{
  "best_metric": 0.37876731157302856,
  "best_model_checkpoint": "miner_id_24/checkpoint-150",
  "epoch": 2.7777777777777777,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013888888888888888,
      "grad_norm": 18.353397369384766,
      "learning_rate": 1e-05,
      "loss": 4.3158,
      "step": 1
    },
    {
      "epoch": 0.013888888888888888,
      "eval_loss": 0.930080771446228,
      "eval_runtime": 16.5039,
      "eval_samples_per_second": 7.332,
      "eval_steps_per_second": 1.878,
      "step": 1
    },
    {
      "epoch": 0.027777777777777776,
      "grad_norm": 21.151273727416992,
      "learning_rate": 2e-05,
      "loss": 4.1819,
      "step": 2
    },
    {
      "epoch": 0.041666666666666664,
      "grad_norm": 17.57390785217285,
      "learning_rate": 3e-05,
      "loss": 4.0791,
      "step": 3
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 10.869729995727539,
      "learning_rate": 4e-05,
      "loss": 3.737,
      "step": 4
    },
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 9.29338550567627,
      "learning_rate": 5e-05,
      "loss": 3.0937,
      "step": 5
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 7.51112699508667,
      "learning_rate": 6e-05,
      "loss": 2.924,
      "step": 6
    },
    {
      "epoch": 0.09722222222222222,
      "grad_norm": 7.464629173278809,
      "learning_rate": 7e-05,
      "loss": 2.8015,
      "step": 7
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 6.8519182205200195,
      "learning_rate": 8e-05,
      "loss": 2.6424,
      "step": 8
    },
    {
      "epoch": 0.125,
      "grad_norm": 7.070345401763916,
      "learning_rate": 9e-05,
      "loss": 2.425,
      "step": 9
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 6.91495418548584,
      "learning_rate": 0.0001,
      "loss": 1.9977,
      "step": 10
    },
    {
      "epoch": 0.1527777777777778,
      "grad_norm": 6.541493892669678,
      "learning_rate": 9.999316524962345e-05,
      "loss": 1.6668,
      "step": 11
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 6.548363208770752,
      "learning_rate": 9.997266286704631e-05,
      "loss": 1.6336,
      "step": 12
    },
    {
      "epoch": 0.18055555555555555,
      "grad_norm": 7.693965911865234,
      "learning_rate": 9.993849845741524e-05,
      "loss": 1.5376,
      "step": 13
    },
    {
      "epoch": 0.19444444444444445,
      "grad_norm": 6.84789514541626,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.3121,
      "step": 14
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 7.3574113845825195,
      "learning_rate": 9.98292246503335e-05,
      "loss": 1.1839,
      "step": 15
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 5.898050785064697,
      "learning_rate": 9.975414512725057e-05,
      "loss": 1.2359,
      "step": 16
    },
    {
      "epoch": 0.2361111111111111,
      "grad_norm": 6.1749091148376465,
      "learning_rate": 9.966546331768191e-05,
      "loss": 0.9733,
      "step": 17
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.932611465454102,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.0949,
      "step": 18
    },
    {
      "epoch": 0.2638888888888889,
      "grad_norm": 5.398266315460205,
      "learning_rate": 9.944739353007344e-05,
      "loss": 2.8265,
      "step": 19
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 4.747178077697754,
      "learning_rate": 9.931806517013612e-05,
      "loss": 2.5941,
      "step": 20
    },
    {
      "epoch": 0.2916666666666667,
      "grad_norm": 4.58260440826416,
      "learning_rate": 9.917525374361912e-05,
      "loss": 2.6221,
      "step": 21
    },
    {
      "epoch": 0.3055555555555556,
      "grad_norm": 4.393043041229248,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.1772,
      "step": 22
    },
    {
      "epoch": 0.3194444444444444,
      "grad_norm": 4.183957576751709,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.3313,
      "step": 23
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 4.469913005828857,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.1241,
      "step": 24
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 4.208711624145508,
      "learning_rate": 9.847001329696653e-05,
      "loss": 1.9774,
