{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 116,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 2.0671271678531755,
      "learning_rate": 0.0,
      "loss": 0.866,
      "step": 1
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 2.2900714872410477,
      "learning_rate": 7.142857142857143e-06,
      "loss": 1.043,
      "step": 2
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 1.861548862699238,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.9408,
      "step": 3
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 1.5918887111051039,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.9754,
      "step": 4
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 2.461157666875476,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.9909,
      "step": 5
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 3.269351348810357,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.9765,
      "step": 6
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 2.891039729725391,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.976,
      "step": 7
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 1.7833187830275166,
      "learning_rate": 5e-05,
      "loss": 0.985,
      "step": 8
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 1.6508430392003424,
      "learning_rate": 4.9993025930300686e-05,
      "loss": 0.9983,
      "step": 9
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 1.3325514749958667,
      "learning_rate": 4.99721076122146e-05,
      "loss": 0.9611,
      "step": 10
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 1.1508459362776429,
      "learning_rate": 4.9937256716606394e-05,
      "loss": 0.8505,
      "step": 11
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.8112441764899212,
      "learning_rate": 4.9888492687682096e-05,
      "loss": 0.8723,
      "step": 12
    },
    {
      "epoch": 0.45614035087719296,
      "grad_norm": 0.8092302145893412,
      "learning_rate": 4.982584273214061e-05,
      "loss": 0.9396,
      "step": 13
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 0.7597090899690478,
      "learning_rate": 4.9749341803994465e-05,
      "loss": 0.8441,
      "step": 14
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.6679585780372033,
      "learning_rate": 4.965903258506806e-05,
      "loss": 0.8723,
      "step": 15
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 0.5980287018483266,
      "learning_rate": 4.955496546118439e-05,
      "loss": 0.8188,
      "step": 16
    },
    {
      "epoch": 0.5964912280701754,
      "grad_norm": 0.7845483827995335,
      "learning_rate": 4.9437198494053464e-05,
      "loss": 0.9303,
      "step": 17
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.5932466105687543,
      "learning_rate": 4.9305797388878264e-05,
      "loss": 0.8304,
      "step": 18
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5537040595976515,
      "learning_rate": 4.916083545769607e-05,
      "loss": 0.8943,
      "step": 19
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.5092981470635746,
      "learning_rate": 4.9002393578475816e-05,
      "loss": 0.8906,
      "step": 20
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 0.5487294714158888,
      "learning_rate": 4.883056014999423e-05,
      "loss": 0.8584,
      "step": 21
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 0.4922358015626094,
      "learning_rate": 4.864543104251587e-05,
      "loss": 0.8129,
      "step": 22
    },
    {
      "epoch": 0.8070175438596491,
      "grad_norm": 0.5397899665966813,
      "learning_rate": 4.8447109544304636e-05,
      "loss": 0.8779,
      "step": 23
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.5043199211637701,
      "learning_rate": 4.823570630399665e-05,
      "loss": 0.8482,
      "step": 24
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.4889597930942473,
      "learning_rate": 4.8011339268866505e-05,
