diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5773 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 0,
+  "global_step": 820,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0012195121951219512,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.987804878048782e-06,
+      "loss": 1.7359,
+      "step": 1
+    },
+    {
+      "epoch": 0.0024390243902439024,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.975609756097561e-06,
+      "loss": 1.7645,
+      "step": 2
+    },
+    {
+      "epoch": 0.003658536585365854,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.963414634146342e-06,
+      "loss": 1.7766,
+      "step": 3
+    },
+    {
+      "epoch": 0.004878048780487805,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.951219512195124e-06,
+      "loss": 1.7373,
+      "step": 4
+    },
+    {
+      "epoch": 0.006097560975609756,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.939024390243903e-06,
+      "loss": 1.7009,
+      "step": 5
+    },
+    {
+      "epoch": 0.007317073170731708,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.926829268292684e-06,
+      "loss": 1.6796,
+      "step": 6
+    },
+    {
+      "epoch": 0.00853658536585366,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.914634146341463e-06,
+      "loss": 1.7517,
+      "step": 7
+    },
+    {
+      "epoch": 0.00975609756097561,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.902439024390245e-06,
+      "loss": 1.7498,
+      "step": 8
+    },
+    {
+      "epoch": 0.01097560975609756,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.890243902439026e-06,
+      "loss": 1.6965,
+      "step": 9
+    },
+    {
+      "epoch": 0.012195121951219513,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.878048780487805e-06,
+      "loss": 1.6836,
+      "step": 10
+    },
+    {
+      "epoch": 0.013414634146341463,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.865853658536586e-06,
+      "loss": 1.7479,
+      "step": 11
+    },
+    {
+      "epoch": 0.014634146341463415,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.853658536585367e-06,
+      "loss": 1.6494,
+      "step": 12
+    },
+    {
+      "epoch": 0.015853658536585366,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.841463414634147e-06,
+      "loss": 1.6509,
+      "step": 13
+    },
+    {
+      "epoch": 0.01707317073170732,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.829268292682928e-06,
+      "loss": 1.6927,
+      "step": 14
+    },
+    {
+      "epoch": 0.018292682926829267,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.817073170731707e-06,
+      "loss": 1.6383,
+      "step": 15
+    },
+    {
+      "epoch": 0.01951219512195122,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.804878048780488e-06,
+      "loss": 1.5846,
+      "step": 16
+    },
+    {
+      "epoch": 0.020731707317073172,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.79268292682927e-06,
+      "loss": 1.624,
+      "step": 17
+    },
+    {
+      "epoch": 0.02195121951219512,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.780487804878049e-06,
+      "loss": 1.5879,
+      "step": 18
+    },
+    {
+      "epoch": 0.023170731707317073,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.76829268292683e-06,
+      "loss": 1.6408,
+      "step": 19
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 1.6303,
+      "step": 20
+    },
+    {
+      "epoch": 0.025609756097560974,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.74390243902439e-06,
+      "loss": 1.6591,
+      "step": 21
+    },
+    {
+      "epoch": 0.026829268292682926,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.731707317073171e-06,
+      "loss": 1.5865,
+      "step": 22
+    },
+    {
+      "epoch": 0.02804878048780488,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.719512195121953e-06,
+      "loss": 1.5393,
+      "step": 23
+    },
+    {
+      "epoch": 0.02926829268292683,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.707317073170732e-06,
+      "loss": 1.6239,
+      "step": 24
+    },
+    {
+      "epoch": 0.03048780487804878,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.695121951219513e-06,
+      "loss": 1.541,
+      "step": 25
+    },
+    {
+      "epoch": 0.03170731707317073,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.682926829268292e-06,
+      "loss": 1.52,
+      "step": 26
+    },
+    {
+      "epoch": 0.032926829268292684,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.670731707317074e-06,
+      "loss": 1.6195,
+      "step": 27
+    },
+    {
+      "epoch": 0.03414634146341464,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.658536585365855e-06,
+      "loss": 1.4989,
+      "step": 28
+    },
+    {
+      "epoch": 0.03536585365853658,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.646341463414634e-06,
+      "loss": 1.5094,
+      "step": 29
+    },
+    {
+      "epoch": 0.036585365853658534,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.634146341463415e-06,
+      "loss": 1.5214,
+      "step": 30
+    },
+    {
+      "epoch": 0.03780487804878049,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.621951219512196e-06,
+      "loss": 1.5122,
+      "step": 31
+    },
+    {
+      "epoch": 0.03902439024390244,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.609756097560976e-06,
+      "loss": 1.5766,
+      "step": 32
+    },
+    {
+      "epoch": 0.04024390243902439,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.597560975609757e-06,
+      "loss": 1.527,
+      "step": 33
+    },
+    {
+      "epoch": 0.041463414634146344,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.585365853658536e-06,
+      "loss": 1.4714,
+      "step": 34
+    },
+    {
+      "epoch": 0.042682926829268296,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.573170731707317e-06,
+      "loss": 1.5223,
+      "step": 35
+    },
+    {
+      "epoch": 0.04390243902439024,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.560975609756098e-06,
+      "loss": 1.5577,
+      "step": 36
+    },
+    {
+      "epoch": 0.045121951219512194,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 9.548780487804878e-06,
+      "loss": 1.4361,
+      "step": 37
+    },
+    {
+      "epoch": 0.046341463414634146,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.536585365853659e-06,
+      "loss": 1.4433,
+      "step": 38
+    },
+    {
+      "epoch": 0.0475609756097561,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 9.52439024390244e-06,
+      "loss": 1.4276,
+      "step": 39
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 9.51219512195122e-06,
+      "loss": 1.467,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.5e-06,
+      "loss": 1.5153,
+      "step": 41
+    },
+    {
+      "epoch": 0.05121951219512195,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.487804878048782e-06,
+      "loss": 1.502,
+      "step": 42
+    },
+    {
+      "epoch": 0.0524390243902439,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.475609756097561e-06,
+      "loss": 1.4238,
+      "step": 43
+    },
+    {
+      "epoch": 0.05365853658536585,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.463414634146342e-06,
+      "loss": 1.4485,
+      "step": 44
+    },
+    {
+      "epoch": 0.054878048780487805,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.451219512195122e-06,
+      "loss": 1.4759,
+      "step": 45
+    },
+    {
+      "epoch": 0.05609756097560976,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 9.439024390243903e-06,
+      "loss": 1.3978,
+      "step": 46
+    },
+    {
+      "epoch": 0.05731707317073171,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 9.426829268292684e-06,
+      "loss": 1.4025,
+      "step": 47
+    },
+    {
+      "epoch": 0.05853658536585366,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.414634146341463e-06,
+      "loss": 1.3911,
+      "step": 48
+    },
+    {
+      "epoch": 0.05975609756097561,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.402439024390244e-06,
+      "loss": 1.3932,
+      "step": 49
+    },
+    {
+      "epoch": 0.06097560975609756,
+      "grad_norm": 0.150390625,
+      "learning_rate": 9.390243902439025e-06,
+      "loss": 1.4234,
+      "step": 50
+    },
+    {
+      "epoch": 0.06219512195121951,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.378048780487805e-06,
+      "loss": 1.4094,
+      "step": 51
+    },
+    {
+      "epoch": 0.06341463414634146,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.365853658536586e-06,
+      "loss": 1.3633,
+      "step": 52
+    },
+    {
+      "epoch": 0.06463414634146342,
+      "grad_norm": 0.14453125,
+      "learning_rate": 9.353658536585367e-06,
+      "loss": 1.3763,
+      "step": 53
+    },
+    {
+      "epoch": 0.06585365853658537,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 9.341463414634148e-06,
+      "loss": 1.3517,
+      "step": 54
+    },
+    {
+      "epoch": 0.06707317073170732,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.329268292682927e-06,
+      "loss": 1.3864,
+      "step": 55
+    },
+    {
+      "epoch": 0.06829268292682927,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.317073170731709e-06,
+      "loss": 1.4261,
+      "step": 56
+    },
+    {
+      "epoch": 0.06951219512195123,
+      "grad_norm": 0.14453125,
+      "learning_rate": 9.30487804878049e-06,
+      "loss": 1.411,
+      "step": 57
+    },
+    {
+      "epoch": 0.07073170731707316,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.292682926829269e-06,
+      "loss": 1.3049,
+      "step": 58
+    },
+    {
+      "epoch": 0.07195121951219512,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.28048780487805e-06,
+      "loss": 1.4078,
+      "step": 59
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 9.268292682926831e-06,
+      "loss": 1.4062,
+      "step": 60
+    },
+    {
+      "epoch": 0.07439024390243902,
+      "grad_norm": 0.150390625,
+      "learning_rate": 9.25609756097561e-06,
+      "loss": 1.3901,
+      "step": 61
+    },
+    {
+      "epoch": 0.07560975609756097,
+      "grad_norm": 0.134765625,
+      "learning_rate": 9.243902439024392e-06,
+      "loss": 1.3779,
+      "step": 62
+    },
+    {
+      "epoch": 0.07682926829268293,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 9.231707317073171e-06,
+      "loss": 1.3235,
+      "step": 63
+    },
+    {
+      "epoch": 0.07804878048780488,
+      "grad_norm": 0.1328125,
+      "learning_rate": 9.219512195121952e-06,
+      "loss": 1.3703,
+      "step": 64
+    },
+    {
+      "epoch": 0.07926829268292683,
+      "grad_norm": 0.13671875,
+      "learning_rate": 9.207317073170733e-06,
+      "loss": 1.3756,
+      "step": 65
+    },
+    {
+      "epoch": 0.08048780487804878,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 9.195121951219513e-06,
+      "loss": 1.3887,
+      "step": 66
+    },
+    {
+      "epoch": 0.08170731707317073,
+      "grad_norm": 0.150390625,
+      "learning_rate": 9.182926829268294e-06,
+      "loss": 1.348,
+      "step": 67
+    },
+    {
+      "epoch": 0.08292682926829269,
+      "grad_norm": 0.13671875,
+      "learning_rate": 9.170731707317075e-06,
+      "loss": 1.3403,
+      "step": 68
+    },
+    {
+      "epoch": 0.08414634146341464,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.158536585365854e-06,
+      "loss": 1.3239,
+      "step": 69
+    },
+    {
+      "epoch": 0.08536585365853659,
+      "grad_norm": 0.14453125,
+      "learning_rate": 9.146341463414635e-06,
+      "loss": 1.3273,
+      "step": 70
+    },
+    {
+      "epoch": 0.08658536585365853,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.134146341463415e-06,
+      "loss": 1.3784,
+      "step": 71
+    },
+    {
+      "epoch": 0.08780487804878048,
+      "grad_norm": 0.1328125,
+      "learning_rate": 9.121951219512196e-06,
+      "loss": 1.3063,
+      "step": 72
+    },
+    {
+      "epoch": 0.08902439024390243,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 9.109756097560977e-06,
+      "loss": 1.3335,
+      "step": 73
+    },
+    {
+      "epoch": 0.09024390243902439,
+      "grad_norm": 0.13671875,
+      "learning_rate": 9.097560975609757e-06,
+      "loss": 1.3193,
+      "step": 74
+    },
+    {
+      "epoch": 0.09146341463414634,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 9.085365853658538e-06,
+      "loss": 1.3003,
+      "step": 75
+    },
+    {
+      "epoch": 0.09268292682926829,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 9.073170731707319e-06,
+      "loss": 1.3387,
+      "step": 76
+    },
+    {
+      "epoch": 0.09390243902439024,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 9.060975609756098e-06,
+      "loss": 1.2503,
+      "step": 77
+    },
+    {
+      "epoch": 0.0951219512195122,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 9.04878048780488e-06,
+      "loss": 1.2954,
+      "step": 78
+    },
+    {
+      "epoch": 0.09634146341463415,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.03658536585366e-06,
+      "loss": 1.3674,
+      "step": 79
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.02439024390244e-06,
+      "loss": 1.3554,
+      "step": 80
+    },
+    {
+      "epoch": 0.09878048780487805,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 9.01219512195122e-06,
+      "loss": 1.3393,
+      "step": 81
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9e-06,
+      "loss": 1.2831,
+      "step": 82
+    },
+    {
+      "epoch": 0.10121951219512196,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.987804878048781e-06,
+      "loss": 1.3079,
+      "step": 83
+    },
+    {
+      "epoch": 0.1024390243902439,
+      "grad_norm": 0.1328125,
+      "learning_rate": 8.975609756097562e-06,
+      "loss": 1.3021,
+      "step": 84
+    },
+    {
+      "epoch": 0.10365853658536585,
+      "grad_norm": 0.142578125,
+      "learning_rate": 8.963414634146342e-06,
+      "loss": 1.2823,
+      "step": 85
+    },
+    {
+      "epoch": 0.1048780487804878,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.951219512195123e-06,
+      "loss": 1.2396,
+      "step": 86
+    },
+    {
+      "epoch": 0.10609756097560975,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.939024390243904e-06,
+      "loss": 1.3479,
+      "step": 87
+    },
+    {
+      "epoch": 0.1073170731707317,
+      "grad_norm": 0.14453125,
+      "learning_rate": 8.926829268292683e-06,
+      "loss": 1.2762,
+      "step": 88
+    },
+    {
+      "epoch": 0.10853658536585366,
+      "grad_norm": 0.14453125,
+      "learning_rate": 8.914634146341465e-06,
+      "loss": 1.3385,
+      "step": 89
+    },
+    {
+      "epoch": 0.10975609756097561,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 8.902439024390244e-06,
+      "loss": 1.3153,
+      "step": 90
+    },
+    {
+      "epoch": 0.11097560975609756,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.890243902439025e-06,
+      "loss": 1.2956,
+      "step": 91
+    },
+    {
+      "epoch": 0.11219512195121951,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.878048780487806e-06,
+      "loss": 1.3069,
+      "step": 92
+    },
+    {
+      "epoch": 0.11341463414634147,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.865853658536586e-06,
+      "loss": 1.2872,
+      "step": 93
+    },
+    {
+      "epoch": 0.11463414634146342,
+      "grad_norm": 0.13671875,
+      "learning_rate": 8.853658536585367e-06,
+      "loss": 1.3395,
+      "step": 94
+    },
+    {
+      "epoch": 0.11585365853658537,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.841463414634148e-06,
+      "loss": 1.2451,
+      "step": 95
+    },
+    {
+      "epoch": 0.11707317073170732,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 8.829268292682927e-06,
+      "loss": 1.2895,
+      "step": 96
+    },
+    {
+      "epoch": 0.11829268292682926,
+      "grad_norm": 0.15625,
+      "learning_rate": 8.817073170731708e-06,
+      "loss": 1.2988,
+      "step": 97
+    },
+    {
+      "epoch": 0.11951219512195121,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 8.80487804878049e-06,
+      "loss": 1.2686,
+      "step": 98
+    },
+    {
+      "epoch": 0.12073170731707317,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.792682926829269e-06,
+      "loss": 1.2651,
+      "step": 99
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 1.2891,
+      "step": 100
+    },
+    {
+      "epoch": 0.12317073170731707,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 8.76829268292683e-06,
+      "loss": 1.286,
+      "step": 101
+    },
+    {
+      "epoch": 0.12439024390243902,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.75609756097561e-06,
+      "loss": 1.2882,