      "step": 25
    },
    {
      "epoch": 0.3611111111111111,
      "grad_norm": 4.95603084564209,
      "learning_rate": 9.826044551386744e-05,
      "loss": 2.1882,
      "step": 26
    },
    {
      "epoch": 0.375,
      "grad_norm": 5.073214054107666,
      "learning_rate": 9.803768380684242e-05,
      "loss": 1.9081,
      "step": 27
    },
    {
      "epoch": 0.3888888888888889,
      "grad_norm": 4.388782978057861,
      "learning_rate": 9.780178907671789e-05,
      "loss": 1.5653,
      "step": 28
    },
    {
      "epoch": 0.4027777777777778,
      "grad_norm": 3.972588300704956,
      "learning_rate": 9.755282581475769e-05,
      "loss": 1.4472,
      "step": 29
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 4.375435829162598,
      "learning_rate": 9.729086208503174e-05,
      "loss": 1.3192,
      "step": 30
    },
    {
      "epoch": 0.4305555555555556,
      "grad_norm": 4.261547088623047,
      "learning_rate": 9.701596950580806e-05,
      "loss": 1.3463,
      "step": 31
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 3.9961156845092773,
      "learning_rate": 9.672822322997305e-05,
      "loss": 1.2567,
      "step": 32
    },
    {
      "epoch": 0.4583333333333333,
      "grad_norm": 4.565048694610596,
      "learning_rate": 9.642770192448536e-05,
      "loss": 1.0793,
      "step": 33
    },
    {
      "epoch": 0.4722222222222222,
      "grad_norm": 3.7915241718292236,
      "learning_rate": 9.611448774886924e-05,
      "loss": 0.7052,
      "step": 34
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 5.736730098724365,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.9681,
      "step": 35
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.29617977142334,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.6169,
      "step": 36
    },
    {
      "epoch": 0.5138888888888888,
      "grad_norm": 3.694272994995117,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.4885,
      "step": 37
    },
    {
      "epoch": 0.5277777777777778,
      "grad_norm": 3.531348466873169,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.5006,
      "step": 38
    },
    {
      "epoch": 0.5416666666666666,
      "grad_norm": 3.6069722175598145,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.2419,
      "step": 39
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 3.6983962059020996,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.2134,
      "step": 40
    },
    {
      "epoch": 0.5694444444444444,
      "grad_norm": 3.409634828567505,
      "learning_rate": 9.357421218136386e-05,
      "loss": 1.9794,
      "step": 41
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 3.5109164714813232,
      "learning_rate": 9.316282404787871e-05,
      "loss": 1.9957,
      "step": 42
    },
    {
      "epoch": 0.5972222222222222,
      "grad_norm": 3.4044387340545654,
      "learning_rate": 9.273963562927695e-05,
      "loss": 1.7358,
      "step": 43
    },
    {
      "epoch": 0.6111111111111112,
      "grad_norm": 3.9102120399475098,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.6241,
      "step": 44
    },
    {
      "epoch": 0.625,
      "grad_norm": 4.303212642669678,
      "learning_rate": 9.185832391312644e-05,
      "loss": 1.7366,
      "step": 45
    },
    {
      "epoch": 0.6388888888888888,
      "grad_norm": 4.366970062255859,
      "learning_rate": 9.140044155740101e-05,
      "loss": 1.4842,
      "step": 46
    },
    {
      "epoch": 0.6527777777777778,
      "grad_norm": 3.8644845485687256,
      "learning_rate": 9.093124073433463e-05,
      "loss": 1.2668,
      "step": 47
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 3.6025264263153076,
      "learning_rate": 9.045084971874738e-05,
      "loss": 1.2008,
      "step": 48
    },
    {
      "epoch": 0.6805555555555556,
      "grad_norm": 3.6874611377716064,
      "learning_rate": 8.995939984474624e-05,
      "loss": 1.137,
      "step": 49
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 3.7785308361053467,
      "learning_rate": 8.945702546981969e-05,
      "loss": 1.0808,
      "step": 50
    },
    {
      "epoch": 0.6944444444444444,
      "eval_loss": 0.4155964255332947,
      "eval_runtime": 16.5584,
      "eval_samples_per_second": 7.307,
