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 42578
    },
    {
      "epoch": 9.959887732428955,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8891,
      "step": 42579
    },
    {
      "epoch": 9.960121623201964,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.408,
      "step": 42580
    },
    {
      "epoch": 9.960355513974974,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 42581
    },
    {
      "epoch": 9.960589404747983,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.6114,
      "step": 42582
    },
    {
      "epoch": 9.960823295520992,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8645,
      "step": 42583
    },
    {
      "epoch": 9.961057186294001,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 42584
    },
    {
      "epoch": 9.96129107706701,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6233,
      "step": 42585
    },
    {
      "epoch": 9.961524967840019,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 42586
    },
    {
      "epoch": 9.961758858613027,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 42587
    },
    {
      "epoch": 9.961992749386036,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7038,
      "step": 42588
    },
    {
      "epoch": 9.962226640159045,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6464,
      "step": 42589
    },
    {
      "epoch": 9.962460530932054,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 42590
    },
    {
      "epoch": 9.962694421705063,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7014,
      "step": 42591
    },
    {
      "epoch": 9.962928312478073,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 42592
    },
    {
      "epoch": 9.963162203251082,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8701,
      "step": 42593
    },
    {
      "epoch": 9.963396094024091,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 42594
    },
    {
      "epoch": 9.9636299847971,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 42595
    },
    {
      "epoch": 9.963863875570109,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 42596
    },
    {
      "epoch": 9.964097766343118,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6893,
      "step": 42597
    },
    {
      "epoch": 9.964331657116126,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 42598
    },
    {
      "epoch": 9.964565547889135,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 42599
    },
    {
      "epoch": 9.964799438662144,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.5441,
      "step": 42600
    },
    {
      "epoch": 9.964799438662144,
      "eval_runtime": 4.6501,
      "eval_samples_per_second": 0.215,
      "eval_steps_per_second": 0.215,
      "step": 42600
    },
    {
      "epoch": 9.965033329435153,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6496,
      "step": 42601
    },
    {
      "epoch": 9.965267220208164,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6245,
      "step": 42602
    },
    {
      "epoch": 9.965501110981172,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 42603
    },
    {
      "epoch": 9.965735001754181,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 42604
    },
    {
      "epoch": 9.96596889252719,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6344,
      "step": 42605
    },
    {
      "epoch": 9.966202783300199,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 42606
    },
    {
      "epoch": 9.966436674073208,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 42607
    },
    {
      "epoch": 9.966670564846217,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.6615,
      "step": 42608
    },
    {
      "epoch": 9.966904455619225,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6676,
      "step": 42609
    },
    {
      "epoch": 9.967138346392234,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 42610
    },
    {
      "epoch": 9.967372237165243,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.4225,
      "step": 42611
    },
    {
      "epoch": 9.967606127938254,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 42612
    },
    {
      "epoch": 9.967840018711263,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 42613
    },
    {
      "epoch": 9.968073909484271,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8535,
      "step": 42614
    },
    {
      "epoch": 9.96830780025728,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 42615
    },
    {
      "epoch": 9.968541691030289,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.2679,
      "step": 42616
    },
    {
      "epoch": 9.968775581803298,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 42617
    },
    {
      "epoch": 9.969009472576307,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 42618
    },
    {
      "epoch": 9.969243363349316,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 42619
    },
    {
      "epoch": 9.969477254122324,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 42620
    },
    {
      "epoch": 9.969711144895333,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7749,
      "step": 42621
    },
    {
      "epoch": 9.969945035668342,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8728,
      "step": 42622
    },
    {
      "epoch": 9.970178926441353,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.4206,
      "step": 42623
    },
    {
      "epoch": 9.970412817214362,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 42624
    },
    {
      "epoch": 9.97064670798737,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.6303,
      "step": 42625
    },
    {
      "epoch": 9.97088059876038,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7587,
      "step": 42626
    },
    {
      "epoch": 9.971114489533388,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.4854,
      "step": 42627
    },
    {
      "epoch": 9.971348380306397,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 42628
    },
    {
      "epoch": 9.971582271079406,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 42629
    },
    {
      "epoch": 9.971816161852415,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7619,
      "step": 42630
    },
    {
      "epoch": 9.972050052625423,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6074,
      "step": 42631
    },
    {
      "epoch": 9.972283943398432,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 42632
    },
    {
      "epoch": 9.972517834171441,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6091,
      "step": 42633
    },
    {
      "epoch": 9.972751724944452,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8111,
      "step": 42634
    },
    {
      "epoch": 9.97298561571746,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6333,
