{
  "best_metric": 0.9090909090909091,
  "best_model_checkpoint": "beit-base-patch16-224-85-fold3/checkpoint-88",
  "epoch": 100.0,
  "eval_steps": 500,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6136363636363636,
      "eval_loss": 0.6424916982650757,
      "eval_runtime": 0.6968,
      "eval_samples_per_second": 63.143,
      "eval_steps_per_second": 2.87,
      "step": 2
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.6036472916603088,
      "eval_runtime": 0.7189,
      "eval_samples_per_second": 61.201,
      "eval_steps_per_second": 2.782,
      "step": 4
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7045454545454546,
      "eval_loss": 0.554934561252594,
      "eval_runtime": 0.6934,
      "eval_samples_per_second": 63.456,
      "eval_steps_per_second": 2.884,
      "step": 6
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.4878787100315094,
      "eval_runtime": 0.6937,
      "eval_samples_per_second": 63.426,
      "eval_steps_per_second": 2.883,
      "step": 8
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.919757843017578,
      "learning_rate": 2.5e-05,
      "loss": 0.5424,
      "step": 10
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7045454545454546,
      "eval_loss": 0.7873035669326782,
      "eval_runtime": 0.6703,
      "eval_samples_per_second": 65.639,
      "eval_steps_per_second": 2.984,
      "step": 10
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7045454545454546,
      "eval_loss": 0.681670606136322,
      "eval_runtime": 0.6871,
      "eval_samples_per_second": 64.039,
      "eval_steps_per_second": 2.911,
      "step": 12
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.48462867736816406,
      "eval_runtime": 0.6748,
      "eval_samples_per_second": 65.2,
      "eval_steps_per_second": 2.964,
      "step": 14
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.5265792608261108,
      "eval_runtime": 0.6828,
      "eval_samples_per_second": 64.442,
      "eval_steps_per_second": 2.929,
      "step": 16
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.4486856162548065,
      "eval_runtime": 0.6788,
      "eval_samples_per_second": 64.825,
      "eval_steps_per_second": 2.947,
      "step": 18
    },
    {
      "epoch": 10.0,
      "grad_norm": 15.24261474609375,
      "learning_rate": 5e-05,
      "loss": 0.5207,
      "step": 20
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.3767998516559601,
      "eval_runtime": 0.678,
      "eval_samples_per_second": 64.901,
      "eval_steps_per_second": 2.95,
      "step": 20
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.7045454545454546,
      "eval_loss": 0.676165759563446,
      "eval_runtime": 0.6783,
      "eval_samples_per_second": 64.868,
      "eval_steps_per_second": 2.949,
      "step": 22
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.3988308012485504,
      "eval_runtime": 0.6822,
      "eval_samples_per_second": 64.498,
      "eval_steps_per_second": 2.932,
      "step": 24
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.3812059462070465,
      "eval_runtime": 0.674,
      "eval_samples_per_second": 65.286,
      "eval_steps_per_second": 2.968,
      "step": 26
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.42074936628341675,
      "eval_runtime": 0.6764,
      "eval_samples_per_second": 65.047,
      "eval_steps_per_second": 2.957,
      "step": 28
    },
    {
      "epoch": 15.0,
      "grad_norm": 5.875618934631348,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.386,
      "step": 30
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.38010966777801514,
      "eval_runtime": 0.6747,
      "eval_samples_per_second": 65.215,
      "eval_steps_per_second": 2.964,
      "step": 30
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.7045454545454546,
      "eval_loss": 0.5402179956436157,
      "eval_runtime": 0.6841,
      "eval_samples_per_second": 64.322,
      "eval_steps_per_second": 2.924,
      "step": 32
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.3865377604961395,
      "eval_runtime": 0.6786,
      "eval_samples_per_second": 64.843,
      "eval_steps_per_second": 2.947,
      "step": 34
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.36351141333580017,
      "eval_runtime": 0.6783,
      "eval_samples_per_second": 64.867,