      "eval_steps_per_second": 1.872,
      "step": 50
    },
    {
      "epoch": 0.7083333333333334,
      "grad_norm": 3.2879114151000977,
      "learning_rate": 8.894386393810563e-05,
      "loss": 0.6805,
      "step": 51
    },
    {
      "epoch": 0.7222222222222222,
      "grad_norm": 3.89845871925354,
      "learning_rate": 8.842005554284296e-05,
      "loss": 2.6619,
      "step": 52
    },
    {
      "epoch": 0.7361111111111112,
      "grad_norm": 3.4884915351867676,
      "learning_rate": 8.788574348801675e-05,
      "loss": 2.5877,
      "step": 53
    },
    {
      "epoch": 0.75,
      "grad_norm": 3.1210224628448486,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.1761,
      "step": 54
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 3.198755979537964,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.1815,
      "step": 55
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 3.187312126159668,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.2522,
      "step": 56
    },
    {
      "epoch": 0.7916666666666666,
      "grad_norm": 3.263937473297119,
      "learning_rate": 8.564642241456986e-05,
      "loss": 1.9056,
      "step": 57
    },
    {
      "epoch": 0.8055555555555556,
      "grad_norm": 3.321568489074707,
      "learning_rate": 8.506183921362443e-05,
      "loss": 1.8593,
      "step": 58
    },
    {
      "epoch": 0.8194444444444444,
      "grad_norm": 3.1380839347839355,
      "learning_rate": 8.44676704559283e-05,
      "loss": 1.6378,
      "step": 59
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 3.390587568283081,
      "learning_rate": 8.386407858128706e-05,
      "loss": 1.5882,
      "step": 60
    },
    {
      "epoch": 0.8472222222222222,
      "grad_norm": 3.768317937850952,
      "learning_rate": 8.32512286056924e-05,
      "loss": 1.5755,
      "step": 61
    },
    {
      "epoch": 0.8611111111111112,
      "grad_norm": 3.800569772720337,
      "learning_rate": 8.262928807620843e-05,
      "loss": 1.4202,
      "step": 62
    },
    {
      "epoch": 0.875,
      "grad_norm": 4.013070106506348,
      "learning_rate": 8.199842702516583e-05,
      "loss": 1.261,
      "step": 63
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 3.542781114578247,
      "learning_rate": 8.135881792367686e-05,
      "loss": 1.1586,
      "step": 64
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 3.6072185039520264,
      "learning_rate": 8.07106356344834e-05,
      "loss": 1.1252,
      "step": 65
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 4.028979301452637,
      "learning_rate": 8.005405736415126e-05,
      "loss": 1.0383,
      "step": 66
    },
    {
      "epoch": 0.9305555555555556,
      "grad_norm": 3.616936445236206,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.9714,
      "step": 67
    },
    {
      "epoch": 0.9444444444444444,
      "grad_norm": 3.4509968757629395,
      "learning_rate": 7.871643313414718e-05,
      "loss": 0.6968,
      "step": 68
    },
    {
      "epoch": 0.9583333333333334,
      "grad_norm": 3.8734400272369385,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.4973,
      "step": 69
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 3.203381299972534,
      "learning_rate": 7.734740790612136e-05,
      "loss": 1.6405,
      "step": 70
    },
    {
      "epoch": 0.9861111111111112,
      "grad_norm": 3.455230712890625,
      "learning_rate": 7.66515864363997e-05,
      "loss": 1.3994,
      "step": 71
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.295337677001953,
      "learning_rate": 7.594847868906076e-05,
      "loss": 0.9192,
      "step": 72
    },
    {
      "epoch": 1.0138888888888888,
      "grad_norm": 2.5301055908203125,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.112,
      "step": 73
    },
    {
      "epoch": 1.0277777777777777,
      "grad_norm": 2.659789800643921,
      "learning_rate": 7.452117519152542e-05,
      "loss": 1.8055,
      "step": 74
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 2.5605313777923584,
      "learning_rate": 7.379736965185368e-05,
      "loss": 1.7834,
      "step": 75
    },
    {
      "epoch": 1.0555555555555556,
      "grad_norm": 2.7616000175476074,
      "learning_rate": 7.30670581489344e-05,
