{
  "best_metric": 0.7741935483870968,
  "best_model_checkpoint": "videomae-base-finetuned-subset\\checkpoint-1792",
  "epoch": 99.00108108108108,
  "eval_steps": 500,
  "global_step": 11100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.504504504504505e-07,
      "loss": 1.6259,
      "step": 10
    },
    {
      "epoch": 0.0,
      "learning_rate": 9.00900900900901e-07,
      "loss": 1.6654,
      "step": 20
    },
    {
      "epoch": 0.0,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 1.6053,
      "step": 30
    },
    {
      "epoch": 0.0,
      "learning_rate": 1.801801801801802e-06,
      "loss": 1.6961,
      "step": 40
    },
    {
      "epoch": 0.0,
      "learning_rate": 2.2522522522522524e-06,
      "loss": 1.6694,
      "step": 50
    },
    {
      "epoch": 0.01,
      "learning_rate": 2.702702702702703e-06,
      "loss": 1.626,
      "step": 60
    },
    {
      "epoch": 0.01,
      "learning_rate": 3.153153153153153e-06,
      "loss": 1.6398,
      "step": 70
    },
    {
      "epoch": 0.01,
      "learning_rate": 3.603603603603604e-06,
      "loss": 1.667,
      "step": 80
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 1.6241,
      "step": 90
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.504504504504505e-06,
      "loss": 1.6307,
      "step": 100
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.954954954954955e-06,
      "loss": 1.638,
      "step": 110
    },
    {
      "epoch": 0.01,
      "eval_accuracy": 0.15668202764976957,
      "eval_loss": 1.573599100112915,
      "eval_runtime": 309.1073,
      "eval_samples_per_second": 0.702,
      "eval_steps_per_second": 0.178,
      "step": 112
    },
    {
      "epoch": 1.0,
      "learning_rate": 5.405405405405406e-06,
      "loss": 1.6344,
      "step": 120
    },
    {
      "epoch": 1.0,
      "learning_rate": 5.855855855855856e-06,
      "loss": 1.598,
      "step": 130
    },
    {
      "epoch": 1.0,
      "learning_rate": 6.306306306306306e-06,
      "loss": 1.6336,
      "step": 140
    },
    {
      "epoch": 1.0,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 1.6173,
      "step": 150
    },
    {
      "epoch": 1.0,
      "learning_rate": 7.207207207207208e-06,
      "loss": 1.6211,
      "step": 160
    },
    {
      "epoch": 1.01,
      "learning_rate": 7.657657657657658e-06,
      "loss": 1.6295,
      "step": 170
    },
    {
      "epoch": 1.01,
      "learning_rate": 8.108108108108109e-06,
      "loss": 1.5838,
      "step": 180
    },
    {
      "epoch": 1.01,
      "learning_rate": 8.558558558558558e-06,
      "loss": 1.6067,
      "step": 190
    },
    {
      "epoch": 1.01,
      "learning_rate": 9.00900900900901e-06,
      "loss": 1.5873,
      "step": 200
    },
    {
      "epoch": 1.01,
      "learning_rate": 9.45945945945946e-06,
      "loss": 1.5708,
      "step": 210
    },
    {
      "epoch": 1.01,
      "learning_rate": 9.90990990990991e-06,
      "loss": 1.5845,
      "step": 220
    },
    {
      "epoch": 1.01,
      "eval_accuracy": 0.271889400921659,
      "eval_loss": 1.5840919017791748,
      "eval_runtime": 275.6132,
      "eval_samples_per_second": 0.787,
      "eval_steps_per_second": 0.2,
      "step": 224
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.0360360360360361e-05,
      "loss": 1.5875,
      "step": 230
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.0810810810810812e-05,
      "loss": 1.5903,
      "step": 240
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.1261261261261261e-05,
      "loss": 1.5704,
      "step": 250
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.1711711711711713e-05,
      "loss": 1.6581,
      "step": 260
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.2162162162162164e-05,
      "loss": 1.6163,
      "step": 270
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.2612612612612611e-05,
      "loss": 1.6129,
      "step": 280
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.3063063063063064e-05,
      "loss": 1.6142,
      "step": 290
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 1.5736,
      "step": 300
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.3963963963963963e-05,
      "loss": 1.4639,
      "step": 310
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.4414414414414416e-05,
      "loss": 1.4171,