+      "step": 102
+    },
+    {
+      "epoch": 0.12560975609756098,
+      "grad_norm": 0.1328125,
+      "learning_rate": 8.743902439024391e-06,
+      "loss": 1.177,
+      "step": 103
+    },
+    {
+      "epoch": 0.12682926829268293,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 8.731707317073171e-06,
+      "loss": 1.2681,
+      "step": 104
+    },
+    {
+      "epoch": 0.12804878048780488,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.719512195121952e-06,
+      "loss": 1.2534,
+      "step": 105
+    },
+    {
+      "epoch": 0.12926829268292683,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.707317073170733e-06,
+      "loss": 1.2651,
+      "step": 106
+    },
+    {
+      "epoch": 0.13048780487804879,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 8.695121951219512e-06,
+      "loss": 1.226,
+      "step": 107
+    },
+    {
+      "epoch": 0.13170731707317074,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 8.682926829268294e-06,
+      "loss": 1.253,
+      "step": 108
+    },
+    {
+      "epoch": 0.1329268292682927,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.670731707317073e-06,
+      "loss": 1.2314,
+      "step": 109
+    },
+    {
+      "epoch": 0.13414634146341464,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.658536585365854e-06,
+      "loss": 1.2127,
+      "step": 110
+    },
+    {
+      "epoch": 0.1353658536585366,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.646341463414635e-06,
+      "loss": 1.2699,
+      "step": 111
+    },
+    {
+      "epoch": 0.13658536585365855,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.634146341463415e-06,
+      "loss": 1.2047,
+      "step": 112
+    },
+    {
+      "epoch": 0.1378048780487805,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 8.621951219512196e-06,
+      "loss": 1.2516,
+      "step": 113
+    },
+    {
+      "epoch": 0.13902439024390245,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 8.609756097560977e-06,
+      "loss": 1.216,
+      "step": 114
+    },
+    {
+      "epoch": 0.1402439024390244,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 8.597560975609756e-06,
+      "loss": 1.2471,
+      "step": 115
+    },
+    {
+      "epoch": 0.14146341463414633,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.585365853658537e-06,
+      "loss": 1.289,
+      "step": 116
+    },
+    {
+      "epoch": 0.14268292682926828,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.573170731707318e-06,
+      "loss": 1.2521,
+      "step": 117
+    },
+    {
+      "epoch": 0.14390243902439023,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.560975609756098e-06,
+      "loss": 1.1712,
+      "step": 118
+    },
+    {
+      "epoch": 0.14512195121951219,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.548780487804879e-06,
+      "loss": 1.1608,
+      "step": 119
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 8.536585365853658e-06,
+      "loss": 1.201,
+      "step": 120
+    },
+    {
+      "epoch": 0.1475609756097561,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.52439024390244e-06,
+      "loss": 1.2613,
+      "step": 121
+    },
+    {
+      "epoch": 0.14878048780487804,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 8.51219512195122e-06,
+      "loss": 1.2407,
+      "step": 122
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.5e-06,
+      "loss": 1.1373,
+      "step": 123
+    },
+    {
+      "epoch": 0.15121951219512195,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.487804878048781e-06,
+      "loss": 1.233,
+      "step": 124
+    },
+    {
+      "epoch": 0.1524390243902439,
+      "grad_norm": 0.13671875,
+      "learning_rate": 8.475609756097562e-06,
+      "loss": 1.1991,
+      "step": 125
+    },
+    {
+      "epoch": 0.15365853658536585,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.463414634146342e-06,
+      "loss": 1.1861,
+      "step": 126
+    },
+    {
+      "epoch": 0.1548780487804878,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.451219512195123e-06,
+      "loss": 1.2282,
+      "step": 127
+    },
+    {
+      "epoch": 0.15609756097560976,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.439024390243902e-06,
+      "loss": 1.2937,
+      "step": 128
+    },
+    {
+      "epoch": 0.1573170731707317,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.426829268292683e-06,
+      "loss": 1.2342,
+      "step": 129
+    },
+    {
+      "epoch": 0.15853658536585366,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.414634146341464e-06,
+      "loss": 1.2159,
+      "step": 130
+    },
+    {
+      "epoch": 0.1597560975609756,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.402439024390244e-06,
+      "loss": 1.1729,
+      "step": 131
+    },
+    {
+      "epoch": 0.16097560975609757,
+      "grad_norm": 0.154296875,
+      "learning_rate": 8.390243902439025e-06,
+      "loss": 1.2317,
+      "step": 132
+    },
+    {
+      "epoch": 0.16219512195121952,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 8.378048780487806e-06,
+      "loss": 1.2474,
+      "step": 133
+    },
+    {
+      "epoch": 0.16341463414634147,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 8.365853658536585e-06,
+      "loss": 1.2129,
+      "step": 134
+    },
+    {
+      "epoch": 0.16463414634146342,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.353658536585366e-06,
+      "loss": 1.1949,
+      "step": 135
+    },
+    {
+      "epoch": 0.16585365853658537,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 8.341463414634147e-06,
+      "loss": 1.2223,
+      "step": 136
+    },
+    {
+      "epoch": 0.16707317073170733,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.329268292682927e-06,
+      "loss": 1.1877,
+      "step": 137
+    },
+    {
+      "epoch": 0.16829268292682928,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.317073170731708e-06,
+      "loss": 1.2069,
+      "step": 138
+    },
+    {
+      "epoch": 0.16951219512195123,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 8.304878048780487e-06,
+      "loss": 1.1599,
+      "step": 139
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 0.15234375,
+      "learning_rate": 8.292682926829268e-06,
+      "loss": 1.2284,
+      "step": 140
+    },
+    {
+      "epoch": 0.1719512195121951,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.28048780487805e-06,
+      "loss": 1.172,
+      "step": 141
+    },
+    {
+      "epoch": 0.17317073170731706,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 8.268292682926829e-06,
+      "loss": 1.1259,
+      "step": 142
+    },
+    {
+      "epoch": 0.174390243902439,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.25609756097561e-06,
+      "loss": 1.2122,
+      "step": 143
+    },
+    {
+      "epoch": 0.17560975609756097,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.243902439024391e-06,
+      "loss": 1.1928,
+      "step": 144
+    },
+    {
+      "epoch": 0.17682926829268292,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 8.23170731707317e-06,
+      "loss": 1.1803,
+      "step": 145
+    },
+    {
+      "epoch": 0.17804878048780487,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.219512195121952e-06,
+      "loss": 1.2284,
+      "step": 146
+    },
+    {
+      "epoch": 0.17926829268292682,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.207317073170731e-06,
+      "loss": 1.2206,
+      "step": 147
+    },
+    {
+      "epoch": 0.18048780487804877,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.195121951219512e-06,
+      "loss": 1.1974,
+      "step": 148
+    },
+    {
+      "epoch": 0.18170731707317073,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.182926829268293e-06,
+      "loss": 1.1073,
+      "step": 149
+    },
+    {
+      "epoch": 0.18292682926829268,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.170731707317073e-06,
+      "loss": 1.2636,
+      "step": 150
+    },
+    {
+      "epoch": 0.18414634146341463,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.158536585365854e-06,
+      "loss": 1.163,
+      "step": 151
+    },
+    {
+      "epoch": 0.18536585365853658,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.146341463414635e-06,
+      "loss": 1.0688,
+      "step": 152
+    },
+    {
+      "epoch": 0.18658536585365854,
+      "grad_norm": 0.15234375,
+      "learning_rate": 8.134146341463414e-06,
+      "loss": 1.1666,
+      "step": 153
+    },
+    {
+      "epoch": 0.1878048780487805,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.121951219512197e-06,
+      "loss": 1.2327,
+      "step": 154
+    },
+    {
+      "epoch": 0.18902439024390244,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.109756097560977e-06,
+      "loss": 1.219,
+      "step": 155
+    },
+    {
+      "epoch": 0.1902439024390244,
+      "grad_norm": 0.154296875,
+      "learning_rate": 8.097560975609758e-06,
+      "loss": 1.1597,
+      "step": 156
+    },
+    {
+      "epoch": 0.19146341463414634,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.085365853658537e-06,
+      "loss": 1.1774,
+      "step": 157
+    },
+    {
+      "epoch": 0.1926829268292683,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.073170731707318e-06,
+      "loss": 1.2422,
+      "step": 158
+    },
+    {
+      "epoch": 0.19390243902439025,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.0609756097561e-06,
+      "loss": 1.2281,
+      "step": 159
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 8.048780487804879e-06,
+      "loss": 1.1927,
+      "step": 160
+    },
+    {
+      "epoch": 0.19634146341463415,
+      "grad_norm": 0.15625,
+      "learning_rate": 8.03658536585366e-06,
+      "loss": 1.2419,
+      "step": 161
+    },
+    {
+      "epoch": 0.1975609756097561,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.02439024390244e-06,
+      "loss": 1.1929,
+      "step": 162
+    },
+    {
+      "epoch": 0.19878048780487806,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.01219512195122e-06,
+      "loss": 1.2225,
+      "step": 163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.1928,
+      "step": 164
+    },
+    {
+      "epoch": 0.20121951219512196,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.98780487804878e-06,
+      "loss": 1.1786,
+      "step": 165
+    },
+    {
+      "epoch": 0.20243902439024392,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.975609756097562e-06,
+      "loss": 1.1522,
+      "step": 166
+    },
+    {
+      "epoch": 0.20365853658536584,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.963414634146343e-06,
+      "loss": 1.2015,
+      "step": 167
+    },
+    {
+      "epoch": 0.2048780487804878,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.951219512195122e-06,
+      "loss": 1.1963,
+      "step": 168
+    },
+    {
+      "epoch": 0.20609756097560974,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.939024390243903e-06,
+      "loss": 1.1305,
+      "step": 169
+    },
+    {
+      "epoch": 0.2073170731707317,
+      "grad_norm": 0.1484375,
+      "learning_rate": 7.926829268292685e-06,
+      "loss": 1.1466,
+      "step": 170
+    },
+    {
+      "epoch": 0.20853658536585365,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.914634146341464e-06,
+      "loss": 1.1028,
+      "step": 171
+    },
+    {
+      "epoch": 0.2097560975609756,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.902439024390245e-06,
+      "loss": 1.2034,
+      "step": 172
+    },
+    {
+      "epoch": 0.21097560975609755,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.890243902439026e-06,
+      "loss": 1.2,
+      "step": 173
+    },
+    {
+      "epoch": 0.2121951219512195,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.878048780487806e-06,
+      "loss": 1.1766,
+      "step": 174
+    },
+    {
+      "epoch": 0.21341463414634146,
+      "grad_norm": 0.154296875,
+      "learning_rate": 7.865853658536587e-06,
+      "loss": 1.1024,
+      "step": 175
+    },
+    {
+      "epoch": 0.2146341463414634,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 7.853658536585366e-06,
+      "loss": 1.1219,
+      "step": 176
+    },
+    {
+      "epoch": 0.21585365853658536,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.841463414634147e-06,
+      "loss": 1.2105,
+      "step": 177
+    },
+    {
+      "epoch": 0.21707317073170732,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.829268292682928e-06,
+      "loss": 1.1326,
+      "step": 178
+    },
+    {
+      "epoch": 0.21829268292682927,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.817073170731708e-06,
+      "loss": 1.1262,
+      "step": 179
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 1.1273,
+      "step": 180
+    },
+    {
+      "epoch": 0.22073170731707317,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.79268292682927e-06,
+      "loss": 1.2313,
+      "step": 181
+    },
+    {
+      "epoch": 0.22195121951219512,
+      "grad_norm": 0.154296875,
+      "learning_rate": 7.78048780487805e-06,
+      "loss": 1.174,
+      "step": 182
+    },
+    {
+      "epoch": 0.22317073170731708,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.76829268292683e-06,
+      "loss": 1.2163,
+      "step": 183
+    },
+    {
+      "epoch": 0.22439024390243903,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.75609756097561e-06,
+      "loss": 1.2275,
+      "step": 184
+    },
+    {
+      "epoch": 0.22560975609756098,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.743902439024391e-06,
+      "loss": 1.1812,
+      "step": 185
+    },
+    {
+      "epoch": 0.22682926829268293,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.731707317073172e-06,
+      "loss": 1.1234,
+      "step": 186
+    },
+    {
+      "epoch": 0.2280487804878049,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.719512195121951e-06,
+      "loss": 1.1762,
+      "step": 187
+    },
+    {
+      "epoch": 0.22926829268292684,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.707317073170732e-06,
+      "loss": 1.0488,
+      "step": 188
+    },
+    {
+      "epoch": 0.2304878048780488,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.695121951219514e-06,
+      "loss": 1.1321,
+      "step": 189
+    },
+    {
+      "epoch": 0.23170731707317074,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.682926829268293e-06,
+      "loss": 1.1436,
+      "step": 190
+    },
+    {
+      "epoch": 0.2329268292682927,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.670731707317074e-06,
+      "loss": 1.1194,
+      "step": 191
+    },
+    {
+      "epoch": 0.23414634146341465,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.658536585365855e-06,
+      "loss": 1.2036,
+      "step": 192
+    },
+    {
+      "epoch": 0.23536585365853657,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.646341463414635e-06,
+      "loss": 1.107,
+      "step": 193
+    },
+    {
+      "epoch": 0.23658536585365852,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.634146341463416e-06,
+      "loss": 1.1002,
+      "step": 194
+    },
+    {
+      "epoch": 0.23780487804878048,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.621951219512196e-06,
+      "loss": 1.1532,
+      "step": 195
+    },
+    {
+      "epoch": 0.23902439024390243,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.609756097560976e-06,
+      "loss": 1.1992,
+      "step": 196
+    },
+    {
+      "epoch": 0.24024390243902438,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.5975609756097565e-06,
+      "loss": 1.139,
+      "step": 197
+    },
+    {
+      "epoch": 0.24146341463414633,
+      "grad_norm": 0.154296875,
+      "learning_rate": 7.5853658536585376e-06,
+      "loss": 1.1214,
+      "step": 198
+    },
+    {
+      "epoch": 0.2426829268292683,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.573170731707318e-06,
+      "loss": 1.1419,
+      "step": 199
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 0.203125,
+      "learning_rate": 7.560975609756098e-06,
+      "loss": 1.1848,
+      "step": 200
+    },
+    {
+      "epoch": 0.2451219512195122,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.548780487804878e-06,
+      "loss": 1.1376,
+      "step": 201
+    },
+    {
+      "epoch": 0.24634146341463414,
+      "grad_norm": 0.298828125,
+      "learning_rate": 7.5365853658536594e-06,
+      "loss": 1.1308,
+      "step": 202
+    },
+    {
+      "epoch": 0.2475609756097561,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.52439024390244e-06,
+      "loss": 1.1523,
+      "step": 203
+    },
+    {
+      "epoch": 0.24878048780487805,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.51219512195122e-06,
+      "loss": 1.1728,