      "loss": 1.7566,
      "step": 76
    },
    {
      "epoch": 1.0694444444444444,
      "grad_norm": 2.7995312213897705,
      "learning_rate": 7.233044034264034e-05,
      "loss": 1.667,
      "step": 77
    },
    {
      "epoch": 1.0833333333333333,
      "grad_norm": 2.6136202812194824,
      "learning_rate": 7.158771761692464e-05,
      "loss": 1.3881,
      "step": 78
    },
    {
      "epoch": 1.0972222222222223,
      "grad_norm": 2.9180049896240234,
      "learning_rate": 7.083909302476453e-05,
      "loss": 1.47,
      "step": 79
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 3.2744598388671875,
      "learning_rate": 7.008477123264848e-05,
      "loss": 1.3704,
      "step": 80
    },
    {
      "epoch": 1.125,
      "grad_norm": 3.1769022941589355,
      "learning_rate": 6.932495846462261e-05,
      "loss": 1.1543,
      "step": 81
    },
    {
      "epoch": 1.1388888888888888,
      "grad_norm": 3.250326633453369,
      "learning_rate": 6.855986244591104e-05,
      "loss": 1.1209,
      "step": 82
    },
    {
      "epoch": 1.1527777777777777,
      "grad_norm": 3.0372560024261475,
      "learning_rate": 6.778969234612584e-05,
      "loss": 0.8542,
      "step": 83
    },
    {
      "epoch": 1.1666666666666667,
      "grad_norm": 3.109739303588867,
      "learning_rate": 6.701465872208216e-05,
      "loss": 0.7596,
      "step": 84
    },
    {
      "epoch": 1.1805555555555556,
      "grad_norm": 3.16326904296875,
      "learning_rate": 6.623497346023418e-05,
      "loss": 0.6677,
      "step": 85
    },
    {
      "epoch": 1.1944444444444444,
      "grad_norm": 2.8650996685028076,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.5838,
      "step": 86
    },
    {
      "epoch": 1.2083333333333333,
      "grad_norm": 3.3142828941345215,
      "learning_rate": 6.466250186922325e-05,
      "loss": 0.5945,
      "step": 87
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 6.8858795166015625,
      "learning_rate": 6.387014543809223e-05,
      "loss": 0.4856,
      "step": 88
    },
    {
      "epoch": 1.2361111111111112,
      "grad_norm": 2.307889223098755,
      "learning_rate": 6.307399704769099e-05,
      "loss": 0.269,
      "step": 89
    },
    {
      "epoch": 1.25,
      "grad_norm": 4.386591911315918,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.2723,
      "step": 90
    },
    {
      "epoch": 1.2638888888888888,
      "grad_norm": 3.985119342803955,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.2554,
      "step": 91
    },
    {
      "epoch": 1.2777777777777777,
      "grad_norm": 3.5602235794067383,
      "learning_rate": 6.066498153718735e-05,
      "loss": 1.9099,
      "step": 92
    },
    {
      "epoch": 1.2916666666666667,
      "grad_norm": 3.168408155441284,
      "learning_rate": 5.985585137257401e-05,
      "loss": 1.787,
      "step": 93
    },
    {
      "epoch": 1.3055555555555556,
      "grad_norm": 2.7823638916015625,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.4852,
      "step": 94
    },
    {
      "epoch": 1.3194444444444444,
      "grad_norm": 3.0199127197265625,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 1.4295,
      "step": 95
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 2.5430126190185547,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.2113,
      "step": 96
    },
    {
      "epoch": 1.3472222222222223,
      "grad_norm": 2.5754640102386475,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.2303,
      "step": 97
    },
    {
      "epoch": 1.3611111111111112,
      "grad_norm": 2.9442992210388184,
      "learning_rate": 5.577423184847932e-05,
      "loss": 1.2014,
      "step": 98
    },
    {
      "epoch": 1.375,
      "grad_norm": 2.7263951301574707,
      "learning_rate": 5.495227651252315e-05,
      "loss": 0.9306,
      "step": 99
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 2.8367345333099365,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 0.9047,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "eval_loss": 0.38123124837875366,
      "eval_runtime": 16.4758,
      "eval_samples_per_second": 7.344,
      "eval_steps_per_second": 1.882,
      "step": 100
    },
    {
      "epoch": 1.4027777777777777,
