{
  "best_metric": 0.8591549295774648,
  "best_model_checkpoint": "beit-base-patch16-224-65-fold3/checkpoint-97",
  "epoch": 92.3076923076923,
  "eval_steps": 500,
  "global_step": 300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9230769230769231,
      "eval_accuracy": 0.5211267605633803,
      "eval_loss": 0.8549354672431946,
      "eval_runtime": 1.1595,
      "eval_samples_per_second": 61.235,
      "eval_steps_per_second": 2.587,
      "step": 3
    },
    {
      "epoch": 1.8461538461538463,
      "eval_accuracy": 0.5633802816901409,
      "eval_loss": 0.697589099407196,
      "eval_runtime": 1.1208,
      "eval_samples_per_second": 63.345,
      "eval_steps_per_second": 2.677,
      "step": 6
    },
    {
      "epoch": 2.769230769230769,
      "eval_accuracy": 0.5633802816901409,
      "eval_loss": 0.6808653473854065,
      "eval_runtime": 1.0746,
      "eval_samples_per_second": 66.072,
      "eval_steps_per_second": 2.792,
      "step": 9
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 5.197551727294922,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7778,
      "step": 10
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6056338028169014,
      "eval_loss": 0.645892322063446,
      "eval_runtime": 1.0976,
      "eval_samples_per_second": 64.689,
      "eval_steps_per_second": 2.733,
      "step": 13
    },
    {
      "epoch": 4.923076923076923,
      "eval_accuracy": 0.6338028169014085,
      "eval_loss": 0.6352745294570923,
      "eval_runtime": 1.1067,
      "eval_samples_per_second": 64.154,
      "eval_steps_per_second": 2.711,
      "step": 16
    },
    {
      "epoch": 5.846153846153846,
      "eval_accuracy": 0.6197183098591549,
      "eval_loss": 0.6140676736831665,
      "eval_runtime": 1.1233,
      "eval_samples_per_second": 63.206,
      "eval_steps_per_second": 2.671,
      "step": 19
    },
    {
      "epoch": 6.153846153846154,
      "grad_norm": 8.811356544494629,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.6542,
      "step": 20
    },
    {
      "epoch": 6.769230769230769,
      "eval_accuracy": 0.6056338028169014,
      "eval_loss": 0.6002594232559204,
      "eval_runtime": 1.0899,
      "eval_samples_per_second": 65.144,
      "eval_steps_per_second": 2.753,
      "step": 22
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.676056338028169,
      "eval_loss": 0.6167530417442322,
      "eval_runtime": 1.1246,
      "eval_samples_per_second": 63.133,
      "eval_steps_per_second": 2.668,
      "step": 26
    },
    {
      "epoch": 8.923076923076923,
      "eval_accuracy": 0.6901408450704225,
      "eval_loss": 0.5781181454658508,
      "eval_runtime": 1.1021,
      "eval_samples_per_second": 64.421,
      "eval_steps_per_second": 2.722,
      "step": 29
    },
    {
      "epoch": 9.23076923076923,
      "grad_norm": 11.113332748413086,
      "learning_rate": 5e-05,
      "loss": 0.5817,
      "step": 30
    },
    {
      "epoch": 9.846153846153847,
      "eval_accuracy": 0.7323943661971831,
      "eval_loss": 0.5710453391075134,
      "eval_runtime": 1.1302,
      "eval_samples_per_second": 62.82,
      "eval_steps_per_second": 2.654,
      "step": 32
    },
    {
      "epoch": 10.76923076923077,
      "eval_accuracy": 0.7464788732394366,
      "eval_loss": 0.5344887971878052,
      "eval_runtime": 1.0983,
      "eval_samples_per_second": 64.645,
      "eval_steps_per_second": 2.731,
      "step": 35
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.647887323943662,
      "eval_loss": 0.6058281064033508,
      "eval_runtime": 1.0953,
      "eval_samples_per_second": 64.82,
      "eval_steps_per_second": 2.739,
      "step": 39
    },
    {
      "epoch": 12.307692307692308,
      "grad_norm": 7.416208744049072,
      "learning_rate": 4.814814814814815e-05,
      "loss": 0.513,
      "step": 40
    },
    {
      "epoch": 12.923076923076923,
      "eval_accuracy": 0.704225352112676,
      "eval_loss": 0.6432794332504272,
      "eval_runtime": 1.1166,
      "eval_samples_per_second": 63.586,
      "eval_steps_per_second": 2.687,