      "step": 320
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.4864864864864867e-05,
      "loss": 1.4522,
      "step": 330
    },
    {
      "epoch": 2.01,
      "eval_accuracy": 0.2350230414746544,
      "eval_loss": 1.6293407678604126,
      "eval_runtime": 315.1151,
      "eval_samples_per_second": 0.689,
      "eval_steps_per_second": 0.175,
      "step": 336
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.5315315315315316e-05,
      "loss": 1.4434,
      "step": 340
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.5765765765765765e-05,
      "loss": 1.4351,
      "step": 350
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 1.3671,
      "step": 360
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.3213,
      "step": 370
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.7117117117117117e-05,
      "loss": 1.2062,
      "step": 380
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.756756756756757e-05,
      "loss": 1.1743,
      "step": 390
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.801801801801802e-05,
      "loss": 1.5591,
      "step": 400
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.846846846846847e-05,
      "loss": 1.2208,
      "step": 410
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.891891891891892e-05,
      "loss": 1.258,
      "step": 420
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.936936936936937e-05,
      "loss": 1.5539,
      "step": 430
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.981981981981982e-05,
      "loss": 1.3111,
      "step": 440
    },
    {
      "epoch": 3.01,
      "eval_accuracy": 0.6036866359447005,
      "eval_loss": 1.045045256614685,
      "eval_runtime": 323.7084,
      "eval_samples_per_second": 0.67,
      "eval_steps_per_second": 0.17,
      "step": 448
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.0270270270270273e-05,
      "loss": 1.3305,
      "step": 450
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.0720720720720722e-05,
      "loss": 1.4166,
      "step": 460
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.117117117117117e-05,
      "loss": 1.2087,
      "step": 470
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.1621621621621624e-05,
      "loss": 1.2995,
      "step": 480
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.2072072072072073e-05,
      "loss": 1.5599,
      "step": 490
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.2522522522522523e-05,
      "loss": 1.2504,
      "step": 500
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.2972972972972976e-05,
      "loss": 1.3355,
      "step": 510
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.3423423423423425e-05,
      "loss": 0.9579,
      "step": 520
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.3873873873873874e-05,
      "loss": 1.1881,
      "step": 530
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.4324324324324327e-05,
      "loss": 1.2578,
      "step": 540
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.4774774774774777e-05,
      "loss": 1.1007,
      "step": 550
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.5225225225225222e-05,
      "loss": 1.2849,
      "step": 560
    },
    {
      "epoch": 4.01,
      "eval_accuracy": 0.4608294930875576,
      "eval_loss": 1.3185501098632812,
      "eval_runtime": 266.5841,
      "eval_samples_per_second": 0.814,
      "eval_steps_per_second": 0.206,
      "step": 560
    },
    {
      "epoch": 5.0,
      "learning_rate": 2.5675675675675675e-05,
      "loss": 1.3457,
      "step": 570
    },
    {
      "epoch": 5.0,
      "learning_rate": 2.6126126126126128e-05,
      "loss": 1.1865,
      "step": 580
    },
    {
      "epoch": 5.0,
      "learning_rate": 2.6576576576576577e-05,
      "loss": 1.2755,
      "step": 590
    },
    {
      "epoch": 5.0,
      "learning_rate": 2.702702702702703e-05,
      "loss": 1.2319,
      "step": 600
    },
    {
      "epoch": 5.0,
      "learning_rate": 2.7477477477477483e-05,
      "loss": 1.2819,
      "step": 610
    },
    {
      "epoch": 5.01,
      "learning_rate": 2.7927927927927926e-05,
      "loss": 1.205,
      "step": 620
    },
    {
      "epoch": 5.01,
      "learning_rate": 2.8378378378378378e-05,
      "loss": 1.1357,
      "step": 630
    },
    {
      "epoch": 5.01,