      "loss": 0.8361,
      "step": 25
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.4769108041878493,
      "learning_rate": 4.7774133619021514e-05,
      "loss": 0.904,
      "step": 26
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 0.41632944677923917,
      "learning_rate": 4.752422169756048e-05,
      "loss": 0.8032,
      "step": 27
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.5107558694810208,
      "learning_rate": 4.726174293673612e-05,
      "loss": 0.8312,
      "step": 28
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5107558694810208,
      "learning_rate": 4.698684378016222e-05,
      "loss": 0.6911,
      "step": 29
    },
    {
      "epoch": 1.0350877192982457,
      "grad_norm": 0.9504413832780126,
      "learning_rate": 4.669967760110908e-05,
      "loss": 0.7434,
      "step": 30
    },
    {
      "epoch": 1.0701754385964912,
      "grad_norm": 0.5169442096859842,
      "learning_rate": 4.6400404616932505e-05,
      "loss": 0.7058,
      "step": 31
    },
    {
      "epoch": 1.1052631578947367,
      "grad_norm": 0.5229168796306622,
      "learning_rate": 4.608919179968457e-05,
      "loss": 0.7504,
      "step": 32
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 0.7632061755291258,
      "learning_rate": 4.576621278295558e-05,
      "loss": 0.7216,
      "step": 33
    },
    {
      "epoch": 1.1754385964912282,
      "grad_norm": 0.6677063053297304,
      "learning_rate": 4.5431647764999455e-05,
      "loss": 0.6885,
      "step": 34
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 0.5388255492178547,
      "learning_rate": 4.5085683408196535e-05,
      "loss": 0.7347,
      "step": 35
    },
    {
      "epoch": 1.2456140350877192,
      "grad_norm": 0.6102262771882692,
      "learning_rate": 4.4728512734909844e-05,
      "loss": 0.7483,
      "step": 36
    },
    {
      "epoch": 1.280701754385965,
      "grad_norm": 0.9784625861220957,
      "learning_rate": 4.436033501979299e-05,
      "loss": 0.7104,
      "step": 37
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 0.5182340717871661,
      "learning_rate": 4.398135567860972e-05,
      "loss": 0.6926,
      "step": 38
    },
    {
      "epoch": 1.3508771929824561,
      "grad_norm": 0.5241406807392717,
      "learning_rate": 4.3591786153627247e-05,
      "loss": 0.6423,
      "step": 39
    },
    {
      "epoch": 1.3859649122807016,
      "grad_norm": 0.5337062588294988,
      "learning_rate": 4.319184379564716e-05,
      "loss": 0.7449,
      "step": 40
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 0.58502320062393,
      "learning_rate": 4.2781751742739885e-05,
      "loss": 0.6863,
      "step": 41
    },
    {
      "epoch": 1.456140350877193,
      "grad_norm": 0.4437314579585024,
      "learning_rate": 4.2361738795750214e-05,
      "loss": 0.6421,
      "step": 42
    },
    {
      "epoch": 1.4912280701754386,
      "grad_norm": 0.4720129405227882,
      "learning_rate": 4.193203929064353e-05,
      "loss": 0.6387,
      "step": 43
    },
    {
      "epoch": 1.526315789473684,
      "grad_norm": 0.49099471978636483,
      "learning_rate": 4.1492892967763686e-05,
      "loss": 0.7621,
      "step": 44
    },
    {
      "epoch": 1.5614035087719298,
      "grad_norm": 0.4870002337399591,
      "learning_rate": 4.1044544838075794e-05,
      "loss": 0.7271,
      "step": 45
    },
    {
      "epoch": 1.5964912280701755,
      "grad_norm": 0.4300326366249684,
      "learning_rate": 4.058724504646834e-05,
      "loss": 0.6553,
      "step": 46
    },
    {
      "epoch": 1.631578947368421,
      "grad_norm": 0.4648072620566617,
      "learning_rate": 4.012124873219094e-05,
      "loss": 0.6585,
      "step": 47
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.4704588320844793,
      "learning_rate": 3.964681588650562e-05,
      "loss": 0.7066,
      "step": 48
    },
    {