+      "step": 204
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.1415,
+      "step": 205
+    },
+    {
+      "epoch": 0.25121951219512195,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.487804878048781e-06,
+      "loss": 1.2315,
+      "step": 206
+    },
+    {
+      "epoch": 0.2524390243902439,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.4756097560975615e-06,
+      "loss": 1.1279,
+      "step": 207
+    },
+    {
+      "epoch": 0.25365853658536586,
+      "grad_norm": 0.208984375,
+      "learning_rate": 7.463414634146342e-06,
+      "loss": 1.2102,
+      "step": 208
+    },
+    {
+      "epoch": 0.2548780487804878,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.451219512195123e-06,
+      "loss": 1.1739,
+      "step": 209
+    },
+    {
+      "epoch": 0.25609756097560976,
+      "grad_norm": 0.240234375,
+      "learning_rate": 7.439024390243903e-06,
+      "loss": 1.1821,
+      "step": 210
+    },
+    {
+      "epoch": 0.2573170731707317,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.426829268292683e-06,
+      "loss": 1.1521,
+      "step": 211
+    },
+    {
+      "epoch": 0.25853658536585367,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.414634146341464e-06,
+      "loss": 1.2337,
+      "step": 212
+    },
+    {
+      "epoch": 0.2597560975609756,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.402439024390245e-06,
+      "loss": 1.1151,
+      "step": 213
+    },
+    {
+      "epoch": 0.26097560975609757,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.390243902439025e-06,
+      "loss": 1.0873,
+      "step": 214
+    },
+    {
+      "epoch": 0.2621951219512195,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.378048780487805e-06,
+      "loss": 1.15,
+      "step": 215
+    },
+    {
+      "epoch": 0.2634146341463415,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.3658536585365855e-06,
+      "loss": 1.2472,
+      "step": 216
+    },
+    {
+      "epoch": 0.2646341463414634,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 7.353658536585367e-06,
+      "loss": 1.1215,
+      "step": 217
+    },
+    {
+      "epoch": 0.2658536585365854,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.341463414634147e-06,
+      "loss": 1.1456,
+      "step": 218
+    },
+    {
+      "epoch": 0.26707317073170733,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.329268292682927e-06,
+      "loss": 1.1391,
+      "step": 219
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.317073170731707e-06,
+      "loss": 1.1412,
+      "step": 220
+    },
+    {
+      "epoch": 0.26951219512195124,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.3048780487804885e-06,
+      "loss": 1.1411,
+      "step": 221
+    },
+    {
+      "epoch": 0.2707317073170732,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.292682926829269e-06,
+      "loss": 1.1139,
+      "step": 222
+    },
+    {
+      "epoch": 0.27195121951219514,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.280487804878049e-06,
+      "loss": 1.0977,
+      "step": 223
+    },
+    {
+      "epoch": 0.2731707317073171,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.268292682926829e-06,
+      "loss": 1.0902,
+      "step": 224
+    },
+    {
+      "epoch": 0.27439024390243905,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.25609756097561e-06,
+      "loss": 1.1449,
+      "step": 225
+    },
+    {
+      "epoch": 0.275609756097561,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.243902439024391e-06,
+      "loss": 1.0953,
+      "step": 226
+    },
+    {
+      "epoch": 0.27682926829268295,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.231707317073171e-06,
+      "loss": 1.1486,
+      "step": 227
+    },
+    {
+      "epoch": 0.2780487804878049,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.219512195121952e-06,
+      "loss": 1.1352,
+      "step": 228
+    },
+    {
+      "epoch": 0.27926829268292686,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.207317073170732e-06,
+      "loss": 1.1072,
+      "step": 229
+    },
+    {
+      "epoch": 0.2804878048780488,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.1951219512195125e-06,
+      "loss": 1.121,
+      "step": 230
+    },
+    {
+      "epoch": 0.2817073170731707,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.182926829268293e-06,
+      "loss": 1.149,
+      "step": 231
+    },
+    {
+      "epoch": 0.28292682926829266,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.170731707317074e-06,
+      "loss": 1.1803,
+      "step": 232
+    },
+    {
+      "epoch": 0.2841463414634146,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.158536585365854e-06,
+      "loss": 1.1377,
+      "step": 233
+    },
+    {
+      "epoch": 0.28536585365853656,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.146341463414634e-06,
+      "loss": 1.1672,
+      "step": 234
+    },
+    {
+      "epoch": 0.2865853658536585,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.1341463414634146e-06,
+      "loss": 1.1429,
+      "step": 235
+    },
+    {
+      "epoch": 0.28780487804878047,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.121951219512196e-06,
+      "loss": 1.0978,
+      "step": 236
+    },
+    {
+      "epoch": 0.2890243902439024,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.109756097560976e-06,
+      "loss": 1.1666,
+      "step": 237
+    },
+    {
+      "epoch": 0.29024390243902437,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.097560975609756e-06,
+      "loss": 1.0938,
+      "step": 238
+    },
+    {
+      "epoch": 0.2914634146341463,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.0853658536585364e-06,
+      "loss": 1.1061,
+      "step": 239
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.0731707317073175e-06,
+      "loss": 1.1368,
+      "step": 240
+    },
+    {
+      "epoch": 0.2939024390243902,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.060975609756098e-06,
+      "loss": 1.219,
+      "step": 241
+    },
+    {
+      "epoch": 0.2951219512195122,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.048780487804878e-06,
+      "loss": 1.0905,
+      "step": 242
+    },
+    {
+      "epoch": 0.29634146341463413,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.036585365853658e-06,
+      "loss": 1.1605,
+      "step": 243
+    },
+    {
+      "epoch": 0.2975609756097561,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.024390243902439e-06,
+      "loss": 1.1597,
+      "step": 244
+    },
+    {
+      "epoch": 0.29878048780487804,
+      "grad_norm": 0.208984375,
+      "learning_rate": 7.01219512195122e-06,
+      "loss": 1.0747,
+      "step": 245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7e-06,
+      "loss": 1.1406,
+      "step": 246
+    },
+    {
+      "epoch": 0.30121951219512194,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.987804878048781e-06,
+      "loss": 1.1297,
+      "step": 247
+    },
+    {
+      "epoch": 0.3024390243902439,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.975609756097561e-06,
+      "loss": 1.1169,
+      "step": 248
+    },
+    {
+      "epoch": 0.30365853658536585,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.9634146341463415e-06,
+      "loss": 1.1603,
+      "step": 249
+    },
+    {
+      "epoch": 0.3048780487804878,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.951219512195122e-06,
+      "loss": 1.1133,
+      "step": 250
+    },
+    {
+      "epoch": 0.30609756097560975,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.939024390243903e-06,
+      "loss": 1.1128,
+      "step": 251
+    },
+    {
+      "epoch": 0.3073170731707317,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.926829268292683e-06,
+      "loss": 1.1305,
+      "step": 252
+    },
+    {
+      "epoch": 0.30853658536585366,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.914634146341463e-06,
+      "loss": 1.1317,
+      "step": 253
+    },
+    {
+      "epoch": 0.3097560975609756,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.902439024390244e-06,
+      "loss": 1.1611,
+      "step": 254
+    },
+    {
+      "epoch": 0.31097560975609756,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.890243902439025e-06,
+      "loss": 1.1758,
+      "step": 255
+    },
+    {
+      "epoch": 0.3121951219512195,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.878048780487805e-06,
+      "loss": 1.1388,
+      "step": 256
+    },
+    {
+      "epoch": 0.31341463414634146,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.865853658536586e-06,
+      "loss": 1.1002,
+      "step": 257
+    },
+    {
+      "epoch": 0.3146341463414634,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.853658536585367e-06,
+      "loss": 1.1417,
+      "step": 258
+    },
+    {
+      "epoch": 0.31585365853658537,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.8414634146341474e-06,
+      "loss": 1.1438,
+      "step": 259
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 1.1272,
+      "step": 260
+    },
+    {
+      "epoch": 0.3182926829268293,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.817073170731709e-06,
+      "loss": 1.124,
+      "step": 261
+    },
+    {
+      "epoch": 0.3195121951219512,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.804878048780489e-06,
+      "loss": 1.1538,
+      "step": 262
+    },
+    {
+      "epoch": 0.3207317073170732,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.792682926829269e-06,
+      "loss": 1.0675,
+      "step": 263
+    },
+    {
+      "epoch": 0.32195121951219513,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.7804878048780495e-06,
+      "loss": 1.1493,
+      "step": 264
+    },
+    {
+      "epoch": 0.3231707317073171,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.768292682926831e-06,
+      "loss": 1.1708,
+      "step": 265
+    },
+    {
+      "epoch": 0.32439024390243903,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.756097560975611e-06,
+      "loss": 1.1503,
+      "step": 266
+    },
+    {
+      "epoch": 0.325609756097561,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.743902439024391e-06,
+      "loss": 1.0744,
+      "step": 267
+    },
+    {
+      "epoch": 0.32682926829268294,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.731707317073171e-06,
+      "loss": 1.1988,
+      "step": 268
+    },
+    {
+      "epoch": 0.3280487804878049,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.7195121951219525e-06,
+      "loss": 1.1247,
+      "step": 269
+    },
+    {
+      "epoch": 0.32926829268292684,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.707317073170733e-06,
+      "loss": 1.1906,
+      "step": 270
+    },
+    {
+      "epoch": 0.3304878048780488,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.695121951219513e-06,
+      "loss": 1.1284,
+      "step": 271
+    },
+    {
+      "epoch": 0.33170731707317075,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.682926829268293e-06,
+      "loss": 1.1857,
+      "step": 272
+    },
+    {
+      "epoch": 0.3329268292682927,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.670731707317074e-06,
+      "loss": 1.0967,
+      "step": 273
+    },
+    {
+      "epoch": 0.33414634146341465,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.658536585365855e-06,
+      "loss": 1.1155,
+      "step": 274
+    },
+    {
+      "epoch": 0.3353658536585366,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.646341463414635e-06,
+      "loss": 1.1599,
+      "step": 275
+    },
+    {
+      "epoch": 0.33658536585365856,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.634146341463415e-06,
+      "loss": 1.1203,
+      "step": 276
+    },
+    {
+      "epoch": 0.3378048780487805,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.621951219512196e-06,
+      "loss": 1.0625,
+      "step": 277
+    },
+    {
+      "epoch": 0.33902439024390246,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.6097560975609765e-06,
+      "loss": 1.1431,
+      "step": 278
+    },
+    {
+      "epoch": 0.3402439024390244,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.597560975609757e-06,
+      "loss": 1.1306,
+      "step": 279
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.585365853658538e-06,
+      "loss": 1.1914,
+      "step": 280
+    },
+    {
+      "epoch": 0.3426829268292683,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.573170731707318e-06,
+      "loss": 1.115,
+      "step": 281
+    },
+    {
+      "epoch": 0.3439024390243902,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.560975609756098e-06,
+      "loss": 1.1587,
+      "step": 282
+    },
+    {
+      "epoch": 0.34512195121951217,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.548780487804879e-06,
+      "loss": 1.105,
+      "step": 283
+    },
+    {
+      "epoch": 0.3463414634146341,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.53658536585366e-06,
+      "loss": 1.1598,
+      "step": 284
+    },
+    {
+      "epoch": 0.3475609756097561,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.52439024390244e-06,
+      "loss": 1.111,
+      "step": 285
+    },
+    {
+      "epoch": 0.348780487804878,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.51219512195122e-06,
+      "loss": 1.1084,
+      "step": 286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.1751,
+      "step": 287
+    },
+    {
+      "epoch": 0.35121951219512193,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.4878048780487815e-06,
+      "loss": 1.138,
+      "step": 288
+    },
+    {
+      "epoch": 0.3524390243902439,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.475609756097562e-06,
+      "loss": 1.1036,
+      "step": 289
+    },
+    {
+      "epoch": 0.35365853658536583,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.463414634146342e-06,
+      "loss": 1.1254,
+      "step": 290
+    },
+    {
+      "epoch": 0.3548780487804878,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.451219512195122e-06,
+      "loss": 1.072,
+      "step": 291
+    },
+    {
+      "epoch": 0.35609756097560974,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.439024390243903e-06,
+      "loss": 1.1864,
+      "step": 292
+    },
+    {
+      "epoch": 0.3573170731707317,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.426829268292684e-06,
+      "loss": 1.0325,
+      "step": 293
+    },
+    {
+      "epoch": 0.35853658536585364,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.414634146341464e-06,
+      "loss": 1.1022,
+      "step": 294
+    },
+    {
+      "epoch": 0.3597560975609756,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.402439024390244e-06,
+      "loss": 1.1639,
+      "step": 295
+    },
+    {
+      "epoch": 0.36097560975609755,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.390243902439025e-06,
+      "loss": 1.0428,
+      "step": 296
+    },
+    {
+      "epoch": 0.3621951219512195,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.3780487804878055e-06,
+      "loss": 1.1188,
+      "step": 297
+    },
+    {
+      "epoch": 0.36341463414634145,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.365853658536586e-06,
+      "loss": 1.0602,
+      "step": 298
+    },
+    {
+      "epoch": 0.3646341463414634,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.353658536585367e-06,
+      "loss": 1.1376,
+      "step": 299
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.341463414634147e-06,
+      "loss": 1.1395,
+      "step": 300
+    },
+    {
+      "epoch": 0.3670731707317073,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.329268292682927e-06,
+      "loss": 1.1327,
+      "step": 301
+    },
+    {
+      "epoch": 0.36829268292682926,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.317073170731708e-06,
+      "loss": 1.1258,
+      "step": 302
+    },
+    {
+      "epoch": 0.3695121951219512,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.304878048780489e-06,
+      "loss": 1.0819,
+      "step": 303
+    },
+    {
+      "epoch": 0.37073170731707317,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.292682926829269e-06,
+      "loss": 1.1154,
+      "step": 304
+    },
+    {
+      "epoch": 0.3719512195121951,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.280487804878049e-06,
+      "loss": 1.0887,
+      "step": 305
+    },
+    {
+      "epoch": 0.37317073170731707,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.2682926829268295e-06,
+      "loss": 1.1633,
+      "step": 306
+    },
+    {
+      "epoch": 0.374390243902439,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.256097560975611e-06,
+      "loss": 1.1101,
+      "step": 307
+    },
+    {