      "step": 42
    },
    {
      "epoch": 13.846153846153847,
      "eval_accuracy": 0.704225352112676,
      "eval_loss": 0.5830168128013611,
      "eval_runtime": 1.0859,
      "eval_samples_per_second": 65.381,
      "eval_steps_per_second": 2.763,
      "step": 45
    },
    {
      "epoch": 14.76923076923077,
      "eval_accuracy": 0.704225352112676,
      "eval_loss": 0.6167191863059998,
      "eval_runtime": 1.1037,
      "eval_samples_per_second": 64.331,
      "eval_steps_per_second": 2.718,
      "step": 48
    },
    {
      "epoch": 15.384615384615385,
      "grad_norm": 4.344594955444336,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.4756,
      "step": 50
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6338028169014085,
      "eval_loss": 0.7304334044456482,
      "eval_runtime": 1.1067,
      "eval_samples_per_second": 64.155,
      "eval_steps_per_second": 2.711,
      "step": 52
    },
    {
      "epoch": 16.923076923076923,
      "eval_accuracy": 0.7605633802816901,
      "eval_loss": 0.5485327839851379,
      "eval_runtime": 1.1071,
      "eval_samples_per_second": 64.132,
      "eval_steps_per_second": 2.71,
      "step": 55
    },
    {
      "epoch": 17.846153846153847,
      "eval_accuracy": 0.7605633802816901,
      "eval_loss": 0.5166436433792114,
      "eval_runtime": 1.1089,
      "eval_samples_per_second": 64.028,
      "eval_steps_per_second": 2.705,
      "step": 58
    },
    {
      "epoch": 18.46153846153846,
      "grad_norm": 11.026878356933594,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.4123,
      "step": 60
    },
    {
      "epoch": 18.76923076923077,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.6266874670982361,
      "eval_runtime": 1.0972,
      "eval_samples_per_second": 64.709,
      "eval_steps_per_second": 2.734,
      "step": 61
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.42525675892829895,
      "eval_runtime": 1.125,
      "eval_samples_per_second": 63.111,
      "eval_steps_per_second": 2.667,
      "step": 65
    },
    {
      "epoch": 20.923076923076923,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.4697897434234619,
      "eval_runtime": 1.069,
      "eval_samples_per_second": 66.415,
      "eval_steps_per_second": 2.806,
      "step": 68
    },
    {
      "epoch": 21.53846153846154,
      "grad_norm": 6.852389335632324,
      "learning_rate": 4.259259259259259e-05,
      "loss": 0.3745,
      "step": 70
    },
    {
      "epoch": 21.846153846153847,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.5312201976776123,
      "eval_runtime": 1.1106,
      "eval_samples_per_second": 63.927,
      "eval_steps_per_second": 2.701,
      "step": 71
    },
    {
      "epoch": 22.76923076923077,
      "eval_accuracy": 0.7464788732394366,
      "eval_loss": 0.5157667398452759,
      "eval_runtime": 1.0999,
      "eval_samples_per_second": 64.553,
      "eval_steps_per_second": 2.728,
      "step": 74
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.5968685746192932,
      "eval_runtime": 1.072,
      "eval_samples_per_second": 66.231,
      "eval_steps_per_second": 2.798,
      "step": 78
    },
    {
      "epoch": 24.615384615384617,
      "grad_norm": 8.991398811340332,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.3751,
      "step": 80
    },
    {
      "epoch": 24.923076923076923,
      "eval_accuracy": 0.7605633802816901,
      "eval_loss": 0.541881799697876,
      "eval_runtime": 1.074,
      "eval_samples_per_second": 66.11,
      "eval_steps_per_second": 2.793,
      "step": 81
    },
    {
      "epoch": 25.846153846153847,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.463002473115921,
      "eval_runtime": 1.09,
      "eval_samples_per_second": 65.141,
      "eval_steps_per_second": 2.752,
      "step": 84
    },
    {
      "epoch": 26.76923076923077,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.5366793870925903,
      "eval_runtime": 1.0752,
      "eval_samples_per_second": 66.032,
      "eval_steps_per_second": 2.79,