      "learning_rate": 2.882882882882883e-05,
      "loss": 1.064,
      "step": 640
    },
    {
      "epoch": 5.01,
      "learning_rate": 2.927927927927928e-05,
      "loss": 1.2596,
      "step": 650
    },
    {
      "epoch": 5.01,
      "learning_rate": 2.9729729729729733e-05,
      "loss": 1.4667,
      "step": 660
    },
    {
      "epoch": 5.01,
      "learning_rate": 3.0180180180180183e-05,
      "loss": 1.3246,
      "step": 670
    },
    {
      "epoch": 5.01,
      "eval_accuracy": 0.5161290322580645,
      "eval_loss": 1.1759101152420044,
      "eval_runtime": 305.6473,
      "eval_samples_per_second": 0.71,
      "eval_steps_per_second": 0.18,
      "step": 672
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.063063063063063e-05,
      "loss": 1.3137,
      "step": 680
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.108108108108108e-05,
      "loss": 1.2802,
      "step": 690
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.153153153153153e-05,
      "loss": 1.2927,
      "step": 700
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.198198198198199e-05,
      "loss": 1.1537,
      "step": 710
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.2432432432432436e-05,
      "loss": 1.2658,
      "step": 720
    },
    {
      "epoch": 6.01,
      "learning_rate": 3.2882882882882886e-05,
      "loss": 1.0551,
      "step": 730
    },
    {
      "epoch": 6.01,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.1632,
      "step": 740
    },
    {
      "epoch": 6.01,
      "learning_rate": 3.3783783783783784e-05,
      "loss": 1.1768,
      "step": 750
    },
    {
      "epoch": 6.01,
      "learning_rate": 3.4234234234234234e-05,
      "loss": 1.0402,
      "step": 760
    },
    {
      "epoch": 6.01,
      "learning_rate": 3.468468468468469e-05,
      "loss": 1.2052,
      "step": 770
    },
    {
      "epoch": 6.01,
      "learning_rate": 3.513513513513514e-05,
      "loss": 1.3801,
      "step": 780
    },
    {
      "epoch": 6.01,
      "eval_accuracy": 0.4608294930875576,
      "eval_loss": 1.2187941074371338,
      "eval_runtime": 283.0514,
      "eval_samples_per_second": 0.767,
      "eval_steps_per_second": 0.194,
      "step": 784
    },
    {
      "epoch": 7.0,
      "learning_rate": 3.558558558558558e-05,
      "loss": 1.2512,
      "step": 790
    },
    {
      "epoch": 7.0,
      "learning_rate": 3.603603603603604e-05,
      "loss": 1.3048,
      "step": 800
    },
    {
      "epoch": 7.0,
      "learning_rate": 3.648648648648649e-05,
      "loss": 0.9037,
      "step": 810
    },
    {
      "epoch": 7.0,
      "learning_rate": 3.693693693693694e-05,
      "loss": 1.1061,
      "step": 820
    },
    {
      "epoch": 7.0,
      "learning_rate": 3.738738738738739e-05,
      "loss": 1.1639,
      "step": 830
    },
    {
      "epoch": 7.01,
      "learning_rate": 3.783783783783784e-05,
      "loss": 1.0467,
      "step": 840
    },
    {
      "epoch": 7.01,
      "learning_rate": 3.8288288288288285e-05,
      "loss": 1.2806,
      "step": 850
    },
    {
      "epoch": 7.01,
      "learning_rate": 3.873873873873874e-05,
      "loss": 1.1093,
      "step": 860
    },
    {
      "epoch": 7.01,
      "learning_rate": 3.918918918918919e-05,
      "loss": 1.1496,
      "step": 870
    },
    {
      "epoch": 7.01,
      "learning_rate": 3.963963963963964e-05,
      "loss": 1.3548,
      "step": 880
    },
    {
      "epoch": 7.01,
      "learning_rate": 4.0090090090090096e-05,
      "loss": 1.3228,
      "step": 890
    },
    {
      "epoch": 7.01,
      "eval_accuracy": 0.6405529953917051,
      "eval_loss": 0.9894869327545166,
      "eval_runtime": 264.9074,
      "eval_samples_per_second": 0.819,
      "eval_steps_per_second": 0.208,
      "step": 896
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.0540540540540545e-05,
      "loss": 0.9621,
      "step": 900
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.099099099099099e-05,
      "loss": 0.7777,
      "step": 910
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.1441441441441444e-05,
      "loss": 1.2822,
      "step": 920
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.189189189189189e-05,
      "loss": 1.3693,
      "step": 930
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.234234234234234e-05,
      "loss": 1.1463,