+      "epoch": 0.375609756097561,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.243902439024391e-06,
+      "loss": 1.1236,
+      "step": 308
+    },
+    {
+      "epoch": 0.37682926829268293,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.231707317073171e-06,
+      "loss": 1.1329,
+      "step": 309
+    },
+    {
+      "epoch": 0.3780487804878049,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.219512195121951e-06,
+      "loss": 1.1476,
+      "step": 310
+    },
+    {
+      "epoch": 0.37926829268292683,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.2073170731707325e-06,
+      "loss": 1.0952,
+      "step": 311
+    },
+    {
+      "epoch": 0.3804878048780488,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.195121951219513e-06,
+      "loss": 1.0978,
+      "step": 312
+    },
+    {
+      "epoch": 0.38170731707317074,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.182926829268293e-06,
+      "loss": 1.0685,
+      "step": 313
+    },
+    {
+      "epoch": 0.3829268292682927,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.170731707317073e-06,
+      "loss": 1.1741,
+      "step": 314
+    },
+    {
+      "epoch": 0.38414634146341464,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.158536585365854e-06,
+      "loss": 1.1777,
+      "step": 315
+    },
+    {
+      "epoch": 0.3853658536585366,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.1463414634146346e-06,
+      "loss": 1.1392,
+      "step": 316
+    },
+    {
+      "epoch": 0.38658536585365855,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.134146341463415e-06,
+      "loss": 1.0785,
+      "step": 317
+    },
+    {
+      "epoch": 0.3878048780487805,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.121951219512196e-06,
+      "loss": 1.1306,
+      "step": 318
+    },
+    {
+      "epoch": 0.38902439024390245,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.109756097560976e-06,
+      "loss": 1.0437,
+      "step": 319
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.0975609756097564e-06,
+      "loss": 1.167,
+      "step": 320
+    },
+    {
+      "epoch": 0.39146341463414636,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.085365853658537e-06,
+      "loss": 1.135,
+      "step": 321
+    },
+    {
+      "epoch": 0.3926829268292683,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.073170731707318e-06,
+      "loss": 1.119,
+      "step": 322
+    },
+    {
+      "epoch": 0.39390243902439026,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.060975609756098e-06,
+      "loss": 1.1277,
+      "step": 323
+    },
+    {
+      "epoch": 0.3951219512195122,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.048780487804878e-06,
+      "loss": 1.1658,
+      "step": 324
+    },
+    {
+      "epoch": 0.39634146341463417,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.0365853658536585e-06,
+      "loss": 1.1558,
+      "step": 325
+    },
+    {
+      "epoch": 0.3975609756097561,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.02439024390244e-06,
+      "loss": 1.1281,
+      "step": 326
+    },
+    {
+      "epoch": 0.39878048780487807,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.01219512195122e-06,
+      "loss": 1.1541,
+      "step": 327
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6e-06,
+      "loss": 1.1222,
+      "step": 328
+    },
+    {
+      "epoch": 0.401219512195122,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.98780487804878e-06,
+      "loss": 1.0978,
+      "step": 329
+    },
+    {
+      "epoch": 0.4024390243902439,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.9756097560975615e-06,
+      "loss": 1.1674,
+      "step": 330
+    },
+    {
+      "epoch": 0.4036585365853659,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.963414634146342e-06,
+      "loss": 1.0761,
+      "step": 331
+    },
+    {
+      "epoch": 0.40487804878048783,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.951219512195122e-06,
+      "loss": 1.059,
+      "step": 332
+    },
+    {
+      "epoch": 0.4060975609756098,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.939024390243902e-06,
+      "loss": 1.1001,
+      "step": 333
+    },
+    {
+      "epoch": 0.4073170731707317,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.926829268292683e-06,
+      "loss": 1.0726,
+      "step": 334
+    },
+    {
+      "epoch": 0.40853658536585363,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.914634146341464e-06,
+      "loss": 1.1478,
+      "step": 335
+    },
+    {
+      "epoch": 0.4097560975609756,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.902439024390244e-06,
+      "loss": 1.0873,
+      "step": 336
+    },
+    {
+      "epoch": 0.41097560975609754,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.890243902439025e-06,
+      "loss": 1.0865,
+      "step": 337
+    },
+    {
+      "epoch": 0.4121951219512195,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.878048780487805e-06,
+      "loss": 1.1033,
+      "step": 338
+    },
+    {
+      "epoch": 0.41341463414634144,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.8658536585365855e-06,
+      "loss": 1.1067,
+      "step": 339
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 1.1376,
+      "step": 340
+    },
+    {
+      "epoch": 0.41585365853658535,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.841463414634147e-06,
+      "loss": 1.0712,
+      "step": 341
+    },
+    {
+      "epoch": 0.4170731707317073,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.829268292682927e-06,
+      "loss": 1.0799,
+      "step": 342
+    },
+    {
+      "epoch": 0.41829268292682925,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.817073170731707e-06,
+      "loss": 1.1425,
+      "step": 343
+    },
+    {
+      "epoch": 0.4195121951219512,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.804878048780488e-06,
+      "loss": 1.1261,
+      "step": 344
+    },
+    {
+      "epoch": 0.42073170731707316,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.792682926829269e-06,
+      "loss": 1.1278,
+      "step": 345
+    },
+    {
+      "epoch": 0.4219512195121951,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.780487804878049e-06,
+      "loss": 1.1108,
+      "step": 346
+    },
+    {
+      "epoch": 0.42317073170731706,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.768292682926829e-06,
+      "loss": 1.1388,
+      "step": 347
+    },
+    {
+      "epoch": 0.424390243902439,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.7560975609756095e-06,
+      "loss": 1.096,
+      "step": 348
+    },
+    {
+      "epoch": 0.42560975609756097,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.7439024390243906e-06,
+      "loss": 1.0673,
+      "step": 349
+    },
+    {
+      "epoch": 0.4268292682926829,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.731707317073171e-06,
+      "loss": 1.1634,
+      "step": 350
+    },
+    {
+      "epoch": 0.42804878048780487,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.719512195121951e-06,
+      "loss": 1.1028,
+      "step": 351
+    },
+    {
+      "epoch": 0.4292682926829268,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.707317073170731e-06,
+      "loss": 1.101,
+      "step": 352
+    },
+    {
+      "epoch": 0.4304878048780488,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.695121951219512e-06,
+      "loss": 1.1242,
+      "step": 353
+    },
+    {
+      "epoch": 0.4317073170731707,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.682926829268293e-06,
+      "loss": 1.124,
+      "step": 354
+    },
+    {
+      "epoch": 0.4329268292682927,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.670731707317073e-06,
+      "loss": 1.0535,
+      "step": 355
+    },
+    {
+      "epoch": 0.43414634146341463,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.658536585365853e-06,
+      "loss": 1.1274,
+      "step": 356
+    },
+    {
+      "epoch": 0.4353658536585366,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.646341463414634e-06,
+      "loss": 1.093,
+      "step": 357
+    },
+    {
+      "epoch": 0.43658536585365854,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.6341463414634145e-06,
+      "loss": 1.1121,
+      "step": 358
+    },
+    {
+      "epoch": 0.4378048780487805,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.6219512195121965e-06,
+      "loss": 1.1554,
+      "step": 359
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.609756097560977e-06,
+      "loss": 1.1646,
+      "step": 360
+    },
+    {
+      "epoch": 0.4402439024390244,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.597560975609757e-06,
+      "loss": 1.1567,
+      "step": 361
+    },
+    {
+      "epoch": 0.44146341463414634,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.585365853658537e-06,
+      "loss": 1.1338,
+      "step": 362
+    },
+    {
+      "epoch": 0.4426829268292683,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.573170731707318e-06,
+      "loss": 1.0403,
+      "step": 363
+    },
+    {
+      "epoch": 0.44390243902439025,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.560975609756099e-06,
+      "loss": 1.0957,
+      "step": 364
+    },
+    {
+      "epoch": 0.4451219512195122,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.548780487804879e-06,
+      "loss": 1.0949,
+      "step": 365
+    },
+    {
+      "epoch": 0.44634146341463415,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.536585365853659e-06,
+      "loss": 1.1318,
+      "step": 366
+    },
+    {
+      "epoch": 0.4475609756097561,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.52439024390244e-06,
+      "loss": 1.1446,
+      "step": 367
+    },
+    {
+      "epoch": 0.44878048780487806,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.5121951219512205e-06,
+      "loss": 1.1105,
+      "step": 368
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.0745,
+      "step": 369
+    },
+    {
+      "epoch": 0.45121951219512196,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.487804878048781e-06,
+      "loss": 1.1463,
+      "step": 370
+    },
+    {
+      "epoch": 0.4524390243902439,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.475609756097562e-06,
+      "loss": 1.1111,
+      "step": 371
+    },
+    {
+      "epoch": 0.45365853658536587,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.463414634146342e-06,
+      "loss": 1.1183,
+      "step": 372
+    },
+    {
+      "epoch": 0.4548780487804878,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.4512195121951226e-06,
+      "loss": 1.1308,
+      "step": 373
+    },
+    {
+      "epoch": 0.4560975609756098,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.439024390243904e-06,
+      "loss": 1.0861,
+      "step": 374
+    },
+    {
+      "epoch": 0.4573170731707317,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.426829268292684e-06,
+      "loss": 1.0827,
+      "step": 375
+    },
+    {
+      "epoch": 0.4585365853658537,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.414634146341464e-06,
+      "loss": 1.1014,
+      "step": 376
+    },
+    {
+      "epoch": 0.45975609756097563,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.4024390243902444e-06,
+      "loss": 1.1291,
+      "step": 377
+    },
+    {
+      "epoch": 0.4609756097560976,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.3902439024390255e-06,
+      "loss": 1.0899,
+      "step": 378
+    },
+    {
+      "epoch": 0.46219512195121953,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.378048780487806e-06,
+      "loss": 1.0797,
+      "step": 379
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.365853658536586e-06,
+      "loss": 1.0599,
+      "step": 380
+    },
+    {
+      "epoch": 0.46463414634146344,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.353658536585366e-06,
+      "loss": 1.1505,
+      "step": 381
+    },
+    {
+      "epoch": 0.4658536585365854,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.341463414634147e-06,
+      "loss": 1.0887,
+      "step": 382
+    },
+    {
+      "epoch": 0.46707317073170734,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.329268292682928e-06,
+      "loss": 1.0973,
+      "step": 383
+    },
+    {
+      "epoch": 0.4682926829268293,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.317073170731708e-06,
+      "loss": 1.08,
+      "step": 384
+    },
+    {
+      "epoch": 0.4695121951219512,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.304878048780488e-06,
+      "loss": 1.1254,
+      "step": 385
+    },
+    {
+      "epoch": 0.47073170731707314,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.292682926829269e-06,
+      "loss": 1.0845,
+      "step": 386
+    },
+    {
+      "epoch": 0.4719512195121951,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.2804878048780495e-06,
+      "loss": 1.111,
+      "step": 387
+    },
+    {
+      "epoch": 0.47317073170731705,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.26829268292683e-06,
+      "loss": 1.0589,
+      "step": 388
+    },
+    {
+      "epoch": 0.474390243902439,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.25609756097561e-06,
+      "loss": 1.1279,
+      "step": 389
+    },
+    {
+      "epoch": 0.47560975609756095,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.243902439024391e-06,
+      "loss": 1.0662,
+      "step": 390
+    },
+    {
+      "epoch": 0.4768292682926829,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.231707317073171e-06,
+      "loss": 1.049,
+      "step": 391
+    },
+    {
+      "epoch": 0.47804878048780486,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.219512195121952e-06,
+      "loss": 1.0644,
+      "step": 392
+    },
+    {
+      "epoch": 0.4792682926829268,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.207317073170733e-06,
+      "loss": 1.141,
+      "step": 393
+    },
+    {
+      "epoch": 0.48048780487804876,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.195121951219513e-06,
+      "loss": 1.0814,
+      "step": 394
+    },
+    {
+      "epoch": 0.4817073170731707,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.182926829268293e-06,
+      "loss": 1.0742,
+      "step": 395
+    },
+    {
+      "epoch": 0.48292682926829267,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.1707317073170735e-06,
+      "loss": 1.0399,
+      "step": 396
+    },
+    {
+      "epoch": 0.4841463414634146,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.1585365853658546e-06,
+      "loss": 1.0634,
+      "step": 397
+    },
+    {
+      "epoch": 0.4853658536585366,
+      "grad_norm": 0.232421875,
+      "learning_rate": 5.146341463414635e-06,
+      "loss": 1.131,
+      "step": 398
+    },
+    {
+      "epoch": 0.4865853658536585,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.134146341463415e-06,
+      "loss": 1.0793,
+      "step": 399
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.121951219512195e-06,
+      "loss": 1.0662,
+      "step": 400
+    },
+    {
+      "epoch": 0.48902439024390243,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.1097560975609764e-06,
+      "loss": 1.178,
+      "step": 401
+    },
+    {
+      "epoch": 0.4902439024390244,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.097560975609757e-06,
+      "loss": 1.0784,
+      "step": 402
+    },
+    {
+      "epoch": 0.49146341463414633,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.085365853658537e-06,
+      "loss": 1.1118,
+      "step": 403
+    },
+    {
+      "epoch": 0.4926829268292683,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.073170731707317e-06,
+      "loss": 1.1315,
+      "step": 404
+    },
+    {
+      "epoch": 0.49390243902439024,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.060975609756098e-06,
+      "loss": 1.0338,
+      "step": 405
+    },
+    {
+      "epoch": 0.4951219512195122,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.0487804878048785e-06,
+      "loss": 1.1451,
+      "step": 406
+    },
+    {
+      "epoch": 0.49634146341463414,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.036585365853659e-06,
+      "loss": 1.0775,
+      "step": 407
+    },
+    {
+      "epoch": 0.4975609756097561,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.024390243902439e-06,
+      "loss": 1.0834,
+      "step": 408
+    },
+    {
+      "epoch": 0.49878048780487805,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.01219512195122e-06,
+      "loss": 1.0719,
+      "step": 409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.37109375,
+      "learning_rate": 5e-06,
+      "loss": 1.2225,
+      "step": 410
+    },
+    {
+      "epoch": 0.501219512195122,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.987804878048781e-06,