      "step": 87
    },
    {
      "epoch": 27.692307692307693,
      "grad_norm": 3.5171091556549072,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.3079,
      "step": 90
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.8309859154929577,
      "eval_loss": 0.5220130085945129,
      "eval_runtime": 1.0931,
      "eval_samples_per_second": 64.955,
      "eval_steps_per_second": 2.745,
      "step": 91
    },
    {
      "epoch": 28.923076923076923,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.5341811180114746,
      "eval_runtime": 1.1125,
      "eval_samples_per_second": 63.819,
      "eval_steps_per_second": 2.697,
      "step": 94
    },
    {
      "epoch": 29.846153846153847,
      "eval_accuracy": 0.8591549295774648,
      "eval_loss": 0.5711377859115601,
      "eval_runtime": 1.0957,
      "eval_samples_per_second": 64.797,
      "eval_steps_per_second": 2.738,
      "step": 97
    },
    {
      "epoch": 30.76923076923077,
      "grad_norm": 7.372060298919678,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.2831,
      "step": 100
    },
    {
      "epoch": 30.76923076923077,
      "eval_accuracy": 0.7605633802816901,
      "eval_loss": 0.575693666934967,
      "eval_runtime": 1.0914,
      "eval_samples_per_second": 65.056,
      "eval_steps_per_second": 2.749,
      "step": 100
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.7464788732394366,
      "eval_loss": 0.5199587345123291,
      "eval_runtime": 1.0974,
      "eval_samples_per_second": 64.696,
      "eval_steps_per_second": 2.734,
      "step": 104
    },
    {
      "epoch": 32.92307692307692,
      "eval_accuracy": 0.8450704225352113,
      "eval_loss": 0.4495673179626465,
      "eval_runtime": 1.0764,
      "eval_samples_per_second": 65.963,
      "eval_steps_per_second": 2.787,
      "step": 107
    },
    {
      "epoch": 33.84615384615385,
      "grad_norm": 14.049703598022461,
      "learning_rate": 3.518518518518519e-05,
      "loss": 0.292,
      "step": 110
    },
    {
      "epoch": 33.84615384615385,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.6480037569999695,
      "eval_runtime": 1.1161,
      "eval_samples_per_second": 63.613,
      "eval_steps_per_second": 2.688,
      "step": 110
    },
    {
      "epoch": 34.76923076923077,
      "eval_accuracy": 0.7464788732394366,
      "eval_loss": 0.6956098079681396,
      "eval_runtime": 1.1028,
      "eval_samples_per_second": 64.382,
      "eval_steps_per_second": 2.72,
      "step": 113
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.5629093647003174,
      "eval_runtime": 1.0914,
      "eval_samples_per_second": 65.052,
      "eval_steps_per_second": 2.749,
      "step": 117
    },
    {
      "epoch": 36.92307692307692,
      "grad_norm": 6.377761363983154,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.2712,
      "step": 120
    },
    {
      "epoch": 36.92307692307692,
      "eval_accuracy": 0.6901408450704225,
      "eval_loss": 0.761414110660553,
      "eval_runtime": 1.1022,
      "eval_samples_per_second": 64.414,
      "eval_steps_per_second": 2.722,
      "step": 120
    },
    {
      "epoch": 37.84615384615385,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.5624626278877258,
      "eval_runtime": 1.076,
      "eval_samples_per_second": 65.988,
      "eval_steps_per_second": 2.788,
      "step": 123
    },
    {
      "epoch": 38.76923076923077,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.571108877658844,
      "eval_runtime": 1.098,
      "eval_samples_per_second": 64.66,
      "eval_steps_per_second": 2.732,
      "step": 126
    },
    {
      "epoch": 40.0,
      "grad_norm": 5.312506198883057,
      "learning_rate": 3.148148148148148e-05,
      "loss": 0.2447,
      "step": 130
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.5476051568984985,
      "eval_runtime": 1.0921,
      "eval_samples_per_second": 65.011,
      "eval_steps_per_second": 2.747,
      "step": 130
    },
    {
      "epoch": 40.92307692307692,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.5354418754577637,