      "eval_steps_per_second": 2.948,
      "step": 36
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.5374018549919128,
      "eval_runtime": 0.6874,
      "eval_samples_per_second": 64.012,
      "eval_steps_per_second": 2.91,
      "step": 38
    },
    {
      "epoch": 20.0,
      "grad_norm": 4.103892803192139,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.3232,
      "step": 40
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.5087771415710449,
      "eval_runtime": 0.6774,
      "eval_samples_per_second": 64.956,
      "eval_steps_per_second": 2.953,
      "step": 40
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.3506772220134735,
      "eval_runtime": 0.676,
      "eval_samples_per_second": 65.088,
      "eval_steps_per_second": 2.959,
      "step": 42
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.2994966208934784,
      "eval_runtime": 0.7055,
      "eval_samples_per_second": 62.363,
      "eval_steps_per_second": 2.835,
      "step": 44
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.3403201401233673,
      "eval_runtime": 0.6828,
      "eval_samples_per_second": 64.438,
      "eval_steps_per_second": 2.929,
      "step": 46
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.3858189880847931,
      "eval_runtime": 0.6785,
      "eval_samples_per_second": 64.85,
      "eval_steps_per_second": 2.948,
      "step": 48
    },
    {
      "epoch": 25.0,
      "grad_norm": 3.5672624111175537,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.249,
      "step": 50
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.4125795066356659,
      "eval_runtime": 0.6893,
      "eval_samples_per_second": 63.831,
      "eval_steps_per_second": 2.901,
      "step": 50
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.4907498061656952,
      "eval_runtime": 0.6875,
      "eval_samples_per_second": 64.0,
      "eval_steps_per_second": 2.909,
      "step": 52
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.37986937165260315,
      "eval_runtime": 0.6841,
      "eval_samples_per_second": 64.317,
      "eval_steps_per_second": 2.923,
      "step": 54
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.35277584195137024,
      "eval_runtime": 0.6755,
      "eval_samples_per_second": 65.134,
      "eval_steps_per_second": 2.961,
      "step": 56
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.3775491416454315,
      "eval_runtime": 0.6817,
      "eval_samples_per_second": 64.541,
      "eval_steps_per_second": 2.934,
      "step": 58
    },
    {
      "epoch": 30.0,
      "grad_norm": 4.824260234832764,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.2064,
      "step": 60
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.3520455062389374,
      "eval_runtime": 0.6863,
      "eval_samples_per_second": 64.112,
      "eval_steps_per_second": 2.914,
      "step": 60
    },
    {
      "epoch": 31.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.43969738483428955,
      "eval_runtime": 0.6825,
      "eval_samples_per_second": 64.466,
      "eval_steps_per_second": 2.93,
      "step": 62
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.4284248650074005,
      "eval_runtime": 0.696,
      "eval_samples_per_second": 63.219,
      "eval_steps_per_second": 2.874,
      "step": 64
    },
    {
      "epoch": 33.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.3833133578300476,
      "eval_runtime": 0.6824,
      "eval_samples_per_second": 64.474,
      "eval_steps_per_second": 2.931,
      "step": 66
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.3557835817337036,
      "eval_runtime": 0.6811,
      "eval_samples_per_second": 64.601,
      "eval_steps_per_second": 2.936,
      "step": 68
    },
    {
      "epoch": 35.0,
      "grad_norm": 8.933732986450195,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.2066,
      "step": 70
    },
    {
      "epoch": 35.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.4880024790763855,
      "eval_runtime": 0.6725,
      "eval_samples_per_second": 65.432,
      "eval_steps_per_second": 2.974,
      "step": 70
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.3738625645637512,
      "eval_runtime": 0.6996,
      "eval_samples_per_second": 62.895,
      "eval_steps_per_second": 2.859,
      "step": 72
    },
    {