      "grad_norm": 2.6654765605926514,
      "learning_rate": 5.330452921628497e-05,
      "loss": 0.7011,
      "step": 101
    },
    {
      "epoch": 1.4166666666666667,
      "grad_norm": 2.9699957370758057,
      "learning_rate": 5.247918773366112e-05,
      "loss": 0.707,
      "step": 102
    },
    {
      "epoch": 1.4305555555555556,
      "grad_norm": 3.1632471084594727,
      "learning_rate": 5.165316846586541e-05,
      "loss": 0.6542,
      "step": 103
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 3.069889783859253,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 0.533,
      "step": 104
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 3.1983110904693604,
      "learning_rate": 5e-05,
      "loss": 0.4616,
      "step": 105
    },
    {
      "epoch": 1.4722222222222223,
      "grad_norm": 2.486084222793579,
      "learning_rate": 4.917330276168208e-05,
      "loss": 0.3338,
      "step": 106
    },
    {
      "epoch": 1.4861111111111112,
      "grad_norm": 3.872039794921875,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.0558,
      "step": 107
    },
    {
      "epoch": 1.5,
      "grad_norm": 3.8578603267669678,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.0648,
      "step": 108
    },
    {
      "epoch": 1.5138888888888888,
      "grad_norm": 3.792673349380493,
      "learning_rate": 4.669547078371504e-05,
      "loss": 1.8315,
      "step": 109
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 3.4060540199279785,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 1.6477,
      "step": 110
    },
    {
      "epoch": 1.5416666666666665,
      "grad_norm": 3.1272222995758057,
      "learning_rate": 4.504772348747687e-05,
      "loss": 1.5035,
      "step": 111
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 2.9216132164001465,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 1.327,
      "step": 112
    },
    {
      "epoch": 1.5694444444444444,
      "grad_norm": 2.967139482498169,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 1.4411,
      "step": 113
    },
    {
      "epoch": 1.5833333333333335,
      "grad_norm": 2.8704493045806885,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 1.3121,
      "step": 114
    },
    {
      "epoch": 1.5972222222222223,
      "grad_norm": 2.790531873703003,
      "learning_rate": 4.17702704859633e-05,
      "loss": 1.1079,
      "step": 115
    },
    {
      "epoch": 1.6111111111111112,
      "grad_norm": 2.704164981842041,
      "learning_rate": 4.095597328339452e-05,
      "loss": 0.9159,
      "step": 116
    },
    {
      "epoch": 1.625,
      "grad_norm": 2.9431660175323486,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 0.8588,
      "step": 117
    },
    {
      "epoch": 1.6388888888888888,
      "grad_norm": 2.6618356704711914,
      "learning_rate": 3.933501846281267e-05,
      "loss": 0.731,
      "step": 118
    },
    {
      "epoch": 1.6527777777777777,
      "grad_norm": 2.4429187774658203,
      "learning_rate": 3.852880399766243e-05,
      "loss": 0.6871,
      "step": 119
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 2.6477255821228027,
      "learning_rate": 3.772572564296005e-05,
      "loss": 0.6108,
      "step": 120
    },
    {
      "epoch": 1.6805555555555556,
      "grad_norm": 2.6095504760742188,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 0.5207,
      "step": 121
    },
    {
      "epoch": 1.6944444444444444,
      "grad_norm": 2.9425339698791504,
      "learning_rate": 3.612985456190778e-05,
      "loss": 0.4459,
      "step": 122
    },
    {
      "epoch": 1.7083333333333335,
      "grad_norm": 2.136305332183838,
      "learning_rate": 3.533749813077677e-05,
      "loss": 0.2755,
      "step": 123
    },
    {
      "epoch": 1.7222222222222223,
      "grad_norm": 3.0910561084747314,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.2408,
      "step": 124
    },
    {
      "epoch": 1.7361111111111112,
      "grad_norm": 3.32804536819458,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.03,
      "step": 125
    },
    {
      "epoch": 1.75,
      "grad_norm": 3.0867159366607666,
      "learning_rate": 3.298534127791785e-05,
      "loss": 1.7317,
      "step": 126
    },
    {