      "step": 940
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.27927927927928e-05,
      "loss": 1.5013,
      "step": 950
    },
    {
      "epoch": 8.01,
      "learning_rate": 4.324324324324325e-05,
      "loss": 1.1016,
      "step": 960
    },
    {
      "epoch": 8.01,
      "learning_rate": 4.369369369369369e-05,
      "loss": 0.9615,
      "step": 970
    },
    {
      "epoch": 8.01,
      "learning_rate": 4.414414414414415e-05,
      "loss": 0.9956,
      "step": 980
    },
    {
      "epoch": 8.01,
      "learning_rate": 4.4594594594594596e-05,
      "loss": 1.0309,
      "step": 990
    },
    {
      "epoch": 8.01,
      "learning_rate": 4.5045045045045046e-05,
      "loss": 0.9706,
      "step": 1000
    },
    {
      "epoch": 8.01,
      "eval_accuracy": 0.6129032258064516,
      "eval_loss": 1.1264568567276,
      "eval_runtime": 304.4356,
      "eval_samples_per_second": 0.713,
      "eval_steps_per_second": 0.181,
      "step": 1008
    },
    {
      "epoch": 9.0,
      "learning_rate": 4.54954954954955e-05,
      "loss": 1.2779,
      "step": 1010
    },
    {
      "epoch": 9.0,
      "learning_rate": 4.594594594594595e-05,
      "loss": 0.9833,
      "step": 1020
    },
    {
      "epoch": 9.0,
      "learning_rate": 4.6396396396396394e-05,
      "loss": 1.1085,
      "step": 1030
    },
    {
      "epoch": 9.0,
      "learning_rate": 4.684684684684685e-05,
      "loss": 1.4607,
      "step": 1040
    },
    {
      "epoch": 9.0,
      "learning_rate": 4.72972972972973e-05,
      "loss": 1.1599,
      "step": 1050
    },
    {
      "epoch": 9.0,
      "learning_rate": 4.774774774774775e-05,
      "loss": 0.9326,
      "step": 1060
    },
    {
      "epoch": 9.01,
      "learning_rate": 4.8198198198198205e-05,
      "loss": 1.2022,
      "step": 1070
    },
    {
      "epoch": 9.01,
      "learning_rate": 4.8648648648648654e-05,
      "loss": 0.9676,
      "step": 1080
    },
    {
      "epoch": 9.01,
      "learning_rate": 4.90990990990991e-05,
      "loss": 0.9456,
      "step": 1090
    },
    {
      "epoch": 9.01,
      "learning_rate": 4.954954954954955e-05,
      "loss": 1.0887,
      "step": 1100
    },
    {
      "epoch": 9.01,
      "learning_rate": 5e-05,
      "loss": 1.1949,
      "step": 1110
    },
    {
      "epoch": 9.01,
      "learning_rate": 4.994994994994995e-05,
      "loss": 1.2483,
      "step": 1120
    },
    {
      "epoch": 9.01,
      "eval_accuracy": 0.5483870967741935,
      "eval_loss": 1.2351685762405396,
      "eval_runtime": 297.6405,
      "eval_samples_per_second": 0.729,
      "eval_steps_per_second": 0.185,
      "step": 1120
    },
    {
      "epoch": 10.0,
      "learning_rate": 4.98998998998999e-05,
      "loss": 0.9994,
      "step": 1130
    },
    {
      "epoch": 10.0,
      "learning_rate": 4.984984984984985e-05,
      "loss": 1.0768,
      "step": 1140
    },
    {
      "epoch": 10.0,
      "learning_rate": 4.97997997997998e-05,
      "loss": 1.0222,
      "step": 1150
    },
    {
      "epoch": 10.0,
      "learning_rate": 4.974974974974975e-05,
      "loss": 0.9446,
      "step": 1160
    },
    {
      "epoch": 10.0,
      "learning_rate": 4.96996996996997e-05,
      "loss": 1.2542,
      "step": 1170
    },
    {
      "epoch": 10.01,
      "learning_rate": 4.964964964964965e-05,
      "loss": 0.9544,
      "step": 1180
    },
    {
      "epoch": 10.01,
      "learning_rate": 4.95995995995996e-05,
      "loss": 1.034,
      "step": 1190
    },
    {
      "epoch": 10.01,
      "learning_rate": 4.954954954954955e-05,
      "loss": 0.9637,
      "step": 1200
    },
    {
      "epoch": 10.01,
      "learning_rate": 4.94994994994995e-05,
      "loss": 0.8433,
      "step": 1210
    },
    {
      "epoch": 10.01,
      "learning_rate": 4.944944944944945e-05,
      "loss": 0.936,
      "step": 1220
    },
    {
      "epoch": 10.01,
      "learning_rate": 4.93993993993994e-05,
      "loss": 0.9394,
      "step": 1230
    },
    {
      "epoch": 10.01,
      "eval_accuracy": 0.4976958525345622,
      "eval_loss": 1.234522819519043,
      "eval_runtime": 263.5838,
      "eval_samples_per_second": 0.823,
      "eval_steps_per_second": 0.209,
      "step": 1232
    },
    {
      "epoch": 11.0,
      "learning_rate": 4.9349349349349347e-05,
      "loss": 1.025,