      "epoch": 37.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.5408774614334106,
      "eval_runtime": 0.7183,
      "eval_samples_per_second": 61.254,
      "eval_steps_per_second": 2.784,
      "step": 74
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.48690366744995117,
      "eval_runtime": 0.6825,
      "eval_samples_per_second": 64.474,
      "eval_steps_per_second": 2.931,
      "step": 76
    },
    {
      "epoch": 39.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.8398366570472717,
      "eval_runtime": 0.7177,
      "eval_samples_per_second": 61.308,
      "eval_steps_per_second": 2.787,
      "step": 78
    },
    {
      "epoch": 40.0,
      "grad_norm": 6.0570878982543945,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.1776,
      "step": 80
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.7954545454545454,
      "eval_loss": 0.540990948677063,
      "eval_runtime": 0.6841,
      "eval_samples_per_second": 64.323,
      "eval_steps_per_second": 2.924,
      "step": 80
    },
    {
      "epoch": 41.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.47399088740348816,
      "eval_runtime": 0.687,
      "eval_samples_per_second": 64.045,
      "eval_steps_per_second": 2.911,
      "step": 82
    },
    {
      "epoch": 42.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.34275954961776733,
      "eval_runtime": 0.6763,
      "eval_samples_per_second": 65.06,
      "eval_steps_per_second": 2.957,
      "step": 84
    },
    {
      "epoch": 43.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.31346604228019714,
      "eval_runtime": 0.6757,
      "eval_samples_per_second": 65.114,
      "eval_steps_per_second": 2.96,
      "step": 86
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.3297279477119446,
      "eval_runtime": 0.7032,
      "eval_samples_per_second": 62.569,
      "eval_steps_per_second": 2.844,
      "step": 88
    },
    {
      "epoch": 45.0,
      "grad_norm": 3.610621213912964,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.1732,
      "step": 90
    },
    {
      "epoch": 45.0,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.39823147654533386,
      "eval_runtime": 0.6794,
      "eval_samples_per_second": 64.766,
      "eval_steps_per_second": 2.944,
      "step": 90
    },
    {
      "epoch": 46.0,
      "eval_accuracy": 0.7954545454545454,
      "eval_loss": 0.5961242318153381,
      "eval_runtime": 0.6745,
      "eval_samples_per_second": 65.236,
      "eval_steps_per_second": 2.965,
      "step": 92
    },
    {
      "epoch": 47.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.47979333996772766,
      "eval_runtime": 0.6815,
      "eval_samples_per_second": 64.566,
      "eval_steps_per_second": 2.935,
      "step": 94
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.5187280178070068,
      "eval_runtime": 0.6783,
      "eval_samples_per_second": 64.868,
      "eval_steps_per_second": 2.949,
      "step": 96
    },
    {
      "epoch": 49.0,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.44300028681755066,
      "eval_runtime": 0.6847,
      "eval_samples_per_second": 64.266,
      "eval_steps_per_second": 2.921,
      "step": 98
    },
    {
      "epoch": 50.0,
      "grad_norm": 3.13413405418396,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.1372,
      "step": 100
    },
    {
      "epoch": 50.0,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.4522327482700348,
      "eval_runtime": 0.6819,
      "eval_samples_per_second": 64.526,
      "eval_steps_per_second": 2.933,
      "step": 100
    },
    {
      "epoch": 51.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.5616609454154968,
      "eval_runtime": 0.6746,
      "eval_samples_per_second": 65.221,
      "eval_steps_per_second": 2.965,
      "step": 102
    },
    {
      "epoch": 52.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.6568149924278259,
      "eval_runtime": 0.6827,
      "eval_samples_per_second": 64.455,
      "eval_steps_per_second": 2.93,
      "step": 104
    },
    {
      "epoch": 53.0,
      "eval_accuracy": 0.7954545454545454,
      "eval_loss": 0.8140836954116821,
      "eval_runtime": 0.6799,
      "eval_samples_per_second": 64.712,
      "eval_steps_per_second": 2.941,
      "step": 106
    },
    {
      "epoch": 54.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.618947446346283,