      "step": 42635
    },
    {
      "epoch": 9.97321950649047,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 42636
    },
    {
      "epoch": 9.973453397263478,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 42637
    },
    {
      "epoch": 9.973687288036487,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 42638
    },
    {
      "epoch": 9.973921178809496,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.5614,
      "step": 42639
    },
    {
      "epoch": 9.974155069582505,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 42640
    },
    {
      "epoch": 9.974388960355514,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 42641
    },
    {
      "epoch": 9.974622851128522,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 42642
    },
    {
      "epoch": 9.974856741901531,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8693,
      "step": 42643
    },
    {
      "epoch": 9.97509063267454,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7687,
      "step": 42644
    },
    {
      "epoch": 9.97532452344755,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.6111,
      "step": 42645
    },
    {
      "epoch": 9.97555841422056,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 42646
    },
    {
      "epoch": 9.975792304993568,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 42647
    },
    {
      "epoch": 9.976026195766577,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.5408,
      "step": 42648
    },
    {
      "epoch": 9.976260086539586,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 42649
    },
    {
      "epoch": 9.976493977312595,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 42650
    },
    {
      "epoch": 9.976727868085604,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 42651
    },
    {
      "epoch": 9.976961758858613,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7875,
      "step": 42652
    },
    {
      "epoch": 9.977195649631621,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7383,
      "step": 42653
    },
    {
      "epoch": 9.97742954040463,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6896,
      "step": 42654
    },
    {
      "epoch": 9.97766343117764,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 42655
    },
    {
      "epoch": 9.97789732195065,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.4834,
      "step": 42656
    },
    {
      "epoch": 9.978131212723659,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7801,
      "step": 42657
    },
    {
      "epoch": 9.978365103496667,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 42658
    },
    {
      "epoch": 9.978598994269676,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.6259,
      "step": 42659
    },
    {
      "epoch": 9.978832885042685,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 42660
    },
    {
      "epoch": 9.979066775815694,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6147,
      "step": 42661
    },
    {
      "epoch": 9.979300666588703,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7653,
      "step": 42662
    },
    {
      "epoch": 9.979534557361712,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7826,
      "step": 42663
    },
    {
      "epoch": 9.97976844813472,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 42664
    },
    {
      "epoch": 9.980002338907731,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 42665
    },
    {
      "epoch": 9.98023622968074,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7997,
      "step": 42666
    },
    {
      "epoch": 9.980470120453749,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.5035,
      "step": 42667
    },
    {
      "epoch": 9.980704011226758,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 42668
    },
    {
      "epoch": 9.980937901999766,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 42669
    },
    {
      "epoch": 9.981171792772775,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6244,
      "step": 42670
    },
    {
      "epoch": 9.981405683545784,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 42671
    },
    {
      "epoch": 9.981639574318793,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 42672
    },
    {
      "epoch": 9.981873465091802,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7234,
      "step": 42673
    },
    {
      "epoch": 9.98210735586481,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.4437,
      "step": 42674
    },
    {
      "epoch": 9.98234124663782,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 42675
    },
    {
      "epoch": 9.98257513741083,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 42676
    },
    {
      "epoch": 9.982809028183839,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 42677
    },
    {
      "epoch": 9.983042918956848,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6947,
      "step": 42678
    },
    {
      "epoch": 9.983276809729857,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 42679
    },
    {
      "epoch": 9.983510700502865,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6608,
      "step": 42680
    },
    {
      "epoch": 9.983744591275874,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 42681
    },
    {
      "epoch": 9.983978482048883,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 42682
    },
    {
      "epoch": 9.984212372821892,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.6436,
      "step": 42683
    },
    {
      "epoch": 9.9844462635949,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.5461,
      "step": 42684
    },
    {
      "epoch": 9.98468015436791,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2192,
      "step": 42685
    },
    {
      "epoch": 9.984914045140918,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 42686
    },
    {
      "epoch": 9.985147935913929,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.5716,
      "step": 42687
    },
    {
      "epoch": 9.985381826686938,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.68,
      "step": 42688
    },
    {
      "epoch": 9.985615717459947,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 42689
    },
    {
      "epoch": 9.985849608232956,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7767,
      "step": 42690
    },
    {
      "epoch": 9.986083499005964,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 42691
    },
    {
      "epoch": 9.986317389778973,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 42692
    },
    {
      "epoch": 9.986551280551982,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 42693
    },
    {
      "epoch": 9.98678517132499,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 42694
    },
    {