      "step": 1240
    },
    {
      "epoch": 11.0,
      "learning_rate": 4.92992992992993e-05,
      "loss": 1.0746,
      "step": 1250
    },
    {
      "epoch": 11.0,
      "learning_rate": 4.9249249249249253e-05,
      "loss": 0.709,
      "step": 1260
    },
    {
      "epoch": 11.0,
      "learning_rate": 4.9199199199199204e-05,
      "loss": 0.8909,
      "step": 1270
    },
    {
      "epoch": 11.0,
      "learning_rate": 4.9149149149149154e-05,
      "loss": 1.2861,
      "step": 1280
    },
    {
      "epoch": 11.01,
      "learning_rate": 4.90990990990991e-05,
      "loss": 0.8274,
      "step": 1290
    },
    {
      "epoch": 11.01,
      "learning_rate": 4.9049049049049054e-05,
      "loss": 1.0774,
      "step": 1300
    },
    {
      "epoch": 11.01,
      "learning_rate": 4.8998998998999004e-05,
      "loss": 0.9917,
      "step": 1310
    },
    {
      "epoch": 11.01,
      "learning_rate": 4.8948948948948954e-05,
      "loss": 1.0932,
      "step": 1320
    },
    {
      "epoch": 11.01,
      "learning_rate": 4.8898898898898904e-05,
      "loss": 1.1893,
      "step": 1330
    },
    {
      "epoch": 11.01,
      "learning_rate": 4.884884884884885e-05,
      "loss": 0.8285,
      "step": 1340
    },
    {
      "epoch": 11.01,
      "eval_accuracy": 0.6682027649769585,
      "eval_loss": 0.8701934218406677,
      "eval_runtime": 297.9313,
      "eval_samples_per_second": 0.728,
      "eval_steps_per_second": 0.185,
      "step": 1344
    },
    {
      "epoch": 12.0,
      "learning_rate": 4.87987987987988e-05,
      "loss": 0.8896,
      "step": 1350
    },
    {
      "epoch": 12.0,
      "learning_rate": 4.8748748748748754e-05,
      "loss": 0.8931,
      "step": 1360
    },
    {
      "epoch": 12.0,
      "learning_rate": 4.8698698698698704e-05,
      "loss": 1.4919,
      "step": 1370
    },
    {
      "epoch": 12.0,
      "learning_rate": 4.8648648648648654e-05,
      "loss": 0.9524,
      "step": 1380
    },
    {
      "epoch": 12.0,
      "learning_rate": 4.85985985985986e-05,
      "loss": 1.249,
      "step": 1390
    },
    {
      "epoch": 12.01,
      "learning_rate": 4.854854854854855e-05,
      "loss": 1.1616,
      "step": 1400
    },
    {
      "epoch": 12.01,
      "learning_rate": 4.8498498498498504e-05,
      "loss": 0.9776,
      "step": 1410
    },
    {
      "epoch": 12.01,
      "learning_rate": 4.8448448448448455e-05,
      "loss": 0.9735,
      "step": 1420
    },
    {
      "epoch": 12.01,
      "learning_rate": 4.83983983983984e-05,
      "loss": 1.1037,
      "step": 1430
    },
    {
      "epoch": 12.01,
      "learning_rate": 4.834834834834835e-05,
      "loss": 1.4429,
      "step": 1440
    },
    {
      "epoch": 12.01,
      "learning_rate": 4.82982982982983e-05,
      "loss": 1.1175,
      "step": 1450
    },
    {
      "epoch": 12.01,
      "eval_accuracy": 0.6405529953917051,
      "eval_loss": 0.9073271155357361,
      "eval_runtime": 287.4262,
      "eval_samples_per_second": 0.755,
      "eval_steps_per_second": 0.191,
      "step": 1456
    },
    {
      "epoch": 13.0,
      "learning_rate": 4.824824824824825e-05,
      "loss": 0.8761,
      "step": 1460
    },
    {
      "epoch": 13.0,
      "learning_rate": 4.8198198198198205e-05,
      "loss": 0.8377,
      "step": 1470
    },
    {
      "epoch": 13.0,
      "learning_rate": 4.814814814814815e-05,
      "loss": 0.9036,
      "step": 1480
    },
    {
      "epoch": 13.0,
      "learning_rate": 4.80980980980981e-05,
      "loss": 0.9554,
      "step": 1490
    },
    {
      "epoch": 13.0,
      "learning_rate": 4.804804804804805e-05,
      "loss": 1.0666,
      "step": 1500
    },
    {
      "epoch": 13.0,
      "learning_rate": 4.7997997997998e-05,
      "loss": 1.4185,
      "step": 1510
    },
    {
      "epoch": 13.01,
      "learning_rate": 4.7947947947947955e-05,
      "loss": 0.8991,
      "step": 1520
    },
    {
      "epoch": 13.01,
      "learning_rate": 4.78978978978979e-05,
      "loss": 0.9192,
      "step": 1530
    },
    {
      "epoch": 13.01,
      "learning_rate": 4.784784784784785e-05,
      "loss": 1.0836,
      "step": 1540
    },
    {
      "epoch": 13.01,
      "learning_rate": 4.77977977977978e-05,
      "loss": 0.9053,
      "step": 1550
    },
    {
      "epoch": 13.01,
      "learning_rate": 4.774774774774775e-05,