      "eval_runtime": 0.6741,
      "eval_samples_per_second": 65.273,
      "eval_steps_per_second": 2.967,
      "step": 108
    },
    {
      "epoch": 55.0,
      "grad_norm": 2.9794626235961914,
      "learning_rate": 2.5e-05,
      "loss": 0.1305,
      "step": 110
    },
    {
      "epoch": 55.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5123546123504639,
      "eval_runtime": 0.6795,
      "eval_samples_per_second": 64.755,
      "eval_steps_per_second": 2.943,
      "step": 110
    },
    {
      "epoch": 56.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5094776153564453,
      "eval_runtime": 0.678,
      "eval_samples_per_second": 64.894,
      "eval_steps_per_second": 2.95,
      "step": 112
    },
    {
      "epoch": 57.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.4101049602031708,
      "eval_runtime": 0.6831,
      "eval_samples_per_second": 64.413,
      "eval_steps_per_second": 2.928,
      "step": 114
    },
    {
      "epoch": 58.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.771160900592804,
      "eval_runtime": 0.6871,
      "eval_samples_per_second": 64.033,
      "eval_steps_per_second": 2.911,
      "step": 116
    },
    {
      "epoch": 59.0,
      "eval_accuracy": 0.7954545454545454,
      "eval_loss": 0.5072603821754456,
      "eval_runtime": 0.6813,
      "eval_samples_per_second": 64.586,
      "eval_steps_per_second": 2.936,
      "step": 118
    },
    {
      "epoch": 60.0,
      "grad_norm": 2.9292502403259277,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.1423,
      "step": 120
    },
    {
      "epoch": 60.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.38904353976249695,
      "eval_runtime": 0.6917,
      "eval_samples_per_second": 63.612,
      "eval_steps_per_second": 2.891,
      "step": 120
    },
    {
      "epoch": 61.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.5701088309288025,
      "eval_runtime": 0.6839,
      "eval_samples_per_second": 64.338,
      "eval_steps_per_second": 2.924,
      "step": 122
    },
    {
      "epoch": 62.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.548241138458252,
      "eval_runtime": 0.6956,
      "eval_samples_per_second": 63.255,
      "eval_steps_per_second": 2.875,
      "step": 124
    },
    {
      "epoch": 63.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.550826370716095,
      "eval_runtime": 0.686,
      "eval_samples_per_second": 64.14,
      "eval_steps_per_second": 2.915,
      "step": 126
    },
    {
      "epoch": 64.0,
      "eval_accuracy": 0.7954545454545454,
      "eval_loss": 0.6588876247406006,
      "eval_runtime": 0.6815,
      "eval_samples_per_second": 64.561,
      "eval_steps_per_second": 2.935,
      "step": 128
    },
    {
      "epoch": 65.0,
      "grad_norm": 3.1001317501068115,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.13,
      "step": 130
    },
    {
      "epoch": 65.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.7183592319488525,
      "eval_runtime": 0.6735,
      "eval_samples_per_second": 65.335,
      "eval_steps_per_second": 2.97,
      "step": 130
    },
    {
      "epoch": 66.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.47017189860343933,
      "eval_runtime": 0.6759,
      "eval_samples_per_second": 65.098,
      "eval_steps_per_second": 2.959,
      "step": 132
    },
    {
      "epoch": 67.0,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.43387213349342346,
      "eval_runtime": 0.6988,
      "eval_samples_per_second": 62.962,
      "eval_steps_per_second": 2.862,
      "step": 134
    },
    {
      "epoch": 68.0,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.4462599456310272,
      "eval_runtime": 0.6774,
      "eval_samples_per_second": 64.955,
      "eval_steps_per_second": 2.952,
      "step": 136
    },
    {
      "epoch": 69.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.4887324571609497,
      "eval_runtime": 0.6809,
      "eval_samples_per_second": 64.617,
      "eval_steps_per_second": 2.937,
      "step": 138
    },
    {
      "epoch": 70.0,
      "grad_norm": 4.127828121185303,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.1232,
      "step": 140
    },
    {
      "epoch": 70.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.512123167514801,