      "eval_runtime": 1.0697,
      "eval_samples_per_second": 66.374,
      "eval_steps_per_second": 2.805,
      "step": 133
    },
    {
      "epoch": 41.84615384615385,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.5217384099960327,
      "eval_runtime": 1.1215,
      "eval_samples_per_second": 63.311,
      "eval_steps_per_second": 2.675,
      "step": 136
    },
    {
      "epoch": 42.76923076923077,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.576663613319397,
      "eval_runtime": 1.1012,
      "eval_samples_per_second": 64.478,
      "eval_steps_per_second": 2.724,
      "step": 139
    },
    {
      "epoch": 43.07692307692308,
      "grad_norm": 9.378477096557617,
      "learning_rate": 2.962962962962963e-05,
      "loss": 0.185,
      "step": 140
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.5605924129486084,
      "eval_runtime": 1.09,
      "eval_samples_per_second": 65.14,
      "eval_steps_per_second": 2.752,
      "step": 143
    },
    {
      "epoch": 44.92307692307692,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.67192542552948,
      "eval_runtime": 1.103,
      "eval_samples_per_second": 64.369,
      "eval_steps_per_second": 2.72,
      "step": 146
    },
    {
      "epoch": 45.84615384615385,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.6073798537254333,
      "eval_runtime": 1.071,
      "eval_samples_per_second": 66.295,
      "eval_steps_per_second": 2.801,
      "step": 149
    },
    {
      "epoch": 46.15384615384615,
      "grad_norm": 11.159360885620117,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.1921,
      "step": 150
    },
    {
      "epoch": 46.76923076923077,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.6350691318511963,
      "eval_runtime": 1.1244,
      "eval_samples_per_second": 63.143,
      "eval_steps_per_second": 2.668,
      "step": 152
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.5916067957878113,
      "eval_runtime": 1.081,
      "eval_samples_per_second": 65.681,
      "eval_steps_per_second": 2.775,
      "step": 156
    },
    {
      "epoch": 48.92307692307692,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.6103379130363464,
      "eval_runtime": 1.0963,
      "eval_samples_per_second": 64.761,
      "eval_steps_per_second": 2.736,
      "step": 159
    },
    {
      "epoch": 49.23076923076923,
      "grad_norm": 5.742426872253418,
      "learning_rate": 2.5925925925925925e-05,
      "loss": 0.1844,
      "step": 160
    },
    {
      "epoch": 49.84615384615385,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.5758461952209473,
      "eval_runtime": 1.1062,
      "eval_samples_per_second": 64.182,
      "eval_steps_per_second": 2.712,
      "step": 162
    },
    {
      "epoch": 50.76923076923077,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.5496791005134583,
      "eval_runtime": 1.106,
      "eval_samples_per_second": 64.196,
      "eval_steps_per_second": 2.713,
      "step": 165
    },
    {
      "epoch": 52.0,
      "eval_accuracy": 0.8309859154929577,
      "eval_loss": 0.5377268195152283,
      "eval_runtime": 1.0974,
      "eval_samples_per_second": 64.698,
      "eval_steps_per_second": 2.734,
      "step": 169
    },
    {
      "epoch": 52.30769230769231,
      "grad_norm": 5.0976691246032715,
      "learning_rate": 2.4074074074074074e-05,
      "loss": 0.17,
      "step": 170
    },
    {
      "epoch": 52.92307692307692,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.6278994679450989,
      "eval_runtime": 1.0996,
      "eval_samples_per_second": 64.568,
      "eval_steps_per_second": 2.728,
      "step": 172
    },
    {
      "epoch": 53.84615384615385,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.5825640559196472,
      "eval_runtime": 1.104,
      "eval_samples_per_second": 64.312,
      "eval_steps_per_second": 2.717,
      "step": 175
    },
    {
      "epoch": 54.76923076923077,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.7172585725784302,
      "eval_runtime": 1.0805,