      "epoch": 1.7017543859649122,
      "grad_norm": 0.4227060942964382,
      "learning_rate": 3.916421120763106e-05,
      "loss": 0.7298,
      "step": 49
    },
    {
      "epoch": 1.736842105263158,
      "grad_norm": 0.47049085759667697,
      "learning_rate": 3.867370395306068e-05,
      "loss": 0.713,
      "step": 50
    },
    {
      "epoch": 1.7719298245614035,
      "grad_norm": 0.41760981184783075,
      "learning_rate": 3.817556778933698e-05,
      "loss": 0.7196,
      "step": 51
    },
    {
      "epoch": 1.807017543859649,
      "grad_norm": 0.5161375525386056,
      "learning_rate": 3.7670080639366004e-05,
      "loss": 0.7425,
      "step": 52
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 0.43748999720890136,
      "learning_rate": 3.715752452735704e-05,
      "loss": 0.5942,
      "step": 53
    },
    {
      "epoch": 1.8771929824561404,
      "grad_norm": 0.5115131837341141,
      "learning_rate": 3.6638185421474084e-05,
      "loss": 0.7346,
      "step": 54
    },
    {
      "epoch": 1.912280701754386,
      "grad_norm": 0.461919359464356,
      "learning_rate": 3.61123530742869e-05,
      "loss": 0.6743,
      "step": 55
    },
    {
      "epoch": 1.9473684210526314,
      "grad_norm": 0.4342888779362966,
      "learning_rate": 3.5580320861110625e-05,
      "loss": 0.7281,
      "step": 56
    },
    {
      "epoch": 1.9824561403508771,
      "grad_norm": 0.48320797418001477,
      "learning_rate": 3.504238561632424e-05,
      "loss": 0.7326,
      "step": 57
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8691230999307578,
      "learning_rate": 3.4498847467759e-05,
      "loss": 0.6139,
      "step": 58
    },
    {
      "epoch": 2.0350877192982457,
      "grad_norm": 0.925588727218562,
      "learning_rate": 3.3950009669249497e-05,
      "loss": 0.5316,
      "step": 59
    },
    {
      "epoch": 2.0701754385964914,
      "grad_norm": 2.7861744302281886,
      "learning_rate": 3.339617843144057e-05,
      "loss": 0.6395,
      "step": 60
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 2.960320751174846,
      "learning_rate": 3.2837662750944535e-05,
      "loss": 0.5571,
      "step": 61
    },
    {
      "epoch": 2.1403508771929824,
      "grad_norm": 0.5858200164045075,
      "learning_rate": 3.227477423794412e-05,
      "loss": 0.5282,
      "step": 62
    },
    {
      "epoch": 2.175438596491228,
      "grad_norm": 0.7249551445260521,
      "learning_rate": 3.170782694233712e-05,
      "loss": 0.568,
      "step": 63
    },
    {
      "epoch": 2.2105263157894735,
      "grad_norm": 0.8322811685066639,
      "learning_rate": 3.1137137178519985e-05,
      "loss": 0.5588,
      "step": 64
    },
    {
      "epoch": 2.245614035087719,
      "grad_norm": 0.4688506121750911,
      "learning_rate": 3.056302334890786e-05,
      "loss": 0.4997,
      "step": 65
    },
    {
      "epoch": 2.280701754385965,
      "grad_norm": 0.5395406531134912,
      "learning_rate": 2.9985805766289817e-05,
      "loss": 0.5389,
      "step": 66
    },
    {
      "epoch": 2.3157894736842106,
      "grad_norm": 0.5972426732470659,
      "learning_rate": 2.9405806475118048e-05,
      "loss": 0.5154,
      "step": 67
    },
    {
      "epoch": 2.3508771929824563,
      "grad_norm": 0.5435761853560287,
      "learning_rate": 2.882334907183115e-05,
      "loss": 0.5696,
      "step": 68
    },
    {
      "epoch": 2.3859649122807016,
      "grad_norm": 0.4733058013404621,
      "learning_rate": 2.8238758524311314e-05,
      "loss": 0.5286,
      "step": 69
    },
    {
      "epoch": 2.4210526315789473,
      "grad_norm": 0.7320896114370156,
      "learning_rate": 2.7652360990576453e-05,
      "loss": 0.6029,
      "step": 70
    },
    {
      "epoch": 2.456140350877193,
      "grad_norm": 0.5143853256014193,
      "learning_rate": 2.7064483636808313e-05,
      "loss": 0.5373,
      "step": 71
    },
    {