      "eval_runtime": 0.6788,
      "eval_samples_per_second": 64.817,
      "eval_steps_per_second": 2.946,
      "step": 140
    },
    {
      "epoch": 71.0,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.4943769574165344,
      "eval_runtime": 0.6818,
      "eval_samples_per_second": 64.537,
      "eval_steps_per_second": 2.934,
      "step": 142
    },
    {
      "epoch": 72.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.5207827091217041,
      "eval_runtime": 0.6878,
      "eval_samples_per_second": 63.972,
      "eval_steps_per_second": 2.908,
      "step": 144
    },
    {
      "epoch": 73.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.6074314713478088,
      "eval_runtime": 0.682,
      "eval_samples_per_second": 64.516,
      "eval_steps_per_second": 2.933,
      "step": 146
    },
    {
      "epoch": 74.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.801326334476471,
      "eval_runtime": 0.677,
      "eval_samples_per_second": 64.988,
      "eval_steps_per_second": 2.954,
      "step": 148
    },
    {
      "epoch": 75.0,
      "grad_norm": 5.893209457397461,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.1241,
      "step": 150
    },
    {
      "epoch": 75.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.7021663188934326,
      "eval_runtime": 0.7076,
      "eval_samples_per_second": 62.178,
      "eval_steps_per_second": 2.826,
      "step": 150
    },
    {
      "epoch": 76.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.5640667676925659,
      "eval_runtime": 0.6795,
      "eval_samples_per_second": 64.757,
      "eval_steps_per_second": 2.944,
      "step": 152
    },
    {
      "epoch": 77.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.6549572944641113,
      "eval_runtime": 0.6807,
      "eval_samples_per_second": 64.639,
      "eval_steps_per_second": 2.938,
      "step": 154
    },
    {
      "epoch": 78.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.6268320083618164,
      "eval_runtime": 0.6782,
      "eval_samples_per_second": 64.879,
      "eval_steps_per_second": 2.949,
      "step": 156
    },
    {
      "epoch": 79.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.5465914011001587,
      "eval_runtime": 0.6776,
      "eval_samples_per_second": 64.934,
      "eval_steps_per_second": 2.952,
      "step": 158
    },
    {
      "epoch": 80.0,
      "grad_norm": 2.7761025428771973,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.1254,
      "step": 160
    },
    {
      "epoch": 80.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.5452571511268616,
      "eval_runtime": 0.6765,
      "eval_samples_per_second": 65.043,
      "eval_steps_per_second": 2.956,
      "step": 160
    },
    {
      "epoch": 81.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5663294792175293,
      "eval_runtime": 0.6842,
      "eval_samples_per_second": 64.313,
      "eval_steps_per_second": 2.923,
      "step": 162
    },
    {
      "epoch": 82.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.5377334952354431,
      "eval_runtime": 0.6837,
      "eval_samples_per_second": 64.352,
      "eval_steps_per_second": 2.925,
      "step": 164
    },
    {
      "epoch": 83.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.5381421446800232,
      "eval_runtime": 0.6869,
      "eval_samples_per_second": 64.059,
      "eval_steps_per_second": 2.912,
      "step": 166
    },
    {
      "epoch": 84.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5458823442459106,
      "eval_runtime": 0.6766,
      "eval_samples_per_second": 65.03,
      "eval_steps_per_second": 2.956,
      "step": 168
    },
    {
      "epoch": 85.0,
      "grad_norm": 2.693737268447876,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.1061,
      "step": 170
    },
    {
      "epoch": 85.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5490157008171082,
      "eval_runtime": 0.692,
      "eval_samples_per_second": 63.581,
      "eval_steps_per_second": 2.89,
      "step": 170
    },
    {
      "epoch": 86.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5444096922874451,
      "eval_runtime": 0.6697,
      "eval_samples_per_second": 65.704,
      "eval_steps_per_second": 2.987,
      "step": 172
    },
    {