      "epoch": 1.7638888888888888,
      "grad_norm": 3.285167932510376,
      "learning_rate": 3.221030765387417e-05,
      "loss": 1.6922,
      "step": 127
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 3.1600706577301025,
      "learning_rate": 3.144013755408895e-05,
      "loss": 1.5817,
      "step": 128
    },
    {
      "epoch": 1.7916666666666665,
      "grad_norm": 3.2354483604431152,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 1.4781,
      "step": 129
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 2.9639623165130615,
      "learning_rate": 2.991522876735154e-05,
      "loss": 1.2128,
      "step": 130
    },
    {
      "epoch": 1.8194444444444444,
      "grad_norm": 2.8074262142181396,
      "learning_rate": 2.916090697523549e-05,
      "loss": 1.1838,
      "step": 131
    },
    {
      "epoch": 1.8333333333333335,
      "grad_norm": 2.8644473552703857,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 1.1049,
      "step": 132
    },
    {
      "epoch": 1.8472222222222223,
      "grad_norm": 3.0072569847106934,
      "learning_rate": 2.766955965735968e-05,
      "loss": 0.9963,
      "step": 133
    },
    {
      "epoch": 1.8611111111111112,
      "grad_norm": 2.9529449939727783,
      "learning_rate": 2.693294185106562e-05,
      "loss": 0.8727,
      "step": 134
    },
    {
      "epoch": 1.875,
      "grad_norm": 2.843312978744507,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 0.6344,
      "step": 135
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 2.7216567993164062,
      "learning_rate": 2.547882480847461e-05,
      "loss": 0.6347,
      "step": 136
    },
    {
      "epoch": 1.9027777777777777,
      "grad_norm": 2.7187931537628174,
      "learning_rate": 2.476172311325783e-05,
      "loss": 0.5701,
      "step": 137
    },
    {
      "epoch": 1.9166666666666665,
      "grad_norm": 2.4936585426330566,
      "learning_rate": 2.405152131093926e-05,
      "loss": 0.4658,
      "step": 138
    },
    {
      "epoch": 1.9305555555555556,
      "grad_norm": 2.5849828720092773,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 0.4262,
      "step": 139
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 2.010850191116333,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 0.2889,
      "step": 140
    },
    {
      "epoch": 1.9583333333333335,
      "grad_norm": 2.7142035961151123,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.8619,
      "step": 141
    },
    {
      "epoch": 1.9722222222222223,
      "grad_norm": 2.805471181869507,
      "learning_rate": 2.128356686585282e-05,
      "loss": 1.2607,
      "step": 142
    },
    {
      "epoch": 1.9861111111111112,
      "grad_norm": 2.920461416244507,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.8266,
      "step": 143
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.9211695194244385,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 0.3794,
      "step": 144
    },
    {
      "epoch": 2.013888888888889,
      "grad_norm": 2.189929246902466,
      "learning_rate": 1.928936436551661e-05,
      "loss": 1.5932,
      "step": 145
    },
    {
      "epoch": 2.0277777777777777,
      "grad_norm": 2.3327298164367676,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.5157,
      "step": 146
    },
    {
      "epoch": 2.0416666666666665,
      "grad_norm": 2.3047962188720703,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.3032,
      "step": 147
    },
    {
      "epoch": 2.0555555555555554,
      "grad_norm": 2.399838924407959,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.3442,
      "step": 148
    },
    {
      "epoch": 2.0694444444444446,
      "grad_norm": 2.26955509185791,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 1.0195,
      "step": 149
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 2.290834903717041,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 0.9126,
      "step": 150
    },
    {
      "epoch": 2.0833333333333335,
      "eval_loss": 0.37876731157302856,
      "eval_runtime": 16.7357,
      "eval_samples_per_second": 7.23,
      "eval_steps_per_second": 1.852,
      "step": 150
    },
    {
      "epoch": 2.0972222222222223,
      "grad_norm": 2.2856011390686035,
      "learning_rate": 1.553232954407171e-05,