      "eval_samples_per_second": 65.708,
      "eval_steps_per_second": 2.776,
      "step": 178
    },
    {
      "epoch": 55.38461538461539,
      "grad_norm": 5.814525604248047,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.1724,
      "step": 180
    },
    {
      "epoch": 56.0,
      "eval_accuracy": 0.8450704225352113,
      "eval_loss": 0.5339981913566589,
      "eval_runtime": 1.1095,
      "eval_samples_per_second": 63.99,
      "eval_steps_per_second": 2.704,
      "step": 182
    },
    {
      "epoch": 56.92307692307692,
      "eval_accuracy": 0.8591549295774648,
      "eval_loss": 0.5527878403663635,
      "eval_runtime": 1.1322,
      "eval_samples_per_second": 62.71,
      "eval_steps_per_second": 2.65,
      "step": 185
    },
    {
      "epoch": 57.84615384615385,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.6547489166259766,
      "eval_runtime": 1.1043,
      "eval_samples_per_second": 64.297,
      "eval_steps_per_second": 2.717,
      "step": 188
    },
    {
      "epoch": 58.46153846153846,
      "grad_norm": 5.154381275177002,
      "learning_rate": 2.037037037037037e-05,
      "loss": 0.1734,
      "step": 190
    },
    {
      "epoch": 58.76923076923077,
      "eval_accuracy": 0.8309859154929577,
      "eval_loss": 0.5986492037773132,
      "eval_runtime": 1.0821,
      "eval_samples_per_second": 65.612,
      "eval_steps_per_second": 2.772,
      "step": 191
    },
    {
      "epoch": 60.0,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.6057180762290955,
      "eval_runtime": 1.1087,
      "eval_samples_per_second": 64.039,
      "eval_steps_per_second": 2.706,
      "step": 195
    },
    {
      "epoch": 60.92307692307692,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.7183056473731995,
      "eval_runtime": 1.0617,
      "eval_samples_per_second": 66.874,
      "eval_steps_per_second": 2.826,
      "step": 198
    },
    {
      "epoch": 61.53846153846154,
      "grad_norm": 4.651764392852783,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.1582,
      "step": 200
    },
    {
      "epoch": 61.84615384615385,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.5912399291992188,
      "eval_runtime": 1.0929,
      "eval_samples_per_second": 64.966,
      "eval_steps_per_second": 2.745,
      "step": 201
    },
    {
      "epoch": 62.76923076923077,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.6002036333084106,
      "eval_runtime": 1.0987,
      "eval_samples_per_second": 64.622,
      "eval_steps_per_second": 2.73,
      "step": 204
    },
    {
      "epoch": 64.0,
      "eval_accuracy": 0.7605633802816901,
      "eval_loss": 0.7886073589324951,
      "eval_runtime": 1.0996,
      "eval_samples_per_second": 64.566,
      "eval_steps_per_second": 2.728,
      "step": 208
    },
    {
      "epoch": 64.61538461538461,
      "grad_norm": 4.254072666168213,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.1372,
      "step": 210
    },
    {
      "epoch": 64.92307692307692,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.7019013166427612,
      "eval_runtime": 1.091,
      "eval_samples_per_second": 65.079,
      "eval_steps_per_second": 2.75,
      "step": 211
    },
    {
      "epoch": 65.84615384615384,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.6460186243057251,
      "eval_runtime": 1.1036,
      "eval_samples_per_second": 64.335,
      "eval_steps_per_second": 2.718,
      "step": 214
    },
    {
      "epoch": 66.76923076923077,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.6934826970100403,
      "eval_runtime": 1.0674,
      "eval_samples_per_second": 66.518,
      "eval_steps_per_second": 2.811,
      "step": 217
    },
    {
      "epoch": 67.6923076923077,
      "grad_norm": 8.02208137512207,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 0.153,
      "step": 220
    },
    {
      "epoch": 68.0,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.8107859492301941,
      "eval_runtime": 1.1001,
      "eval_samples_per_second": 64.542,
      "eval_steps_per_second": 2.727,
      "step": 221