      "loss": 1.093,
      "step": 1560
    },
    {
      "epoch": 13.01,
      "eval_accuracy": 0.5576036866359447,
      "eval_loss": 0.9210311770439148,
      "eval_runtime": 291.6874,
      "eval_samples_per_second": 0.744,
      "eval_steps_per_second": 0.189,
      "step": 1568
    },
    {
      "epoch": 14.0,
      "learning_rate": 4.76976976976977e-05,
      "loss": 0.9382,
      "step": 1570
    },
    {
      "epoch": 14.0,
      "learning_rate": 4.764764764764765e-05,
      "loss": 0.9564,
      "step": 1580
    },
    {
      "epoch": 14.0,
      "learning_rate": 4.75975975975976e-05,
      "loss": 1.0024,
      "step": 1590
    },
    {
      "epoch": 14.0,
      "learning_rate": 4.754754754754755e-05,
      "loss": 0.9034,
      "step": 1600
    },
    {
      "epoch": 14.0,
      "learning_rate": 4.74974974974975e-05,
      "loss": 0.9554,
      "step": 1610
    },
    {
      "epoch": 14.0,
      "learning_rate": 4.744744744744745e-05,
      "loss": 1.0446,
      "step": 1620
    },
    {
      "epoch": 14.01,
      "learning_rate": 4.73973973973974e-05,
      "loss": 0.6378,
      "step": 1630
    },
    {
      "epoch": 14.01,
      "learning_rate": 4.734734734734735e-05,
      "loss": 1.4743,
      "step": 1640
    },
    {
      "epoch": 14.01,
      "learning_rate": 4.72972972972973e-05,
      "loss": 1.0941,
      "step": 1650
    },
    {
      "epoch": 14.01,
      "learning_rate": 4.724724724724725e-05,
      "loss": 0.9804,
      "step": 1660
    },
    {
      "epoch": 14.01,
      "learning_rate": 4.71971971971972e-05,
      "loss": 1.0513,
      "step": 1670
    },
    {
      "epoch": 14.01,
      "learning_rate": 4.714714714714715e-05,
      "loss": 0.8364,
      "step": 1680
    },
    {
      "epoch": 14.01,
      "eval_accuracy": 0.6589861751152074,
      "eval_loss": 0.9316214919090271,
      "eval_runtime": 298.8525,
      "eval_samples_per_second": 0.726,
      "eval_steps_per_second": 0.184,
      "step": 1680
    },
    {
      "epoch": 15.0,
      "learning_rate": 4.70970970970971e-05,
      "loss": 1.098,
      "step": 1690
    },
    {
      "epoch": 15.0,
      "learning_rate": 4.704704704704705e-05,
      "loss": 1.0153,
      "step": 1700
    },
    {
      "epoch": 15.0,
      "learning_rate": 4.6996996996997e-05,
      "loss": 0.7871,
      "step": 1710
    },
    {
      "epoch": 15.0,
      "learning_rate": 4.694694694694695e-05,
      "loss": 1.0481,
      "step": 1720
    },
    {
      "epoch": 15.0,
      "learning_rate": 4.68968968968969e-05,
      "loss": 1.0053,
      "step": 1730
    },
    {
      "epoch": 15.01,
      "learning_rate": 4.684684684684685e-05,
      "loss": 0.9024,
      "step": 1740
    },
    {
      "epoch": 15.01,
      "learning_rate": 4.67967967967968e-05,
      "loss": 0.6558,
      "step": 1750
    },
    {
      "epoch": 15.01,
      "learning_rate": 4.674674674674675e-05,
      "loss": 1.0366,
      "step": 1760
    },
    {
      "epoch": 15.01,
      "learning_rate": 4.66966966966967e-05,
      "loss": 0.9607,
      "step": 1770
    },
    {
      "epoch": 15.01,
      "learning_rate": 4.6646646646646644e-05,
      "loss": 0.8788,
      "step": 1780
    },
    {
      "epoch": 15.01,
      "learning_rate": 4.65965965965966e-05,
      "loss": 0.766,
      "step": 1790
    },
    {
      "epoch": 15.01,
      "eval_accuracy": 0.7741935483870968,
      "eval_loss": 0.7628481984138489,
      "eval_runtime": 265.3812,
      "eval_samples_per_second": 0.818,
      "eval_steps_per_second": 0.207,
      "step": 1792
    },
    {
      "epoch": 16.0,
      "learning_rate": 4.654654654654655e-05,
      "loss": 0.8845,
      "step": 1800
    },
    {
      "epoch": 16.0,
      "learning_rate": 4.64964964964965e-05,
      "loss": 0.7004,
      "step": 1810
    },
    {
      "epoch": 16.0,
      "learning_rate": 4.644644644644645e-05,
      "loss": 0.9095,
      "step": 1820
    },
    {
      "epoch": 16.0,
      "learning_rate": 4.6396396396396394e-05,
      "loss": 0.8129,
      "step": 1830
    },
    {
      "epoch": 16.0,
      "learning_rate": 4.634634634634635e-05,
      "loss": 1.1457,
      "step": 1840
    },
    {
      "epoch": 16.01,
      "learning_rate": 4.62962962962963e-05,
      "loss": 1.15,
      "step": 1850
    },
    {
      "epoch": 16.01,