+      "loss": 1.0514,
+      "step": 411
+    },
+    {
+      "epoch": 0.5024390243902439,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.975609756097562e-06,
+      "loss": 1.1518,
+      "step": 412
+    },
+    {
+      "epoch": 0.5036585365853659,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.963414634146342e-06,
+      "loss": 1.1065,
+      "step": 413
+    },
+    {
+      "epoch": 0.5048780487804878,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.951219512195122e-06,
+      "loss": 1.0973,
+      "step": 414
+    },
+    {
+      "epoch": 0.5060975609756098,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.9390243902439025e-06,
+      "loss": 1.1072,
+      "step": 415
+    },
+    {
+      "epoch": 0.5073170731707317,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.926829268292684e-06,
+      "loss": 1.0223,
+      "step": 416
+    },
+    {
+      "epoch": 0.5085365853658537,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.914634146341464e-06,
+      "loss": 1.106,
+      "step": 417
+    },
+    {
+      "epoch": 0.5097560975609756,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.902439024390244e-06,
+      "loss": 1.1571,
+      "step": 418
+    },
+    {
+      "epoch": 0.5109756097560976,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.890243902439024e-06,
+      "loss": 1.1613,
+      "step": 419
+    },
+    {
+      "epoch": 0.5121951219512195,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.0864,
+      "step": 420
+    },
+    {
+      "epoch": 0.5134146341463415,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.865853658536586e-06,
+      "loss": 1.0378,
+      "step": 421
+    },
+    {
+      "epoch": 0.5146341463414634,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.853658536585366e-06,
+      "loss": 1.1185,
+      "step": 422
+    },
+    {
+      "epoch": 0.5158536585365854,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.841463414634146e-06,
+      "loss": 1.0443,
+      "step": 423
+    },
+    {
+      "epoch": 0.5170731707317073,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.829268292682927e-06,
+      "loss": 1.0556,
+      "step": 424
+    },
+    {
+      "epoch": 0.5182926829268293,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.817073170731708e-06,
+      "loss": 1.1298,
+      "step": 425
+    },
+    {
+      "epoch": 0.5195121951219512,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.804878048780488e-06,
+      "loss": 1.065,
+      "step": 426
+    },
+    {
+      "epoch": 0.5207317073170732,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.792682926829268e-06,
+      "loss": 1.1335,
+      "step": 427
+    },
+    {
+      "epoch": 0.5219512195121951,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.780487804878049e-06,
+      "loss": 1.0745,
+      "step": 428
+    },
+    {
+      "epoch": 0.5231707317073171,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.7682926829268295e-06,
+      "loss": 1.0771,
+      "step": 429
+    },
+    {
+      "epoch": 0.524390243902439,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.75609756097561e-06,
+      "loss": 1.0682,
+      "step": 430
+    },
+    {
+      "epoch": 0.525609756097561,
+      "grad_norm": 0.318359375,
+      "learning_rate": 4.743902439024391e-06,
+      "loss": 1.1436,
+      "step": 431
+    },
+    {
+      "epoch": 0.526829268292683,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.731707317073171e-06,
+      "loss": 1.1088,
+      "step": 432
+    },
+    {
+      "epoch": 0.5280487804878049,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.719512195121951e-06,
+      "loss": 1.048,
+      "step": 433
+    },
+    {
+      "epoch": 0.5292682926829269,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.7073170731707316e-06,
+      "loss": 1.0735,
+      "step": 434
+    },
+    {
+      "epoch": 0.5304878048780488,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.695121951219513e-06,
+      "loss": 1.0368,
+      "step": 435
+    },
+    {
+      "epoch": 0.5317073170731708,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.682926829268293e-06,
+      "loss": 1.0993,
+      "step": 436
+    },
+    {
+      "epoch": 0.5329268292682927,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.670731707317074e-06,
+      "loss": 1.0675,
+      "step": 437
+    },
+    {
+      "epoch": 0.5341463414634147,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.658536585365854e-06,
+      "loss": 1.0814,
+      "step": 438
+    },
+    {
+      "epoch": 0.5353658536585366,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.6463414634146345e-06,
+      "loss": 1.1575,
+      "step": 439
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.634146341463416e-06,
+      "loss": 1.0666,
+      "step": 440
+    },
+    {
+      "epoch": 0.5378048780487805,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.621951219512196e-06,
+      "loss": 1.117,
+      "step": 441
+    },
+    {
+      "epoch": 0.5390243902439025,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.609756097560976e-06,
+      "loss": 1.1458,
+      "step": 442
+    },
+    {
+      "epoch": 0.5402439024390244,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.597560975609756e-06,
+      "loss": 1.1491,
+      "step": 443
+    },
+    {
+      "epoch": 0.5414634146341464,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.5853658536585375e-06,
+      "loss": 1.1205,
+      "step": 444
+    },
+    {
+      "epoch": 0.5426829268292683,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.573170731707318e-06,
+      "loss": 1.0387,
+      "step": 445
+    },
+    {
+      "epoch": 0.5439024390243903,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.560975609756098e-06,
+      "loss": 1.0962,
+      "step": 446
+    },
+    {
+      "epoch": 0.5451219512195122,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.548780487804878e-06,
+      "loss": 1.1036,
+      "step": 447
+    },
+    {
+      "epoch": 0.5463414634146342,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.536585365853659e-06,
+      "loss": 1.0776,
+      "step": 448
+    },
+    {
+      "epoch": 0.5475609756097561,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.52439024390244e-06,
+      "loss": 1.1556,
+      "step": 449
+    },
+    {
+      "epoch": 0.5487804878048781,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.51219512195122e-06,
+      "loss": 1.0779,
+      "step": 450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.5e-06,
+      "loss": 1.0078,
+      "step": 451
+    },
+    {
+      "epoch": 0.551219512195122,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.487804878048781e-06,
+      "loss": 1.0247,
+      "step": 452
+    },
+    {
+      "epoch": 0.552439024390244,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.4756097560975615e-06,
+      "loss": 1.1226,
+      "step": 453
+    },
+    {
+      "epoch": 0.5536585365853659,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.463414634146342e-06,
+      "loss": 1.0441,
+      "step": 454
+    },
+    {
+      "epoch": 0.5548780487804879,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.451219512195122e-06,
+      "loss": 1.0545,
+      "step": 455
+    },
+    {
+      "epoch": 0.5560975609756098,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.439024390243903e-06,
+      "loss": 1.0808,
+      "step": 456
+    },
+    {
+      "epoch": 0.5573170731707318,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.426829268292683e-06,
+      "loss": 1.0545,
+      "step": 457
+    },
+    {
+      "epoch": 0.5585365853658537,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.414634146341464e-06,
+      "loss": 1.0619,
+      "step": 458
+    },
+    {
+      "epoch": 0.5597560975609757,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.402439024390245e-06,
+      "loss": 1.0775,
+      "step": 459
+    },
+    {
+      "epoch": 0.5609756097560976,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.390243902439025e-06,
+      "loss": 1.0985,
+      "step": 460
+    },
+    {
+      "epoch": 0.5621951219512196,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.378048780487805e-06,
+      "loss": 1.0534,
+      "step": 461
+    },
+    {
+      "epoch": 0.5634146341463414,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.3658536585365854e-06,
+      "loss": 1.0525,
+      "step": 462
+    },
+    {
+      "epoch": 0.5646341463414634,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.3536585365853665e-06,
+      "loss": 1.0208,
+      "step": 463
+    },
+    {
+      "epoch": 0.5658536585365853,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.341463414634147e-06,
+      "loss": 1.0627,
+      "step": 464
+    },
+    {
+      "epoch": 0.5670731707317073,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.329268292682927e-06,
+      "loss": 1.0544,
+      "step": 465
+    },
+    {
+      "epoch": 0.5682926829268292,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.317073170731707e-06,
+      "loss": 1.1068,
+      "step": 466
+    },
+    {
+      "epoch": 0.5695121951219512,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.304878048780488e-06,
+      "loss": 1.0316,
+      "step": 467
+    },
+    {
+      "epoch": 0.5707317073170731,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.292682926829269e-06,
+      "loss": 1.061,
+      "step": 468
+    },
+    {
+      "epoch": 0.5719512195121951,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.280487804878049e-06,
+      "loss": 1.117,
+      "step": 469
+    },
+    {
+      "epoch": 0.573170731707317,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.268292682926829e-06,
+      "loss": 1.137,
+      "step": 470
+    },
+    {
+      "epoch": 0.574390243902439,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.25609756097561e-06,
+      "loss": 1.0557,
+      "step": 471
+    },
+    {
+      "epoch": 0.5756097560975609,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.2439024390243905e-06,
+      "loss": 1.1299,
+      "step": 472
+    },
+    {
+      "epoch": 0.5768292682926829,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.231707317073171e-06,
+      "loss": 1.0561,
+      "step": 473
+    },
+    {
+      "epoch": 0.5780487804878048,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.219512195121951e-06,
+      "loss": 1.0669,
+      "step": 474
+    },
+    {
+      "epoch": 0.5792682926829268,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.207317073170732e-06,
+      "loss": 1.0306,
+      "step": 475
+    },
+    {
+      "epoch": 0.5804878048780487,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.195121951219512e-06,
+      "loss": 1.1473,
+      "step": 476
+    },
+    {
+      "epoch": 0.5817073170731707,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.182926829268293e-06,
+      "loss": 1.0403,
+      "step": 477
+    },
+    {
+      "epoch": 0.5829268292682926,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.170731707317074e-06,
+      "loss": 1.0267,
+      "step": 478
+    },
+    {
+      "epoch": 0.5841463414634146,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.158536585365854e-06,
+      "loss": 1.1301,
+      "step": 479
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.146341463414634e-06,
+      "loss": 1.0481,
+      "step": 480
+    },
+    {
+      "epoch": 0.5865853658536585,
+      "grad_norm": 0.232421875,
+      "learning_rate": 4.1341463414634145e-06,
+      "loss": 1.1321,
+      "step": 481
+    },
+    {
+      "epoch": 0.5878048780487805,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.121951219512196e-06,
+      "loss": 1.0029,
+      "step": 482
+    },
+    {
+      "epoch": 0.5890243902439024,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.109756097560976e-06,
+      "loss": 1.0561,
+      "step": 483
+    },
+    {
+      "epoch": 0.5902439024390244,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.097560975609756e-06,
+      "loss": 1.0826,
+      "step": 484
+    },
+    {
+      "epoch": 0.5914634146341463,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.085365853658536e-06,
+      "loss": 1.0699,
+      "step": 485
+    },
+    {
+      "epoch": 0.5926829268292683,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.0731707317073175e-06,
+      "loss": 1.1371,
+      "step": 486
+    },
+    {
+      "epoch": 0.5939024390243902,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.0609756097560986e-06,
+      "loss": 1.0756,
+      "step": 487
+    },
+    {
+      "epoch": 0.5951219512195122,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.048780487804879e-06,
+      "loss": 1.0674,
+      "step": 488
+    },
+    {
+      "epoch": 0.5963414634146341,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.036585365853659e-06,
+      "loss": 1.0942,
+      "step": 489
+    },
+    {
+      "epoch": 0.5975609756097561,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.024390243902439e-06,
+      "loss": 1.0999,
+      "step": 490
+    },
+    {
+      "epoch": 0.598780487804878,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.01219512195122e-06,
+      "loss": 1.141,
+      "step": 491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.1518,
+      "step": 492
+    },
+    {
+      "epoch": 0.6012195121951219,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.987804878048781e-06,
+      "loss": 1.0763,
+      "step": 493
+    },
+    {
+      "epoch": 0.6024390243902439,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.975609756097561e-06,
+      "loss": 1.0236,
+      "step": 494
+    },
+    {
+      "epoch": 0.6036585365853658,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.963414634146342e-06,
+      "loss": 1.0964,
+      "step": 495
+    },
+    {
+      "epoch": 0.6048780487804878,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.9512195121951225e-06,
+      "loss": 1.0978,
+      "step": 496
+    },
+    {
+      "epoch": 0.6060975609756097,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.939024390243903e-06,
+      "loss": 1.0856,
+      "step": 497
+    },
+    {
+      "epoch": 0.6073170731707317,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.926829268292683e-06,
+      "loss": 1.1262,
+      "step": 498
+    },
+    {
+      "epoch": 0.6085365853658536,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.914634146341464e-06,
+      "loss": 1.1766,
+      "step": 499
+    },
+    {
+      "epoch": 0.6097560975609756,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.0285,
+      "step": 500
+    },
+    {
+      "epoch": 0.6109756097560975,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.890243902439025e-06,
+      "loss": 1.0808,
+      "step": 501
+    },
+    {
+      "epoch": 0.6121951219512195,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.878048780487805e-06,
+      "loss": 1.098,
+      "step": 502
+    },
+    {
+      "epoch": 0.6134146341463415,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.865853658536586e-06,
+      "loss": 1.1352,
+      "step": 503
+    },
+    {
+      "epoch": 0.6146341463414634,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.853658536585366e-06,
+      "loss": 1.116,
+      "step": 504
+    },
+    {
+      "epoch": 0.6158536585365854,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.8414634146341465e-06,
+      "loss": 1.0465,
+      "step": 505
+    },
+    {
+      "epoch": 0.6170731707317073,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.829268292682928e-06,
+      "loss": 1.0884,
+      "step": 506
+    },
+    {
+      "epoch": 0.6182926829268293,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.817073170731708e-06,
+      "loss": 1.0589,
+      "step": 507
+    },
+    {
+      "epoch": 0.6195121951219512,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.804878048780488e-06,
+      "loss": 1.038,
+      "step": 508
+    },
+    {
+      "epoch": 0.6207317073170732,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.7926829268292688e-06,
+      "loss": 1.029,
+      "step": 509
+    },
+    {
+      "epoch": 0.6219512195121951,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.780487804878049e-06,
+      "loss": 1.1342,
+      "step": 510
+    },
+    {
+      "epoch": 0.6231707317073171,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.7682926829268297e-06,
+      "loss": 1.1217,
+      "step": 511
+    },
+    {
+      "epoch": 0.624390243902439,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.75609756097561e-06,
+      "loss": 1.073,
+      "step": 512
+    },
+    {
+      "epoch": 0.625609756097561,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.7439024390243906e-06,
+      "loss": 1.0929,
+      "step": 513
+    },
+    {