      "epoch": 2.4912280701754383,
      "grad_norm": 0.47578504969519847,
      "learning_rate": 2.6475454454818073e-05,
      "loss": 0.5372,
      "step": 72
    },
    {
      "epoch": 2.526315789473684,
      "grad_norm": 0.47485068019332843,
      "learning_rate": 2.5885602079051353e-05,
      "loss": 0.4714,
      "step": 73
    },
    {
      "epoch": 2.56140350877193,
      "grad_norm": 0.5454183913380498,
      "learning_rate": 2.529525560323462e-05,
      "loss": 0.559,
      "step": 74
    },
    {
      "epoch": 2.5964912280701755,
      "grad_norm": 0.4805059890683054,
      "learning_rate": 2.470474439676539e-05,
      "loss": 0.5432,
      "step": 75
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.549573444590586,
      "learning_rate": 2.4114397920948657e-05,
      "loss": 0.544,
      "step": 76
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.44171024079549115,
      "learning_rate": 2.3524545545181933e-05,
      "loss": 0.5719,
      "step": 77
    },
    {
      "epoch": 2.7017543859649122,
      "grad_norm": 0.40375286197219823,
      "learning_rate": 2.2935516363191693e-05,
      "loss": 0.5574,
      "step": 78
    },
    {
      "epoch": 2.736842105263158,
      "grad_norm": 0.44079765151450895,
      "learning_rate": 2.2347639009423553e-05,
      "loss": 0.5473,
      "step": 79
    },
    {
      "epoch": 2.7719298245614032,
      "grad_norm": 0.3837140844283907,
      "learning_rate": 2.1761241475688695e-05,
      "loss": 0.5,
      "step": 80
    },
    {
      "epoch": 2.807017543859649,
      "grad_norm": 0.36498294953041693,
      "learning_rate": 2.117665092816885e-05,
      "loss": 0.4579,
      "step": 81
    },
    {
      "epoch": 2.8421052631578947,
      "grad_norm": 0.4176708774075213,
      "learning_rate": 2.059419352488196e-05,
      "loss": 0.5302,
      "step": 82
    },
    {
      "epoch": 2.8771929824561404,
      "grad_norm": 0.3801878678793892,
      "learning_rate": 2.0014194233710193e-05,
      "loss": 0.5338,
      "step": 83
    },
    {
      "epoch": 2.912280701754386,
      "grad_norm": 0.46768889533074776,
      "learning_rate": 1.9436976651092144e-05,
      "loss": 0.5636,
      "step": 84
    },
    {
      "epoch": 2.9473684210526314,
      "grad_norm": 0.41361908756083027,
      "learning_rate": 1.8862862821480025e-05,
      "loss": 0.4871,
      "step": 85
    },
    {
      "epoch": 2.982456140350877,
      "grad_norm": 0.35900229029204755,
      "learning_rate": 1.829217305766289e-05,
      "loss": 0.545,
      "step": 86
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.35900229029204755,
      "learning_rate": 1.7725225762055887e-05,
      "loss": 0.3114,
      "step": 87
    },
    {
      "epoch": 3.0350877192982457,
      "grad_norm": 1.039088884520627,
      "learning_rate": 1.7162337249055477e-05,
      "loss": 0.4039,
      "step": 88
    },
    {
      "epoch": 3.0701754385964914,
      "grad_norm": 1.2225625714935646,
      "learning_rate": 1.6603821568559437e-05,
      "loss": 0.4477,
      "step": 89
    },
    {
      "epoch": 3.1052631578947367,
      "grad_norm": 0.517211355092857,
      "learning_rate": 1.604999033075051e-05,
      "loss": 0.4303,
      "step": 90
    },
    {
      "epoch": 3.1403508771929824,
      "grad_norm": 1.2522323591078175,
      "learning_rate": 1.5501152532241005e-05,
      "loss": 0.3948,
      "step": 91
    },
    {
      "epoch": 3.175438596491228,
      "grad_norm": 1.4369568509391157,
      "learning_rate": 1.495761438367577e-05,
      "loss": 0.3998,
      "step": 92
    },
    {
      "epoch": 3.2105263157894735,
      "grad_norm": 1.080289936134096,
      "learning_rate": 1.4419679138889378e-05,
      "loss": 0.4489,
      "step": 93
    },
    {
      "epoch": 3.245614035087719,
      "grad_norm": 0.3959735616331335,
      "learning_rate": 1.3887646925713116e-05,
      "loss": 0.3701,
      "step": 94
    },
    {