    },
    {
      "epoch": 68.92307692307692,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.7539263963699341,
      "eval_runtime": 1.1222,
      "eval_samples_per_second": 63.268,
      "eval_steps_per_second": 2.673,
      "step": 224
    },
    {
      "epoch": 69.84615384615384,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.7089611887931824,
      "eval_runtime": 1.1229,
      "eval_samples_per_second": 63.227,
      "eval_steps_per_second": 2.672,
      "step": 227
    },
    {
      "epoch": 70.76923076923077,
      "grad_norm": 4.6820597648620605,
      "learning_rate": 1.2962962962962962e-05,
      "loss": 0.1512,
      "step": 230
    },
    {
      "epoch": 70.76923076923077,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.7146645188331604,
      "eval_runtime": 1.1135,
      "eval_samples_per_second": 63.762,
      "eval_steps_per_second": 2.694,
      "step": 230
    },
    {
      "epoch": 72.0,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.8680395483970642,
      "eval_runtime": 1.0821,
      "eval_samples_per_second": 65.611,
      "eval_steps_per_second": 2.772,
      "step": 234
    },
    {
      "epoch": 72.92307692307692,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.8785440325737,
      "eval_runtime": 1.1317,
      "eval_samples_per_second": 62.738,
      "eval_steps_per_second": 2.651,
      "step": 237
    },
    {
      "epoch": 73.84615384615384,
      "grad_norm": 5.328825950622559,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.1381,
      "step": 240
    },
    {
      "epoch": 73.84615384615384,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.7413442134857178,
      "eval_runtime": 1.1102,
      "eval_samples_per_second": 63.955,
      "eval_steps_per_second": 2.702,
      "step": 240
    },
    {
      "epoch": 74.76923076923077,
      "eval_accuracy": 0.8169014084507042,
      "eval_loss": 0.7254679799079895,
      "eval_runtime": 1.0871,
      "eval_samples_per_second": 65.312,
      "eval_steps_per_second": 2.76,
      "step": 243
    },
    {
      "epoch": 76.0,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.7123715877532959,
      "eval_runtime": 1.1015,
      "eval_samples_per_second": 64.455,
      "eval_steps_per_second": 2.723,
      "step": 247
    },
    {
      "epoch": 76.92307692307692,
      "grad_norm": 6.364928245544434,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.1432,
      "step": 250
    },
    {
      "epoch": 76.92307692307692,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.734306275844574,
      "eval_runtime": 1.093,
      "eval_samples_per_second": 64.958,
      "eval_steps_per_second": 2.745,
      "step": 250
    },
    {
      "epoch": 77.84615384615384,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.7403556704521179,
      "eval_runtime": 1.1105,
      "eval_samples_per_second": 63.937,
      "eval_steps_per_second": 2.702,
      "step": 253
    },
    {
      "epoch": 78.76923076923077,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.6940977573394775,
      "eval_runtime": 1.08,
      "eval_samples_per_second": 65.743,
      "eval_steps_per_second": 2.778,
      "step": 256
    },
    {
      "epoch": 80.0,
      "grad_norm": 8.516518592834473,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.1135,
      "step": 260
    },
    {
      "epoch": 80.0,
      "eval_accuracy": 0.8309859154929577,
      "eval_loss": 0.6721236109733582,
      "eval_runtime": 1.102,
      "eval_samples_per_second": 64.43,
      "eval_steps_per_second": 2.722,
      "step": 260
    },
    {
      "epoch": 80.92307692307692,
      "eval_accuracy": 0.8309859154929577,
      "eval_loss": 0.6691632866859436,
      "eval_runtime": 1.0893,
      "eval_samples_per_second": 65.178,
      "eval_steps_per_second": 2.754,
      "step": 263
    },
    {
      "epoch": 81.84615384615384,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.6879910826683044,
      "eval_runtime": 1.0969,
      "eval_samples_per_second": 64.728,
      "eval_steps_per_second": 2.735,