+      "epoch": 0.6268292682926829,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.731707317073171e-06,
+      "loss": 1.1077,
+      "step": 514
+    },
+    {
+      "epoch": 0.6280487804878049,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.7195121951219516e-06,
+      "loss": 1.0438,
+      "step": 515
+    },
+    {
+      "epoch": 0.6292682926829268,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.707317073170732e-06,
+      "loss": 1.1266,
+      "step": 516
+    },
+    {
+      "epoch": 0.6304878048780488,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.6951219512195125e-06,
+      "loss": 1.0465,
+      "step": 517
+    },
+    {
+      "epoch": 0.6317073170731707,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.6829268292682928e-06,
+      "loss": 1.0808,
+      "step": 518
+    },
+    {
+      "epoch": 0.6329268292682927,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.6707317073170734e-06,
+      "loss": 1.1272,
+      "step": 519
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.6585365853658537e-06,
+      "loss": 1.0783,
+      "step": 520
+    },
+    {
+      "epoch": 0.6353658536585366,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.6463414634146344e-06,
+      "loss": 1.0423,
+      "step": 521
+    },
+    {
+      "epoch": 0.6365853658536585,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.6341463414634146e-06,
+      "loss": 1.1099,
+      "step": 522
+    },
+    {
+      "epoch": 0.6378048780487805,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.6219512195121953e-06,
+      "loss": 1.0481,
+      "step": 523
+    },
+    {
+      "epoch": 0.6390243902439025,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.609756097560976e-06,
+      "loss": 1.0337,
+      "step": 524
+    },
+    {
+      "epoch": 0.6402439024390244,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.5975609756097562e-06,
+      "loss": 1.0519,
+      "step": 525
+    },
+    {
+      "epoch": 0.6414634146341464,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.585365853658537e-06,
+      "loss": 1.1016,
+      "step": 526
+    },
+    {
+      "epoch": 0.6426829268292683,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.573170731707317e-06,
+      "loss": 1.1309,
+      "step": 527
+    },
+    {
+      "epoch": 0.6439024390243903,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.560975609756098e-06,
+      "loss": 1.1033,
+      "step": 528
+    },
+    {
+      "epoch": 0.6451219512195122,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.548780487804878e-06,
+      "loss": 1.09,
+      "step": 529
+    },
+    {
+      "epoch": 0.6463414634146342,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.5365853658536588e-06,
+      "loss": 1.0413,
+      "step": 530
+    },
+    {
+      "epoch": 0.6475609756097561,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.524390243902439e-06,
+      "loss": 1.1486,
+      "step": 531
+    },
+    {
+      "epoch": 0.6487804878048781,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.5121951219512197e-06,
+      "loss": 1.072,
+      "step": 532
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.5e-06,
+      "loss": 1.0675,
+      "step": 533
+    },
+    {
+      "epoch": 0.651219512195122,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.4878048780487806e-06,
+      "loss": 1.0719,
+      "step": 534
+    },
+    {
+      "epoch": 0.6524390243902439,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.475609756097561e-06,
+      "loss": 1.1005,
+      "step": 535
+    },
+    {
+      "epoch": 0.6536585365853659,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.4634146341463416e-06,
+      "loss": 1.1054,
+      "step": 536
+    },
+    {
+      "epoch": 0.6548780487804878,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.451219512195122e-06,
+      "loss": 1.0733,
+      "step": 537
+    },
+    {
+      "epoch": 0.6560975609756098,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.4390243902439025e-06,
+      "loss": 1.0682,
+      "step": 538
+    },
+    {
+      "epoch": 0.6573170731707317,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.4268292682926836e-06,
+      "loss": 1.055,
+      "step": 539
+    },
+    {
+      "epoch": 0.6585365853658537,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.414634146341464e-06,
+      "loss": 1.0479,
+      "step": 540
+    },
+    {
+      "epoch": 0.6597560975609756,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.4024390243902445e-06,
+      "loss": 1.0739,
+      "step": 541
+    },
+    {
+      "epoch": 0.6609756097560976,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.3902439024390248e-06,
+      "loss": 1.0636,
+      "step": 542
+    },
+    {
+      "epoch": 0.6621951219512195,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.3780487804878054e-06,
+      "loss": 1.0841,
+      "step": 543
+    },
+    {
+      "epoch": 0.6634146341463415,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.3658536585365857e-06,
+      "loss": 1.0363,
+      "step": 544
+    },
+    {
+      "epoch": 0.6646341463414634,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.3536585365853664e-06,
+      "loss": 1.0939,
+      "step": 545
+    },
+    {
+      "epoch": 0.6658536585365854,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.3414634146341466e-06,
+      "loss": 1.0318,
+      "step": 546
+    },
+    {
+      "epoch": 0.6670731707317074,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.3292682926829273e-06,
+      "loss": 1.0734,
+      "step": 547
+    },
+    {
+      "epoch": 0.6682926829268293,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.3170731707317076e-06,
+      "loss": 1.1419,
+      "step": 548
+    },
+    {
+      "epoch": 0.6695121951219513,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.3048780487804882e-06,
+      "loss": 1.128,
+      "step": 549
+    },
+    {
+      "epoch": 0.6707317073170732,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.292682926829269e-06,
+      "loss": 1.0225,
+      "step": 550
+    },
+    {
+      "epoch": 0.6719512195121952,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.280487804878049e-06,
+      "loss": 1.0659,
+      "step": 551
+    },
+    {
+      "epoch": 0.6731707317073171,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.26829268292683e-06,
+      "loss": 1.1373,
+      "step": 552
+    },
+    {
+      "epoch": 0.6743902439024391,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.25609756097561e-06,
+      "loss": 1.0377,
+      "step": 553
+    },
+    {
+      "epoch": 0.675609756097561,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.2439024390243908e-06,
+      "loss": 1.0002,
+      "step": 554
+    },
+    {
+      "epoch": 0.676829268292683,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.231707317073171e-06,
+      "loss": 1.0513,
+      "step": 555
+    },
+    {
+      "epoch": 0.6780487804878049,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.2195121951219517e-06,
+      "loss": 1.0748,
+      "step": 556
+    },
+    {
+      "epoch": 0.6792682926829269,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.207317073170732e-06,
+      "loss": 1.1373,
+      "step": 557
+    },
+    {
+      "epoch": 0.6804878048780488,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.1951219512195126e-06,
+      "loss": 1.085,
+      "step": 558
+    },
+    {
+      "epoch": 0.6817073170731708,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.182926829268293e-06,
+      "loss": 1.0097,
+      "step": 559
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.1707317073170736e-06,
+      "loss": 1.1246,
+      "step": 560
+    },
+    {
+      "epoch": 0.6841463414634147,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.158536585365854e-06,
+      "loss": 1.0998,
+      "step": 561
+    },
+    {
+      "epoch": 0.6853658536585366,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.1463414634146345e-06,
+      "loss": 1.0452,
+      "step": 562
+    },
+    {
+      "epoch": 0.6865853658536586,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.1341463414634147e-06,
+      "loss": 1.0788,
+      "step": 563
+    },
+    {
+      "epoch": 0.6878048780487804,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.1219512195121954e-06,
+      "loss": 1.0457,
+      "step": 564
+    },
+    {
+      "epoch": 0.6890243902439024,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.1097560975609757e-06,
+      "loss": 1.0646,
+      "step": 565
+    },
+    {
+      "epoch": 0.6902439024390243,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.0975609756097564e-06,
+      "loss": 1.0374,
+      "step": 566
+    },
+    {
+      "epoch": 0.6914634146341463,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.0853658536585366e-06,
+      "loss": 1.0734,
+      "step": 567
+    },
+    {
+      "epoch": 0.6926829268292682,
+      "grad_norm": 0.208984375,
+      "learning_rate": 3.0731707317073173e-06,
+      "loss": 1.0418,
+      "step": 568
+    },
+    {
+      "epoch": 0.6939024390243902,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.060975609756098e-06,
+      "loss": 1.0873,
+      "step": 569
+    },
+    {
+      "epoch": 0.6951219512195121,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.0487804878048782e-06,
+      "loss": 1.0968,
+      "step": 570
+    },
+    {
+      "epoch": 0.6963414634146341,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.036585365853659e-06,
+      "loss": 1.1014,
+      "step": 571
+    },
+    {
+      "epoch": 0.697560975609756,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.024390243902439e-06,
+      "loss": 1.0726,
+      "step": 572
+    },
+    {
+      "epoch": 0.698780487804878,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.01219512195122e-06,
+      "loss": 1.0602,
+      "step": 573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3e-06,
+      "loss": 1.0828,
+      "step": 574
+    },
+    {
+      "epoch": 0.7012195121951219,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.9878048780487808e-06,
+      "loss": 1.1386,
+      "step": 575
+    },
+    {
+      "epoch": 0.7024390243902439,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.975609756097561e-06,
+      "loss": 1.104,
+      "step": 576
+    },
+    {
+      "epoch": 0.7036585365853658,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.9634146341463417e-06,
+      "loss": 1.0109,
+      "step": 577
+    },
+    {
+      "epoch": 0.7048780487804878,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.951219512195122e-06,
+      "loss": 1.0697,
+      "step": 578
+    },
+    {
+      "epoch": 0.7060975609756097,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.9390243902439026e-06,
+      "loss": 1.1323,
+      "step": 579
+    },
+    {
+      "epoch": 0.7073170731707317,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.083,
+      "step": 580
+    },
+    {
+      "epoch": 0.7085365853658536,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.9146341463414635e-06,
+      "loss": 1.0529,
+      "step": 581
+    },
+    {
+      "epoch": 0.7097560975609756,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.902439024390244e-06,
+      "loss": 1.0708,
+      "step": 582
+    },
+    {
+      "epoch": 0.7109756097560975,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.8902439024390245e-06,
+      "loss": 1.0725,
+      "step": 583
+    },
+    {
+      "epoch": 0.7121951219512195,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.8780487804878047e-06,
+      "loss": 1.0535,
+      "step": 584
+    },
+    {
+      "epoch": 0.7134146341463414,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.8658536585365854e-06,
+      "loss": 1.0558,
+      "step": 585
+    },
+    {
+      "epoch": 0.7146341463414634,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.8536585365853657e-06,
+      "loss": 1.0615,
+      "step": 586
+    },
+    {
+      "epoch": 0.7158536585365853,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.8414634146341463e-06,
+      "loss": 1.0907,
+      "step": 587
+    },
+    {
+      "epoch": 0.7170731707317073,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.8292682926829266e-06,
+      "loss": 1.0498,
+      "step": 588
+    },
+    {
+      "epoch": 0.7182926829268292,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.8170731707317073e-06,
+      "loss": 1.0522,
+      "step": 589
+    },
+    {
+      "epoch": 0.7195121951219512,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.8048780487804884e-06,
+      "loss": 1.1584,
+      "step": 590
+    },
+    {
+      "epoch": 0.7207317073170731,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.7926829268292686e-06,
+      "loss": 1.1063,
+      "step": 591
+    },
+    {
+      "epoch": 0.7219512195121951,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.7804878048780493e-06,
+      "loss": 1.0556,
+      "step": 592
+    },
+    {
+      "epoch": 0.723170731707317,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.7682926829268295e-06,
+      "loss": 1.0481,
+      "step": 593
+    },
+    {
+      "epoch": 0.724390243902439,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.7560975609756102e-06,
+      "loss": 1.0418,
+      "step": 594
+    },
+    {
+      "epoch": 0.725609756097561,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.7439024390243905e-06,
+      "loss": 1.0569,
+      "step": 595
+    },
+    {
+      "epoch": 0.7268292682926829,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.731707317073171e-06,
+      "loss": 1.0622,
+      "step": 596
+    },
+    {
+      "epoch": 0.7280487804878049,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.719512195121952e-06,
+      "loss": 1.0188,
+      "step": 597
+    },
+    {
+      "epoch": 0.7292682926829268,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.707317073170732e-06,
+      "loss": 1.126,
+      "step": 598
+    },
+    {
+      "epoch": 0.7304878048780488,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.6951219512195128e-06,
+      "loss": 1.0651,
+      "step": 599
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.682926829268293e-06,
+      "loss": 1.0791,
+      "step": 600
+    },
+    {
+      "epoch": 0.7329268292682927,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.6707317073170737e-06,
+      "loss": 1.1254,
+      "step": 601
+    },
+    {
+      "epoch": 0.7341463414634146,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.658536585365854e-06,
+      "loss": 1.0697,
+      "step": 602
+    },
+    {
+      "epoch": 0.7353658536585366,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.6463414634146346e-06,
+      "loss": 1.1039,
+      "step": 603
+    },
+    {
+      "epoch": 0.7365853658536585,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.634146341463415e-06,
+      "loss": 1.0711,
+      "step": 604
+    },
+    {
+      "epoch": 0.7378048780487805,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.6219512195121956e-06,
+      "loss": 1.0276,
+      "step": 605
+    },
+    {
+      "epoch": 0.7390243902439024,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.609756097560976e-06,
+      "loss": 1.0611,
+      "step": 606
+    },
+    {
+      "epoch": 0.7402439024390244,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.5975609756097565e-06,
+      "loss": 1.1051,
+      "step": 607
+    },
+    {
+      "epoch": 0.7414634146341463,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.5853658536585367e-06,
+      "loss": 1.0689,
+      "step": 608
+    },
+    {
+      "epoch": 0.7426829268292683,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.5731707317073174e-06,
+      "loss": 1.0563,
+      "step": 609
+    },
+    {
+      "epoch": 0.7439024390243902,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.5609756097560977e-06,
+      "loss": 1.0239,
+      "step": 610
+    },
+    {
+      "epoch": 0.7451219512195122,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.5487804878048783e-06,
+      "loss": 1.1052,
+      "step": 611
+    },
+    {
+      "epoch": 0.7463414634146341,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.5365853658536586e-06,
+      "loss": 1.1,
+      "step": 612
+    },
+    {
+      "epoch": 0.7475609756097561,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.5243902439024393e-06,
+      "loss": 1.1258,
+      "step": 613
+    },
+    {
+      "epoch": 0.748780487804878,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.5121951219512195e-06,
+      "loss": 1.1154,
+      "step": 614
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.5e-06,
+      "loss": 1.0448,
+      "step": 615
+    },
+    {
+      "epoch": 0.751219512195122,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.487804878048781e-06,
+      "loss": 1.0559,