      "loss": 0.8388,
      "step": 151
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 2.38346791267395,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 0.7963,
      "step": 152
    },
    {
      "epoch": 2.125,
      "grad_norm": 2.418088674545288,
      "learning_rate": 1.435357758543015e-05,
      "loss": 0.7272,
      "step": 153
    },
    {
      "epoch": 2.138888888888889,
      "grad_norm": 2.605368137359619,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 0.6916,
      "step": 154
    },
    {
      "epoch": 2.1527777777777777,
      "grad_norm": 2.2164158821105957,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 0.3841,
      "step": 155
    },
    {
      "epoch": 2.1666666666666665,
      "grad_norm": 2.296902894973755,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 0.3451,
      "step": 156
    },
    {
      "epoch": 2.1805555555555554,
      "grad_norm": 2.050575017929077,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 0.3102,
      "step": 157
    },
    {
      "epoch": 2.1944444444444446,
      "grad_norm": 2.33125901222229,
      "learning_rate": 1.157994445715706e-05,
      "loss": 0.3001,
      "step": 158
    },
    {
      "epoch": 2.2083333333333335,
      "grad_norm": 2.4090118408203125,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 0.2666,
      "step": 159
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 1.8565489053726196,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 0.1895,
      "step": 160
    },
    {
      "epoch": 2.236111111111111,
      "grad_norm": 1.2469310760498047,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 0.0927,
      "step": 161
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.9802675247192383,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.6083,
      "step": 162
    },
    {
      "epoch": 2.263888888888889,
      "grad_norm": 2.974001884460449,
      "learning_rate": 9.068759265665384e-06,
      "loss": 1.444,
      "step": 163
    },
    {
      "epoch": 2.2777777777777777,
      "grad_norm": 2.8431155681610107,
      "learning_rate": 8.599558442598998e-06,
      "loss": 1.3003,
      "step": 164
    },
    {
      "epoch": 2.2916666666666665,
      "grad_norm": 2.768345355987549,
      "learning_rate": 8.141676086873572e-06,
      "loss": 1.1234,
      "step": 165
    },
    {
      "epoch": 2.3055555555555554,
      "grad_norm": 2.8433845043182373,
      "learning_rate": 7.695237378953223e-06,
      "loss": 1.0373,
      "step": 166
    },
    {
      "epoch": 2.3194444444444446,
      "grad_norm": 2.789776563644409,
      "learning_rate": 7.260364370723044e-06,
      "loss": 0.9523,
      "step": 167
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 2.6452817916870117,
      "learning_rate": 6.837175952121306e-06,
      "loss": 0.788,
      "step": 168
    },
    {
      "epoch": 2.3472222222222223,
      "grad_norm": 2.6535677909851074,
      "learning_rate": 6.425787818636131e-06,
      "loss": 0.7767,
      "step": 169
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 2.7237205505371094,
      "learning_rate": 6.026312439675552e-06,
      "loss": 0.6581,
      "step": 170
    },
    {
      "epoch": 2.375,
      "grad_norm": 2.4387779235839844,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 0.5044,
      "step": 171
    },
    {
      "epoch": 2.388888888888889,
      "grad_norm": 2.4389240741729736,
      "learning_rate": 5.263533508961827e-06,
      "loss": 0.4171,
      "step": 172
    },
    {
      "epoch": 2.4027777777777777,
      "grad_norm": 2.2974355220794678,
      "learning_rate": 4.900438493352055e-06,
      "loss": 0.3782,
      "step": 173
    },
    {
      "epoch": 2.4166666666666665,
      "grad_norm": 2.248307943344116,
      "learning_rate": 4.549673247541875e-06,
      "loss": 0.3273,
      "step": 174
    },
    {
      "epoch": 2.4305555555555554,
      "grad_norm": 2.141526222229004,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 0.2397,
      "step": 175
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 2.2435672283172607,
      "learning_rate": 3.885512251130763e-06,
      "loss": 0.2287,
      "step": 176
    },
    {
      "epoch": 2.4583333333333335,