      "epoch": 87.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.534376859664917,
      "eval_runtime": 0.6765,
      "eval_samples_per_second": 65.038,
      "eval_steps_per_second": 2.956,
      "step": 174
    },
    {
      "epoch": 88.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5250511765480042,
      "eval_runtime": 0.682,
      "eval_samples_per_second": 64.515,
      "eval_steps_per_second": 2.933,
      "step": 176
    },
    {
      "epoch": 89.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.5178171992301941,
      "eval_runtime": 0.684,
      "eval_samples_per_second": 64.33,
      "eval_steps_per_second": 2.924,
      "step": 178
    },
    {
      "epoch": 90.0,
      "grad_norm": 2.427743911743164,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.1084,
      "step": 180
    },
    {
      "epoch": 90.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.5160835981369019,
      "eval_runtime": 0.6844,
      "eval_samples_per_second": 64.293,
      "eval_steps_per_second": 2.922,
      "step": 180
    },
    {
      "epoch": 91.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.518388032913208,
      "eval_runtime": 0.6894,
      "eval_samples_per_second": 63.827,
      "eval_steps_per_second": 2.901,
      "step": 182
    },
    {
      "epoch": 92.0,
      "eval_accuracy": 0.8863636363636364,
      "eval_loss": 0.5185267925262451,
      "eval_runtime": 0.6979,
      "eval_samples_per_second": 63.042,
      "eval_steps_per_second": 2.866,
      "step": 184
    },
    {
      "epoch": 93.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5300322771072388,
      "eval_runtime": 0.6969,
      "eval_samples_per_second": 63.132,
      "eval_steps_per_second": 2.87,
      "step": 186
    },
    {
      "epoch": 94.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5598505139350891,
      "eval_runtime": 0.6962,
      "eval_samples_per_second": 63.196,
      "eval_steps_per_second": 2.873,
      "step": 188
    },
    {
      "epoch": 95.0,
      "grad_norm": 3.235063314437866,
      "learning_rate": 2.777777777777778e-06,
      "loss": 0.1025,
      "step": 190
    },
    {
      "epoch": 95.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.5971865057945251,
      "eval_runtime": 0.6979,
      "eval_samples_per_second": 63.049,
      "eval_steps_per_second": 2.866,
      "step": 190
    },
    {
      "epoch": 96.0,
      "eval_accuracy": 0.8181818181818182,
      "eval_loss": 0.6083278059959412,
      "eval_runtime": 0.6792,
      "eval_samples_per_second": 64.783,
      "eval_steps_per_second": 2.945,
      "step": 192
    },
    {
      "epoch": 97.0,
      "eval_accuracy": 0.8409090909090909,
      "eval_loss": 0.5969064831733704,
      "eval_runtime": 0.6832,
      "eval_samples_per_second": 64.403,
      "eval_steps_per_second": 2.927,
      "step": 194
    },
    {
      "epoch": 98.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.576902449131012,
      "eval_runtime": 0.6819,
      "eval_samples_per_second": 64.523,
      "eval_steps_per_second": 2.933,
      "step": 196
    },
    {
      "epoch": 99.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.56729656457901,
      "eval_runtime": 0.6923,
      "eval_samples_per_second": 63.558,
      "eval_steps_per_second": 2.889,
      "step": 198
    },
    {
      "epoch": 100.0,
      "grad_norm": 2.5425853729248047,
      "learning_rate": 0.0,
      "loss": 0.1184,
      "step": 200
    },
    {
      "epoch": 100.0,
      "eval_accuracy": 0.8636363636363636,
      "eval_loss": 0.5642234683036804,
      "eval_runtime": 0.6837,
      "eval_samples_per_second": 64.359,
      "eval_steps_per_second": 2.925,
      "step": 200
    },
    {
      "epoch": 100.0,
      "step": 200,
      "total_flos": 1.8822505078960128e+18,
      "train_loss": 0.20665995717048646,
      "train_runtime": 1214.9613,
      "train_samples_per_second": 20.001,
      "train_steps_per_second": 0.165
    },
    {
      "epoch": 100.0,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.3297279477119446,
      "eval_runtime": 0.7284,
      "eval_samples_per_second": 60.403,
      "eval_steps_per_second": 2.746,
      "step": 200
    }
  ],
  "logging_steps": 10,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 500,
  "total_flos": 1.8822505078960128e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}