      "epoch": 3.280701754385965,
      "grad_norm": 0.6536961872630214,
      "learning_rate": 1.3361814578525922e-05,
      "loss": 0.3669,
      "step": 95
    },
    {
      "epoch": 3.3157894736842106,
      "grad_norm": 0.9236561795720727,
      "learning_rate": 1.2842475472642968e-05,
      "loss": 0.4101,
      "step": 96
    },
    {
      "epoch": 3.3508771929824563,
      "grad_norm": 0.624339506754103,
      "learning_rate": 1.2329919360634002e-05,
      "loss": 0.3797,
      "step": 97
    },
    {
      "epoch": 3.3859649122807016,
      "grad_norm": 0.4687946351643177,
      "learning_rate": 1.182443221066303e-05,
      "loss": 0.3414,
      "step": 98
    },
    {
      "epoch": 3.4210526315789473,
      "grad_norm": 0.47331039977421663,
      "learning_rate": 1.1326296046939333e-05,
      "loss": 0.3378,
      "step": 99
    },
    {
      "epoch": 3.456140350877193,
      "grad_norm": 0.4780339258304509,
      "learning_rate": 1.083578879236895e-05,
      "loss": 0.3148,
      "step": 100
    },
    {
      "epoch": 3.4912280701754383,
      "grad_norm": 0.472991860964808,
      "learning_rate": 1.0353184113494386e-05,
      "loss": 0.362,
      "step": 101
    },
    {
      "epoch": 3.526315789473684,
      "grad_norm": 0.5473896318173218,
      "learning_rate": 9.878751267809069e-06,
      "loss": 0.366,
      "step": 102
    },
    {
      "epoch": 3.56140350877193,
      "grad_norm": 0.49421552981457206,
      "learning_rate": 9.412754953531663e-06,
      "loss": 0.3753,
      "step": 103
    },
    {
      "epoch": 3.5964912280701755,
      "grad_norm": 0.41594310192216705,
      "learning_rate": 8.955455161924217e-06,
      "loss": 0.3392,
      "step": 104
    },
    {
      "epoch": 3.6315789473684212,
      "grad_norm": 0.3728949229584135,
      "learning_rate": 8.507107032236322e-06,
      "loss": 0.3265,
      "step": 105
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.4105187771371921,
      "learning_rate": 8.067960709356478e-06,
      "loss": 0.3411,
      "step": 106
    },
    {
      "epoch": 3.7017543859649122,
      "grad_norm": 0.4016073311581445,
      "learning_rate": 7.638261204249784e-06,
      "loss": 0.3637,
      "step": 107
    },
    {
      "epoch": 3.736842105263158,
      "grad_norm": 0.5286770481220647,
      "learning_rate": 7.218248257260127e-06,
      "loss": 0.4157,
      "step": 108
    },
    {
      "epoch": 3.7719298245614032,
      "grad_norm": 0.3890055642916449,
      "learning_rate": 6.8081562043528445e-06,
      "loss": 0.4434,
      "step": 109
    },
    {
      "epoch": 3.807017543859649,
      "grad_norm": 0.38635140940193147,
      "learning_rate": 6.40821384637276e-06,
      "loss": 0.367,
      "step": 110
    },
    {
      "epoch": 3.8421052631578947,
      "grad_norm": 0.37252947829119043,
      "learning_rate": 6.018644321390288e-06,
      "loss": 0.3714,
      "step": 111
    },
    {
      "epoch": 3.8771929824561404,
      "grad_norm": 0.3916205714633107,
      "learning_rate": 5.639664980207024e-06,
      "loss": 0.333,
      "step": 112
    },
    {
      "epoch": 3.912280701754386,
      "grad_norm": 0.39308963657720947,
      "learning_rate": 5.271487265090163e-06,
      "loss": 0.3425,
      "step": 113
    },
    {
      "epoch": 3.9473684210526314,
      "grad_norm": 0.34722098404157115,
      "learning_rate": 4.914316591803475e-06,
      "loss": 0.3638,
      "step": 114
    },
    {
      "epoch": 3.982456140350877,
      "grad_norm": 0.33647207931419637,
      "learning_rate": 4.56835223500055e-06,
      "loss": 0.3789,
      "step": 115
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5236255764345163,
      "learning_rate": 4.23378721704443e-06,
      "loss": 0.2909,
      "step": 116
    }
  ],
  "logging_steps": 1,
  "max_steps": 140,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 52227047817216.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}