      "grad_norm": 2.3485684394836426,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 0.2137,
      "step": 177
    },
    {
      "epoch": 2.4722222222222223,
      "grad_norm": 1.5031319856643677,
      "learning_rate": 3.271776770026963e-06,
      "loss": 0.1144,
      "step": 178
    },
    {
      "epoch": 2.486111111111111,
      "grad_norm": 2.6786105632781982,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 1.7655,
      "step": 179
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.7928459644317627,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 1.4685,
      "step": 180
    },
    {
      "epoch": 2.513888888888889,
      "grad_norm": 2.703256130218506,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.2011,
      "step": 181
    },
    {
      "epoch": 2.5277777777777777,
      "grad_norm": 2.643092155456543,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.1017,
      "step": 182
    },
    {
      "epoch": 2.5416666666666665,
      "grad_norm": 2.6408233642578125,
      "learning_rate": 1.962316193157593e-06,
      "loss": 1.0977,
      "step": 183
    },
    {
      "epoch": 2.5555555555555554,
      "grad_norm": 2.7568745613098145,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 0.9662,
      "step": 184
    },
    {
      "epoch": 2.5694444444444446,
      "grad_norm": 2.6440587043762207,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 0.8838,
      "step": 185
    },
    {
      "epoch": 2.5833333333333335,
      "grad_norm": 2.6023125648498535,
      "learning_rate": 1.333670137599713e-06,
      "loss": 0.7707,
      "step": 186
    },
    {
      "epoch": 2.5972222222222223,
      "grad_norm": 2.5014240741729736,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 0.7291,
      "step": 187
    },
    {
      "epoch": 2.611111111111111,
      "grad_norm": 2.311945915222168,
      "learning_rate": 9.810017062595322e-07,
      "loss": 0.5362,
      "step": 188
    },
    {
      "epoch": 2.625,
      "grad_norm": 2.521380662918091,
      "learning_rate": 8.247462563808817e-07,
      "loss": 0.4541,
      "step": 189
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 2.2873525619506836,
      "learning_rate": 6.819348298638839e-07,
      "loss": 0.3515,
      "step": 190
    },
    {
      "epoch": 2.6527777777777777,
      "grad_norm": 2.233489751815796,
      "learning_rate": 5.526064699265753e-07,
      "loss": 0.3229,
      "step": 191
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 2.0663254261016846,
      "learning_rate": 4.367965336512403e-07,
      "loss": 0.2556,
      "step": 192
    },
    {
      "epoch": 2.6805555555555554,
      "grad_norm": 2.246070146560669,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 0.1935,
      "step": 193
    },
    {
      "epoch": 2.6944444444444446,
      "grad_norm": 1.6796926259994507,
      "learning_rate": 2.458548727494292e-07,
      "loss": 0.1674,
      "step": 194
    },
    {
      "epoch": 2.7083333333333335,
      "grad_norm": 1.2821648120880127,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 0.0973,
      "step": 195
    },
    {
      "epoch": 2.7222222222222223,
      "grad_norm": 2.6151907444000244,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.532,
      "step": 196
    },
    {
      "epoch": 2.736111111111111,
      "grad_norm": 2.7919063568115234,
      "learning_rate": 6.150154258476315e-08,
      "loss": 1.3863,
      "step": 197
    },
    {
      "epoch": 2.75,
      "grad_norm": 2.7798070907592773,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.2623,
      "step": 198
    },
    {
      "epoch": 2.763888888888889,
      "grad_norm": 2.6710925102233887,
      "learning_rate": 6.834750376549792e-09,
      "loss": 1.1619,
      "step": 199
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 2.729437828063965,
      "learning_rate": 0.0,
      "loss": 0.995,
      "step": 200
    },
    {
      "epoch": 2.7777777777777777,
      "eval_loss": 0.3891352117061615,
      "eval_runtime": 17.0476,
      "eval_samples_per_second": 7.098,
      "eval_steps_per_second": 1.818,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 1
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.00223523021652e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}