+      "step": 616
+    },
+    {
+      "epoch": 0.7524390243902439,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.475609756097561e-06,
+      "loss": 1.0948,
+      "step": 617
+    },
+    {
+      "epoch": 0.7536585365853659,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.463414634146342e-06,
+      "loss": 1.0541,
+      "step": 618
+    },
+    {
+      "epoch": 0.7548780487804878,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.451219512195122e-06,
+      "loss": 1.0377,
+      "step": 619
+    },
+    {
+      "epoch": 0.7560975609756098,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.4390243902439027e-06,
+      "loss": 1.0511,
+      "step": 620
+    },
+    {
+      "epoch": 0.7573170731707317,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.426829268292683e-06,
+      "loss": 1.1331,
+      "step": 621
+    },
+    {
+      "epoch": 0.7585365853658537,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.4146341463414637e-06,
+      "loss": 1.0521,
+      "step": 622
+    },
+    {
+      "epoch": 0.7597560975609756,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.402439024390244e-06,
+      "loss": 1.0358,
+      "step": 623
+    },
+    {
+      "epoch": 0.7609756097560976,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.3902439024390246e-06,
+      "loss": 1.1193,
+      "step": 624
+    },
+    {
+      "epoch": 0.7621951219512195,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.378048780487805e-06,
+      "loss": 1.0871,
+      "step": 625
+    },
+    {
+      "epoch": 0.7634146341463415,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.3658536585365855e-06,
+      "loss": 1.1028,
+      "step": 626
+    },
+    {
+      "epoch": 0.7646341463414634,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.3536585365853658e-06,
+      "loss": 1.1165,
+      "step": 627
+    },
+    {
+      "epoch": 0.7658536585365854,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.3414634146341465e-06,
+      "loss": 1.0831,
+      "step": 628
+    },
+    {
+      "epoch": 0.7670731707317073,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.329268292682927e-06,
+      "loss": 1.0924,
+      "step": 629
+    },
+    {
+      "epoch": 0.7682926829268293,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.317073170731708e-06,
+      "loss": 1.1502,
+      "step": 630
+    },
+    {
+      "epoch": 0.7695121951219512,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.304878048780488e-06,
+      "loss": 1.059,
+      "step": 631
+    },
+    {
+      "epoch": 0.7707317073170732,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.2926829268292687e-06,
+      "loss": 1.0582,
+      "step": 632
+    },
+    {
+      "epoch": 0.7719512195121951,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.280487804878049e-06,
+      "loss": 1.0578,
+      "step": 633
+    },
+    {
+      "epoch": 0.7731707317073171,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.2682926829268297e-06,
+      "loss": 1.027,
+      "step": 634
+    },
+    {
+      "epoch": 0.774390243902439,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.25609756097561e-06,
+      "loss": 1.1362,
+      "step": 635
+    },
+    {
+      "epoch": 0.775609756097561,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.2439024390243906e-06,
+      "loss": 1.0586,
+      "step": 636
+    },
+    {
+      "epoch": 0.776829268292683,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.231707317073171e-06,
+      "loss": 1.0916,
+      "step": 637
+    },
+    {
+      "epoch": 0.7780487804878049,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.2195121951219515e-06,
+      "loss": 1.0785,
+      "step": 638
+    },
+    {
+      "epoch": 0.7792682926829269,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.207317073170732e-06,
+      "loss": 1.0886,
+      "step": 639
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.1951219512195125e-06,
+      "loss": 1.063,
+      "step": 640
+    },
+    {
+      "epoch": 0.7817073170731708,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.1829268292682927e-06,
+      "loss": 1.0256,
+      "step": 641
+    },
+    {
+      "epoch": 0.7829268292682927,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.1707317073170734e-06,
+      "loss": 1.0759,
+      "step": 642
+    },
+    {
+      "epoch": 0.7841463414634147,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.1585365853658537e-06,
+      "loss": 1.0809,
+      "step": 643
+    },
+    {
+      "epoch": 0.7853658536585366,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.1463414634146343e-06,
+      "loss": 1.0953,
+      "step": 644
+    },
+    {
+      "epoch": 0.7865853658536586,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.1341463414634146e-06,
+      "loss": 1.0791,
+      "step": 645
+    },
+    {
+      "epoch": 0.7878048780487805,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.1219512195121953e-06,
+      "loss": 1.1424,
+      "step": 646
+    },
+    {
+      "epoch": 0.7890243902439025,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.1097560975609755e-06,
+      "loss": 1.0963,
+      "step": 647
+    },
+    {
+      "epoch": 0.7902439024390244,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.097560975609756e-06,
+      "loss": 1.0406,
+      "step": 648
+    },
+    {
+      "epoch": 0.7914634146341464,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.085365853658537e-06,
+      "loss": 1.0071,
+      "step": 649
+    },
+    {
+      "epoch": 0.7926829268292683,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.073170731707317e-06,
+      "loss": 1.0457,
+      "step": 650
+    },
+    {
+      "epoch": 0.7939024390243903,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.060975609756098e-06,
+      "loss": 1.1034,
+      "step": 651
+    },
+    {
+      "epoch": 0.7951219512195122,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.048780487804878e-06,
+      "loss": 1.1096,
+      "step": 652
+    },
+    {
+      "epoch": 0.7963414634146342,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.0365853658536587e-06,
+      "loss": 1.0638,
+      "step": 653
+    },
+    {
+      "epoch": 0.7975609756097561,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.0243902439024394e-06,
+      "loss": 1.0905,
+      "step": 654
+    },
+    {
+      "epoch": 0.7987804878048781,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.0121951219512197e-06,
+      "loss": 1.0643,
+      "step": 655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.1126,
+      "step": 656
+    },
+    {
+      "epoch": 0.801219512195122,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.9878048780487806e-06,
+      "loss": 1.0715,
+      "step": 657
+    },
+    {
+      "epoch": 0.802439024390244,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.9756097560975613e-06,
+      "loss": 1.0718,
+      "step": 658
+    },
+    {
+      "epoch": 0.8036585365853659,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.9634146341463415e-06,
+      "loss": 1.0326,
+      "step": 659
+    },
+    {
+      "epoch": 0.8048780487804879,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.0687,
+      "step": 660
+    },
+    {
+      "epoch": 0.8060975609756098,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.9390243902439024e-06,
+      "loss": 1.027,
+      "step": 661
+    },
+    {
+      "epoch": 0.8073170731707318,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.926829268292683e-06,
+      "loss": 1.1043,
+      "step": 662
+    },
+    {
+      "epoch": 0.8085365853658537,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.914634146341464e-06,
+      "loss": 1.061,
+      "step": 663
+    },
+    {
+      "epoch": 0.8097560975609757,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.902439024390244e-06,
+      "loss": 1.0429,
+      "step": 664
+    },
+    {
+      "epoch": 0.8109756097560976,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.8902439024390245e-06,
+      "loss": 1.0563,
+      "step": 665
+    },
+    {
+      "epoch": 0.8121951219512196,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.878048780487805e-06,
+      "loss": 1.0664,
+      "step": 666
+    },
+    {
+      "epoch": 0.8134146341463414,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.8658536585365854e-06,
+      "loss": 1.022,
+      "step": 667
+    },
+    {
+      "epoch": 0.8146341463414634,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.853658536585366e-06,
+      "loss": 1.0534,
+      "step": 668
+    },
+    {
+      "epoch": 0.8158536585365853,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.8414634146341464e-06,
+      "loss": 1.1069,
+      "step": 669
+    },
+    {
+      "epoch": 0.8170731707317073,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.8292682926829268e-06,
+      "loss": 1.0231,
+      "step": 670
+    },
+    {
+      "epoch": 0.8182926829268292,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.8170731707317073e-06,
+      "loss": 1.0328,
+      "step": 671
+    },
+    {
+      "epoch": 0.8195121951219512,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.804878048780488e-06,
+      "loss": 1.0963,
+      "step": 672
+    },
+    {
+      "epoch": 0.8207317073170731,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.7926829268292685e-06,
+      "loss": 1.0811,
+      "step": 673
+    },
+    {
+      "epoch": 0.8219512195121951,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.780487804878049e-06,
+      "loss": 1.032,
+      "step": 674
+    },
+    {
+      "epoch": 0.823170731707317,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.7682926829268294e-06,
+      "loss": 1.1036,
+      "step": 675
+    },
+    {
+      "epoch": 0.824390243902439,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.7560975609756098e-06,
+      "loss": 1.0545,
+      "step": 676
+    },
+    {
+      "epoch": 0.8256097560975609,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.7439024390243903e-06,
+      "loss": 1.0845,
+      "step": 677
+    },
+    {
+      "epoch": 0.8268292682926829,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.7317073170731708e-06,
+      "loss": 1.088,
+      "step": 678
+    },
+    {
+      "epoch": 0.8280487804878048,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.7195121951219512e-06,
+      "loss": 1.0467,
+      "step": 679
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.707317073170732e-06,
+      "loss": 0.9774,
+      "step": 680
+    },
+    {
+      "epoch": 0.8304878048780487,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.6951219512195124e-06,
+      "loss": 1.0679,
+      "step": 681
+    },
+    {
+      "epoch": 0.8317073170731707,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.6829268292682928e-06,
+      "loss": 1.0333,
+      "step": 682
+    },
+    {
+      "epoch": 0.8329268292682926,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.6707317073170733e-06,
+      "loss": 1.0098,
+      "step": 683
+    },
+    {
+      "epoch": 0.8341463414634146,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.6585365853658538e-06,
+      "loss": 1.1388,
+      "step": 684
+    },
+    {
+      "epoch": 0.8353658536585366,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.6463414634146345e-06,
+      "loss": 1.0618,
+      "step": 685
+    },
+    {
+      "epoch": 0.8365853658536585,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.634146341463415e-06,
+      "loss": 1.076,
+      "step": 686
+    },
+    {
+      "epoch": 0.8378048780487805,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.6219512195121954e-06,
+      "loss": 1.0945,
+      "step": 687
+    },
+    {
+      "epoch": 0.8390243902439024,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.6097560975609759e-06,
+      "loss": 1.0487,
+      "step": 688
+    },
+    {
+      "epoch": 0.8402439024390244,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.5975609756097563e-06,
+      "loss": 1.1462,
+      "step": 689
+    },
+    {
+      "epoch": 0.8414634146341463,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.5853658536585368e-06,
+      "loss": 1.0855,
+      "step": 690
+    },
+    {
+      "epoch": 0.8426829268292683,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.5731707317073172e-06,
+      "loss": 1.0957,
+      "step": 691
+    },
+    {
+      "epoch": 0.8439024390243902,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.5609756097560977e-06,
+      "loss": 1.0581,
+      "step": 692
+    },
+    {
+      "epoch": 0.8451219512195122,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.5487804878048782e-06,
+      "loss": 1.0793,
+      "step": 693
+    },
+    {
+      "epoch": 0.8463414634146341,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.5365853658536586e-06,
+      "loss": 1.0879,
+      "step": 694
+    },
+    {
+      "epoch": 0.8475609756097561,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.5243902439024391e-06,
+      "loss": 1.0585,
+      "step": 695
+    },
+    {
+      "epoch": 0.848780487804878,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.5121951219512196e-06,
+      "loss": 1.0186,
+      "step": 696
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.5e-06,
+      "loss": 1.1329,
+      "step": 697
+    },
+    {
+      "epoch": 0.8512195121951219,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.4878048780487805e-06,
+      "loss": 1.1513,
+      "step": 698
+    },
+    {
+      "epoch": 0.8524390243902439,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.475609756097561e-06,
+      "loss": 1.1021,
+      "step": 699
+    },
+    {
+      "epoch": 0.8536585365853658,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.4634146341463414e-06,
+      "loss": 1.1114,
+      "step": 700
+    },
+    {
+      "epoch": 0.8548780487804878,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.451219512195122e-06,
+      "loss": 1.0979,
+      "step": 701
+    },
+    {
+      "epoch": 0.8560975609756097,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.4390243902439024e-06,
+      "loss": 1.0381,
+      "step": 702
+    },
+    {
+      "epoch": 0.8573170731707317,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.4268292682926828e-06,
+      "loss": 1.0386,
+      "step": 703
+    },
+    {
+      "epoch": 0.8585365853658536,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.4146341463414633e-06,
+      "loss": 1.0427,
+      "step": 704
+    },
+    {
+      "epoch": 0.8597560975609756,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.4024390243902442e-06,
+      "loss": 1.0731,
+      "step": 705
+    },
+    {
+      "epoch": 0.8609756097560975,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.3902439024390246e-06,
+      "loss": 1.135,
+      "step": 706
+    },
+    {
+      "epoch": 0.8621951219512195,
+      "grad_norm": 0.25,
+      "learning_rate": 1.3780487804878051e-06,
+      "loss": 1.0106,
+      "step": 707
+    },
+    {
+      "epoch": 0.8634146341463415,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.3658536585365856e-06,
+      "loss": 1.0915,
+      "step": 708
+    },
+    {
+      "epoch": 0.8646341463414634,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.353658536585366e-06,
+      "loss": 1.0352,
+      "step": 709
+    },
+    {
+      "epoch": 0.8658536585365854,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.3414634146341465e-06,
+      "loss": 1.0295,
+      "step": 710
+    },
+    {
+      "epoch": 0.8670731707317073,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.329268292682927e-06,
+      "loss": 1.0741,
+      "step": 711
+    },
+    {
+      "epoch": 0.8682926829268293,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.3170731707317074e-06,
+      "loss": 1.0662,
+      "step": 712
+    },
+    {
+      "epoch": 0.8695121951219512,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.304878048780488e-06,
+      "loss": 1.0697,
+      "step": 713
+    },
+    {
+      "epoch": 0.8707317073170732,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.2926829268292684e-06,
+      "loss": 0.9954,
+      "step": 714
+    },
+    {
+      "epoch": 0.8719512195121951,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.2804878048780488e-06,
+      "loss": 1.0151,
+      "step": 715
+    },
+    {
+      "epoch": 0.8731707317073171,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.2682926829268293e-06,
+      "loss": 0.9893,
+      "step": 716
+    },
+    {
+      "epoch": 0.874390243902439,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.2560975609756098e-06,
+      "loss": 1.0561,
+      "step": 717
+    },
+    {
+      "epoch": 0.875609756097561,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.2439024390243904e-06,
+      "loss": 1.1524,
+      "step": 718