      "learning_rate": 4.624624624624625e-05,
      "loss": 0.8391,
      "step": 1860
    },
    {
      "epoch": 16.01,
      "learning_rate": 4.61961961961962e-05,
      "loss": 1.0107,
      "step": 1870
    },
    {
      "epoch": 16.01,
      "learning_rate": 4.6146146146146144e-05,
      "loss": 1.103,
      "step": 1880
    },
    {
      "epoch": 16.01,
      "learning_rate": 4.6096096096096094e-05,
      "loss": 0.7583,
      "step": 1890
    },
    {
      "epoch": 16.01,
      "learning_rate": 4.604604604604605e-05,
      "loss": 0.7702,
      "step": 1900
    },
    {
      "epoch": 16.01,
      "eval_accuracy": 0.6682027649769585,
      "eval_loss": 0.8981941938400269,
      "eval_runtime": 299.8513,
      "eval_samples_per_second": 0.724,
      "eval_steps_per_second": 0.183,
      "step": 1904
    },
    {
      "epoch": 17.0,
      "learning_rate": 4.5995995995996e-05,
      "loss": 0.9112,
      "step": 1910
    },
    {
      "epoch": 17.0,
      "learning_rate": 4.594594594594595e-05,
      "loss": 0.808,
      "step": 1920
    },
    {
      "epoch": 17.0,
      "learning_rate": 4.5895895895895895e-05,
      "loss": 0.8087,
      "step": 1930
    },
    {
      "epoch": 17.0,
      "learning_rate": 4.5845845845845845e-05,
      "loss": 1.0996,
      "step": 1940
    },
    {
      "epoch": 17.0,
      "learning_rate": 4.57957957957958e-05,
      "loss": 0.9929,
      "step": 1950
    },
    {
      "epoch": 17.01,
      "learning_rate": 4.574574574574575e-05,
      "loss": 0.5999,
      "step": 1960
    },
    {
      "epoch": 17.01,
      "learning_rate": 4.56956956956957e-05,
      "loss": 0.8528,
      "step": 1970
    },
    {
      "epoch": 17.01,
      "learning_rate": 4.5645645645645645e-05,
      "loss": 0.9779,
      "step": 1980
    },
    {
      "epoch": 17.01,
      "learning_rate": 4.5595595595595595e-05,
      "loss": 1.172,
      "step": 1990
    },
    {
      "epoch": 17.01,
      "learning_rate": 4.5545545545545545e-05,
      "loss": 0.8387,
      "step": 2000
    },
    {
      "epoch": 17.01,
      "learning_rate": 4.54954954954955e-05,
      "loss": 0.9184,
      "step": 2010
    },
    {
      "epoch": 17.01,
      "eval_accuracy": 0.6221198156682027,
      "eval_loss": 1.1010408401489258,
      "eval_runtime": 291.2628,
      "eval_samples_per_second": 0.745,
      "eval_steps_per_second": 0.189,
      "step": 2016
    },
    {
      "epoch": 18.0,
      "learning_rate": 4.544544544544545e-05,
      "loss": 0.9717,
      "step": 2020
    },
    {
      "epoch": 18.0,
      "learning_rate": 4.5395395395395395e-05,
      "loss": 1.1521,
      "step": 2030
    },
    {
      "epoch": 18.0,
      "learning_rate": 4.5345345345345345e-05,
      "loss": 1.0965,
      "step": 2040
    },
    {
      "epoch": 18.0,
      "learning_rate": 4.5295295295295295e-05,
      "loss": 0.6683,
      "step": 2050
    },
    {
      "epoch": 18.0,
      "learning_rate": 4.524524524524525e-05,
      "loss": 1.145,
      "step": 2060
    },
    {
      "epoch": 18.0,
      "learning_rate": 4.5195195195195196e-05,
      "loss": 1.3037,
      "step": 2070
    },
    {
      "epoch": 18.01,
      "learning_rate": 4.5145145145145146e-05,
      "loss": 0.7085,
      "step": 2080
    },
    {
      "epoch": 18.01,
      "learning_rate": 4.5095095095095096e-05,
      "loss": 0.7228,
      "step": 2090
    },
    {
      "epoch": 18.01,
      "learning_rate": 4.5045045045045046e-05,
      "loss": 1.1086,
      "step": 2100
    },
    {
      "epoch": 18.01,
      "learning_rate": 4.4994994994994996e-05,
      "loss": 0.8006,
      "step": 2110
    },
    {
      "epoch": 18.01,
      "learning_rate": 4.4944944944944946e-05,
      "loss": 0.7309,
      "step": 2120
    },
    {
      "epoch": 18.01,
      "eval_accuracy": 0.6866359447004609,
      "eval_loss": 0.8245224952697754,
      "eval_runtime": 270.1099,
      "eval_samples_per_second": 0.803,
      "eval_steps_per_second": 0.204,
      "step": 2128
    },
    {
      "epoch": 19.0,
      "learning_rate": 4.4894894894894896e-05,
      "loss": 0.7444,
      "step": 2130
    },
    {
      "epoch": 19.0,
      "learning_rate": 4.4844844844844846e-05,
      "loss": 0.9354,
      "step": 2140
    },
    {
      "epoch": 19.0,
      "learning_rate": 4.4794794794794796e-05,