      "epoch": 9.987019062098,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.4603,
      "step": 42695
    },
    {
      "epoch": 9.987252952871009,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 42696
    },
    {
      "epoch": 9.987486843644017,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.699,
      "step": 42697
    },
    {
      "epoch": 9.987720734417028,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 42698
    },
    {
      "epoch": 9.987954625190037,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7025,
      "step": 42699
    },
    {
      "epoch": 9.988188515963046,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7393,
      "step": 42700
    },
    {
      "epoch": 9.988188515963046,
      "eval_runtime": 4.5897,
      "eval_samples_per_second": 0.218,
      "eval_steps_per_second": 0.218,
      "step": 42700
    },
    {
      "epoch": 9.988422406736055,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 42701
    },
    {
      "epoch": 9.988656297509063,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 42702
    },
    {
      "epoch": 9.988890188282072,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6627,
      "step": 42703
    },
    {
      "epoch": 9.989124079055081,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 42704
    },
    {
      "epoch": 9.98935796982809,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 42705
    },
    {
      "epoch": 9.989591860601099,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 42706
    },
    {
      "epoch": 9.989825751374108,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6772,
      "step": 42707
    },
    {
      "epoch": 9.990059642147116,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 42708
    },
    {
      "epoch": 9.990293532920127,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 42709
    },
    {
      "epoch": 9.990527423693136,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7294,
      "step": 42710
    },
    {
      "epoch": 9.990761314466145,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.5893,
      "step": 42711
    },
    {
      "epoch": 9.990995205239154,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 42712
    },
    {
      "epoch": 9.991229096012162,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 42713
    },
    {
      "epoch": 9.991462986785171,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 42714
    },
    {
      "epoch": 9.99169687755818,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.4043,
      "step": 42715
    },
    {
      "epoch": 9.991930768331189,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 42716
    },
    {
      "epoch": 9.992164659104198,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.6766,
      "step": 42717
    },
    {
      "epoch": 9.992398549877207,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 42718
    },
    {
      "epoch": 9.992632440650217,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.5948,
      "step": 42719
    },
    {
      "epoch": 9.992866331423226,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.653,
      "step": 42720
    },
    {
      "epoch": 9.993100222196235,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 42721
    },
    {
      "epoch": 9.993334112969244,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 42722
    },
    {
      "epoch": 9.993568003742253,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 42723
    },
    {
      "epoch": 9.993801894515261,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7401,
      "step": 42724
    },
    {
      "epoch": 9.99403578528827,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7566,
      "step": 42725
    },
    {
      "epoch": 9.994269676061279,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 42726
    },
    {
      "epoch": 9.994503566834288,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7798,
      "step": 42727
    },
    {
      "epoch": 9.994737457607297,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7826,
      "step": 42728
    },
    {
      "epoch": 9.994971348380307,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 42729
    },
    {
      "epoch": 9.995205239153316,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.2821,
      "step": 42730
    },
    {
      "epoch": 9.995439129926325,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 42731
    },
    {
      "epoch": 9.995673020699334,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7104,
      "step": 42732
    },
    {
      "epoch": 9.995906911472343,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8289,
      "step": 42733
    },
    {
      "epoch": 9.996140802245352,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 42734
    },
    {
      "epoch": 9.99637469301836,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8405,
      "step": 42735
    },
    {
      "epoch": 9.99660858379137,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 42736
    },
    {
      "epoch": 9.996842474564378,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7573,
      "step": 42737
    },
    {
      "epoch": 9.997076365337387,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.4318,
      "step": 42738
    },
    {
      "epoch": 9.997310256110396,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 42739
    },
    {
      "epoch": 9.997544146883406,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6168,
      "step": 42740
    },
    {
      "epoch": 9.997778037656415,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 42741
    },
    {
      "epoch": 9.998011928429424,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 42742
    },
    {
      "epoch": 9.998245819202433,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 42743
    },
    {
      "epoch": 9.998479709975442,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.5629,
      "step": 42744
    },
    {
      "epoch": 9.99871360074845,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7169,
      "step": 42745
    },
    {
      "epoch": 9.99894749152146,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6558,
      "step": 42746
    },
    {
      "epoch": 9.999181382294468,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.5726,
      "step": 42747
    },
    {
      "epoch": 9.999415273067477,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6509,
      "step": 42748
    },
    {
      "epoch": 9.999649163840486,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 42749
    },
    {
      "epoch": 9.999883054613495,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 42750
    }
  ],
  "logging_steps": 1,
  "max_steps": 42750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.993889859088941e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}