+    },
+    {
+      "epoch": 0.8768292682926829,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.231707317073171e-06,
+      "loss": 1.054,
+      "step": 719
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.2195121951219514e-06,
+      "loss": 1.1322,
+      "step": 720
+    },
+    {
+      "epoch": 0.8792682926829268,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.2073170731707318e-06,
+      "loss": 1.0766,
+      "step": 721
+    },
+    {
+      "epoch": 0.8804878048780488,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.1951219512195123e-06,
+      "loss": 1.0509,
+      "step": 722
+    },
+    {
+      "epoch": 0.8817073170731707,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.1829268292682928e-06,
+      "loss": 1.0303,
+      "step": 723
+    },
+    {
+      "epoch": 0.8829268292682927,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.1707317073170732e-06,
+      "loss": 1.065,
+      "step": 724
+    },
+    {
+      "epoch": 0.8841463414634146,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.158536585365854e-06,
+      "loss": 1.0904,
+      "step": 725
+    },
+    {
+      "epoch": 0.8853658536585366,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.1463414634146344e-06,
+      "loss": 1.0988,
+      "step": 726
+    },
+    {
+      "epoch": 0.8865853658536585,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.1341463414634148e-06,
+      "loss": 1.0732,
+      "step": 727
+    },
+    {
+      "epoch": 0.8878048780487805,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.1219512195121953e-06,
+      "loss": 1.0844,
+      "step": 728
+    },
+    {
+      "epoch": 0.8890243902439025,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.1097560975609758e-06,
+      "loss": 1.0841,
+      "step": 729
+    },
+    {
+      "epoch": 0.8902439024390244,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.0975609756097562e-06,
+      "loss": 1.1141,
+      "step": 730
+    },
+    {
+      "epoch": 0.8914634146341464,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.0853658536585367e-06,
+      "loss": 1.0769,
+      "step": 731
+    },
+    {
+      "epoch": 0.8926829268292683,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.0731707317073172e-06,
+      "loss": 1.0601,
+      "step": 732
+    },
+    {
+      "epoch": 0.8939024390243903,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.0609756097560976e-06,
+      "loss": 1.0131,
+      "step": 733
+    },
+    {
+      "epoch": 0.8951219512195122,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.048780487804878e-06,
+      "loss": 1.1302,
+      "step": 734
+    },
+    {
+      "epoch": 0.8963414634146342,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.0365853658536586e-06,
+      "loss": 1.0548,
+      "step": 735
+    },
+    {
+      "epoch": 0.8975609756097561,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.024390243902439e-06,
+      "loss": 1.052,
+      "step": 736
+    },
+    {
+      "epoch": 0.8987804878048781,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.0121951219512197e-06,
+      "loss": 1.0466,
+      "step": 737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.0713,
+      "step": 738
+    },
+    {
+      "epoch": 0.901219512195122,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.878048780487806e-07,
+      "loss": 1.101,
+      "step": 739
+    },
+    {
+      "epoch": 0.9024390243902439,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.0413,
+      "step": 740
+    },
+    {
+      "epoch": 0.9036585365853659,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.634146341463416e-07,
+      "loss": 1.0669,
+      "step": 741
+    },
+    {
+      "epoch": 0.9048780487804878,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.51219512195122e-07,
+      "loss": 1.0594,
+      "step": 742
+    },
+    {
+      "epoch": 0.9060975609756098,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.390243902439025e-07,
+      "loss": 1.0646,
+      "step": 743
+    },
+    {
+      "epoch": 0.9073170731707317,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.26829268292683e-07,
+      "loss": 1.0803,
+      "step": 744
+    },
+    {
+      "epoch": 0.9085365853658537,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.146341463414634e-07,
+      "loss": 1.1442,
+      "step": 745
+    },
+    {
+      "epoch": 0.9097560975609756,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.02439024390244e-07,
+      "loss": 1.079,
+      "step": 746
+    },
+    {
+      "epoch": 0.9109756097560976,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.902439024390245e-07,
+      "loss": 1.0908,
+      "step": 747
+    },
+    {
+      "epoch": 0.9121951219512195,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.780487804878049e-07,
+      "loss": 1.0497,
+      "step": 748
+    },
+    {
+      "epoch": 0.9134146341463415,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.658536585365854e-07,
+      "loss": 1.0826,
+      "step": 749
+    },
+    {
+      "epoch": 0.9146341463414634,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.53658536585366e-07,
+      "loss": 1.1029,
+      "step": 750
+    },
+    {
+      "epoch": 0.9158536585365854,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.414634146341464e-07,
+      "loss": 1.1093,
+      "step": 751
+    },
+    {
+      "epoch": 0.9170731707317074,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.292682926829269e-07,
+      "loss": 1.0173,
+      "step": 752
+    },
+    {
+      "epoch": 0.9182926829268293,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.170731707317075e-07,
+      "loss": 1.0709,
+      "step": 753
+    },
+    {
+      "epoch": 0.9195121951219513,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.048780487804879e-07,
+      "loss": 1.1359,
+      "step": 754
+    },
+    {
+      "epoch": 0.9207317073170732,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.926829268292684e-07,
+      "loss": 1.0086,
+      "step": 755
+    },
+    {
+      "epoch": 0.9219512195121952,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.804878048780489e-07,
+      "loss": 1.0692,
+      "step": 756
+    },
+    {
+      "epoch": 0.9231707317073171,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.682926829268293e-07,
+      "loss": 1.0772,
+      "step": 757
+    },
+    {
+      "epoch": 0.9243902439024391,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.560975609756098e-07,
+      "loss": 1.1352,
+      "step": 758
+    },
+    {
+      "epoch": 0.925609756097561,
+      "grad_norm": 0.205078125,
+      "learning_rate": 7.439024390243903e-07,
+      "loss": 1.0776,
+      "step": 759
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.317073170731707e-07,
+      "loss": 1.0599,
+      "step": 760
+    },
+    {
+      "epoch": 0.9280487804878049,
+      "grad_norm": 0.240234375,
+      "learning_rate": 7.195121951219512e-07,
+      "loss": 1.1065,
+      "step": 761
+    },
+    {
+      "epoch": 0.9292682926829269,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.073170731707316e-07,
+      "loss": 1.1051,
+      "step": 762
+    },
+    {
+      "epoch": 0.9304878048780488,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.951219512195123e-07,
+      "loss": 1.083,
+      "step": 763
+    },
+    {
+      "epoch": 0.9317073170731708,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.829268292682928e-07,
+      "loss": 1.0656,
+      "step": 764
+    },
+    {
+      "epoch": 0.9329268292682927,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.707317073170733e-07,
+      "loss": 1.0788,
+      "step": 765
+    },
+    {
+      "epoch": 0.9341463414634147,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6.585365853658537e-07,
+      "loss": 1.0992,
+      "step": 766
+    },
+    {
+      "epoch": 0.9353658536585366,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.463414634146342e-07,
+      "loss": 1.0471,
+      "step": 767
+    },
+    {
+      "epoch": 0.9365853658536586,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.341463414634146e-07,
+      "loss": 1.078,
+      "step": 768
+    },
+    {
+      "epoch": 0.9378048780487804,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.219512195121952e-07,
+      "loss": 1.0927,
+      "step": 769
+    },
+    {
+      "epoch": 0.9390243902439024,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.097560975609757e-07,
+      "loss": 1.0055,
+      "step": 770
+    },
+    {
+      "epoch": 0.9402439024390243,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.975609756097562e-07,
+      "loss": 1.1183,
+      "step": 771
+    },
+    {
+      "epoch": 0.9414634146341463,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.853658536585366e-07,
+      "loss": 1.0887,
+      "step": 772
+    },
+    {
+      "epoch": 0.9426829268292682,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.731707317073172e-07,
+      "loss": 1.0923,
+      "step": 773
+    },
+    {
+      "epoch": 0.9439024390243902,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.609756097560977e-07,
+      "loss": 1.0986,
+      "step": 774
+    },
+    {
+      "epoch": 0.9451219512195121,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.487804878048781e-07,
+      "loss": 1.0457,
+      "step": 775
+    },
+    {
+      "epoch": 0.9463414634146341,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.365853658536586e-07,
+      "loss": 1.0444,
+      "step": 776
+    },
+    {
+      "epoch": 0.947560975609756,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.24390243902439e-07,
+      "loss": 1.0999,
+      "step": 777
+    },
+    {
+      "epoch": 0.948780487804878,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.121951219512195e-07,
+      "loss": 1.0839,
+      "step": 778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.0789,
+      "step": 779
+    },
+    {
+      "epoch": 0.9512195121951219,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.878048780487805e-07,
+      "loss": 1.0361,
+      "step": 780
+    },
+    {
+      "epoch": 0.9524390243902439,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.75609756097561e-07,
+      "loss": 1.037,
+      "step": 781
+    },
+    {
+      "epoch": 0.9536585365853658,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.634146341463415e-07,
+      "loss": 1.1276,
+      "step": 782
+    },
+    {
+      "epoch": 0.9548780487804878,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.51219512195122e-07,
+      "loss": 1.0421,
+      "step": 783
+    },
+    {
+      "epoch": 0.9560975609756097,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.3902439024390246e-07,
+      "loss": 1.0561,
+      "step": 784
+    },
+    {
+      "epoch": 0.9573170731707317,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.26829268292683e-07,
+      "loss": 1.0926,
+      "step": 785
+    },
+    {
+      "epoch": 0.9585365853658536,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.1463414634146344e-07,
+      "loss": 1.1011,
+      "step": 786
+    },
+    {
+      "epoch": 0.9597560975609756,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.0243902439024396e-07,
+      "loss": 1.0871,
+      "step": 787
+    },
+    {
+      "epoch": 0.9609756097560975,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.9024390243902443e-07,
+      "loss": 1.0532,
+      "step": 788
+    },
+    {
+      "epoch": 0.9621951219512195,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.780487804878049e-07,
+      "loss": 1.1131,
+      "step": 789
+    },
+    {
+      "epoch": 0.9634146341463414,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.6585365853658536e-07,
+      "loss": 1.094,
+      "step": 790
+    },
+    {
+      "epoch": 0.9646341463414634,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.536585365853658e-07,
+      "loss": 1.0334,
+      "step": 791
+    },
+    {
+      "epoch": 0.9658536585365853,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.414634146341464e-07,
+      "loss": 1.0802,
+      "step": 792
+    },
+    {
+      "epoch": 0.9670731707317073,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.2926829268292686e-07,
+      "loss": 1.0745,
+      "step": 793
+    },
+    {
+      "epoch": 0.9682926829268292,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.170731707317073e-07,
+      "loss": 1.0459,
+      "step": 794
+    },
+    {
+      "epoch": 0.9695121951219512,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.0487804878048784e-07,
+      "loss": 1.0958,
+      "step": 795
+    },
+    {
+      "epoch": 0.9707317073170731,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.926829268292683e-07,
+      "loss": 1.0859,
+      "step": 796
+    },
+    {
+      "epoch": 0.9719512195121951,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.804878048780488e-07,
+      "loss": 1.1665,
+      "step": 797
+    },
+    {
+      "epoch": 0.973170731707317,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.682926829268293e-07,
+      "loss": 1.0844,
+      "step": 798
+    },
+    {
+      "epoch": 0.974390243902439,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.5609756097560976e-07,
+      "loss": 1.1316,
+      "step": 799
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.439024390243903e-07,
+      "loss": 0.9933,
+      "step": 800
+    },
+    {
+      "epoch": 0.9768292682926829,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.3170731707317074e-07,
+      "loss": 1.0578,
+      "step": 801
+    },
+    {
+      "epoch": 0.9780487804878049,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.1951219512195123e-07,
+      "loss": 1.1175,
+      "step": 802
+    },
+    {
+      "epoch": 0.9792682926829268,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.0731707317073172e-07,
+      "loss": 1.0502,
+      "step": 803
+    },
+    {
+      "epoch": 0.9804878048780488,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.9512195121951221e-07,
+      "loss": 1.0945,
+      "step": 804
+    },
+    {
+      "epoch": 0.9817073170731707,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.8292682926829268e-07,
+      "loss": 1.0537,
+      "step": 805
+    },
+    {
+      "epoch": 0.9829268292682927,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.707317073170732e-07,
+      "loss": 1.1431,
+      "step": 806
+    },
+    {
+      "epoch": 0.9841463414634146,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.5853658536585366e-07,
+      "loss": 1.1126,
+      "step": 807
+    },
+    {
+      "epoch": 0.9853658536585366,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.4634146341463415e-07,
+      "loss": 1.078,
+      "step": 808
+    },
+    {
+      "epoch": 0.9865853658536585,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.3414634146341465e-07,
+      "loss": 1.0876,
+      "step": 809
+    },
+    {
+      "epoch": 0.9878048780487805,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.2195121951219514e-07,
+      "loss": 1.0637,
+      "step": 810
+    },
+    {
+      "epoch": 0.9890243902439024,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.0975609756097562e-07,
+      "loss": 1.0522,
+      "step": 811
+    },
+    {
+      "epoch": 0.9902439024390244,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.756097560975611e-08,
+      "loss": 1.0683,
+      "step": 812
+    },
+    {
+      "epoch": 0.9914634146341463,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.53658536585366e-08,
+      "loss": 1.093,
+      "step": 813
+    },
+    {
+      "epoch": 0.9926829268292683,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.317073170731708e-08,
+      "loss": 1.1015,
+      "step": 814
+    },
+    {
+      "epoch": 0.9939024390243902,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.097560975609757e-08,
+      "loss": 1.117,
+      "step": 815
+    },
+    {
+      "epoch": 0.9951219512195122,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.8780487804878054e-08,
+      "loss": 1.1107,
+      "step": 816
+    },
+    {
+      "epoch": 0.9963414634146341,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.658536585365854e-08,
+      "loss": 1.129,
+      "step": 817
+    },
+    {
+      "epoch": 0.9975609756097561,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.4390243902439027e-08,
+      "loss": 1.1927,
+      "step": 818
+    },
+    {
+      "epoch": 0.998780487804878,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.2195121951219513e-08,
+      "loss": 1.0641,
+      "step": 819
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0,
+      "loss": 1.1285,
+      "step": 820
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 820,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.657644303878193e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}