      "step": 266
    },
    {
      "epoch": 82.76923076923077,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.6856514811515808,
      "eval_runtime": 1.0765,
      "eval_samples_per_second": 65.955,
      "eval_steps_per_second": 2.787,
      "step": 269
    },
    {
      "epoch": 83.07692307692308,
      "grad_norm": 4.385526657104492,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.1182,
      "step": 270
    },
    {
      "epoch": 84.0,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.6849518418312073,
      "eval_runtime": 1.1279,
      "eval_samples_per_second": 62.95,
      "eval_steps_per_second": 2.66,
      "step": 273
    },
    {
      "epoch": 84.92307692307692,
      "eval_accuracy": 0.7887323943661971,
      "eval_loss": 0.6815563440322876,
      "eval_runtime": 1.0923,
      "eval_samples_per_second": 64.998,
      "eval_steps_per_second": 2.746,
      "step": 276
    },
    {
      "epoch": 85.84615384615384,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.7047742605209351,
      "eval_runtime": 1.1236,
      "eval_samples_per_second": 63.187,
      "eval_steps_per_second": 2.67,
      "step": 279
    },
    {
      "epoch": 86.15384615384616,
      "grad_norm": 5.567570209503174,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.1019,
      "step": 280
    },
    {
      "epoch": 86.76923076923077,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.7803800702095032,
      "eval_runtime": 1.0996,
      "eval_samples_per_second": 64.567,
      "eval_steps_per_second": 2.728,
      "step": 282
    },
    {
      "epoch": 88.0,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.8013417720794678,
      "eval_runtime": 1.1133,
      "eval_samples_per_second": 63.773,
      "eval_steps_per_second": 2.695,
      "step": 286
    },
    {
      "epoch": 88.92307692307692,
      "eval_accuracy": 0.7605633802816901,
      "eval_loss": 0.7505632638931274,
      "eval_runtime": 1.1302,
      "eval_samples_per_second": 62.82,
      "eval_steps_per_second": 2.654,
      "step": 289
    },
    {
      "epoch": 89.23076923076923,
      "grad_norm": 3.6284656524658203,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.1163,
      "step": 290
    },
    {
      "epoch": 89.84615384615384,
      "eval_accuracy": 0.7746478873239436,
      "eval_loss": 0.7047209143638611,
      "eval_runtime": 1.1154,
      "eval_samples_per_second": 63.653,
      "eval_steps_per_second": 2.69,
      "step": 292
    },
    {
      "epoch": 90.76923076923077,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.6762815713882446,
      "eval_runtime": 1.1157,
      "eval_samples_per_second": 63.635,
      "eval_steps_per_second": 2.689,
      "step": 295
    },
    {
      "epoch": 92.0,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.6605757474899292,
      "eval_runtime": 1.1108,
      "eval_samples_per_second": 63.918,
      "eval_steps_per_second": 2.701,
      "step": 299
    },
    {
      "epoch": 92.3076923076923,
      "grad_norm": 3.279923439025879,
      "learning_rate": 0.0,
      "loss": 0.1258,
      "step": 300
    },
    {
      "epoch": 92.3076923076923,
      "eval_accuracy": 0.8028169014084507,
      "eval_loss": 0.6592366695404053,
      "eval_runtime": 1.0928,
      "eval_samples_per_second": 64.968,
      "eval_steps_per_second": 2.745,
      "step": 300
    },
    {
      "epoch": 92.3076923076923,
      "step": 300,
      "total_flos": 2.839022453308834e+18,
      "train_loss": 0.26989306092262266,
      "train_runtime": 1732.8571,
      "train_samples_per_second": 22.91,
      "train_steps_per_second": 0.173
    },
    {
      "epoch": 92.3076923076923,
      "eval_accuracy": 0.8591549295774648,
      "eval_loss": 0.5711377859115601,
      "eval_runtime": 1.0772,
      "eval_samples_per_second": 65.91,
      "eval_steps_per_second": 2.785,
      "step": 300
    }
  ],
  "logging_steps": 10,
  "max_steps": 300,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 500,
  "total_flos": 2.839022453308834e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}