      "loss": 1.1357,
      "step": 2150
    },
    {
      "epoch": 19.0,
      "learning_rate": 4.4744744744744746e-05,
      "loss": 1.2744,
      "step": 2160
    },
    {
      "epoch": 19.0,
      "learning_rate": 4.4694694694694696e-05,
      "loss": 0.9527,
      "step": 2170
    },
    {
      "epoch": 19.0,
      "learning_rate": 4.4644644644644646e-05,
      "loss": 1.0036,
      "step": 2180
    },
    {
      "epoch": 19.01,
      "learning_rate": 4.4594594594594596e-05,
      "loss": 0.7309,
      "step": 2190
    },
    {
      "epoch": 19.01,
      "learning_rate": 4.4544544544544546e-05,
      "loss": 0.8148,
      "step": 2200
    },
    {
      "epoch": 19.01,
      "learning_rate": 4.4494494494494497e-05,
      "loss": 0.697,
      "step": 2210
    },
    {
      "epoch": 19.01,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 1.006,
      "step": 2220
    },
    {
      "epoch": 19.01,
      "learning_rate": 4.43943943943944e-05,
      "loss": 0.9584,
      "step": 2230
    },
    {
      "epoch": 19.01,
      "learning_rate": 4.434434434434435e-05,
      "loss": 0.9575,
      "step": 2240
    },
    {
      "epoch": 19.01,
      "eval_accuracy": 0.7096774193548387,
      "eval_loss": 0.9028933048248291,
      "eval_runtime": 304.4293,
      "eval_samples_per_second": 0.713,
      "eval_steps_per_second": 0.181,
      "step": 2240
    },
    {
      "epoch": 20.0,
      "learning_rate": 4.42942942942943e-05,
      "loss": 0.9425,
      "step": 2250
    },
    {
      "epoch": 20.0,
      "learning_rate": 4.424424424424425e-05,
      "loss": 0.9152,
      "step": 2260
    },
    {
      "epoch": 20.0,
      "learning_rate": 4.41941941941942e-05,
      "loss": 0.7259,
      "step": 2270
    },
    {
      "epoch": 20.0,
      "learning_rate": 4.414414414414415e-05,
      "loss": 1.0089,
      "step": 2280
    },
    {
      "epoch": 20.0,
      "learning_rate": 4.40940940940941e-05,
      "loss": 0.737,
      "step": 2290
    },
    {
      "epoch": 20.01,
      "learning_rate": 4.404404404404405e-05,
      "loss": 0.9018,
      "step": 2300
    },
    {
      "epoch": 20.01,
      "learning_rate": 4.3993993993994e-05,
      "loss": 0.7937,
      "step": 2310
    },
    {
      "epoch": 20.01,
      "learning_rate": 4.394394394394394e-05,
      "loss": 0.9109,
      "step": 2320
    },
    {
      "epoch": 20.01,
      "learning_rate": 4.38938938938939e-05,
      "loss": 1.1044,
      "step": 2330
    },
    {
      "epoch": 20.01,
      "learning_rate": 4.384384384384385e-05,
      "loss": 0.9854,
      "step": 2340
    },
    {
      "epoch": 20.01,
      "learning_rate": 4.37937937937938e-05,
      "loss": 0.8233,
      "step": 2350
    },
    {
      "epoch": 20.01,
      "eval_accuracy": 0.5161290322580645,
      "eval_loss": 1.2444663047790527,
      "eval_runtime": 259.9613,
      "eval_samples_per_second": 0.835,
      "eval_steps_per_second": 0.212,
      "step": 2352
    },
    {
      "epoch": 21.0,
      "learning_rate": 4.374374374374375e-05,
      "loss": 0.9704,
      "step": 2360
    },
    {
      "epoch": 21.0,
      "learning_rate": 4.369369369369369e-05,
      "loss": 0.6429,
      "step": 2370
    },
    {
      "epoch": 21.0,
      "learning_rate": 4.364364364364365e-05,
      "loss": 0.9135,
      "step": 2380
    },
    {
      "epoch": 21.0,
      "learning_rate": 4.35935935935936e-05,
      "loss": 0.7036,
      "step": 2390
    },
    {
      "epoch": 21.0,
      "learning_rate": 4.354354354354355e-05,
      "loss": 0.8668,
      "step": 2400
    },
    {
      "epoch": 21.01,
      "learning_rate": 4.34934934934935e-05,
      "loss": 0.8978,
      "step": 2410
    },
    {
      "epoch": 21.01,
      "learning_rate": 4.344344344344344e-05,
      "loss": 0.5814,
      "step": 2420
    },
    {
      "epoch": 21.01,
      "learning_rate": 4.339339339339339e-05,
      "loss": 1.0719,
      "step": 2430
    },
    {
      "epoch": 21.01,
      "learning_rate": 4.334334334334335e-05,
      "loss": 0.9802,
      "step": 2440
    },
    {
      "epoch": 21.01,
      "learning_rate": 4.32932932932933e-05,
      "loss": 0.5188,
      "step": 2450
    },
    {
      "epoch": 21.01,
      "learning_rate": 4.324324324324325e-05,
      "loss": 0.7643,
      "step": 2460
    },
    {
      "epoch": 21.01,
      "eval_accuracy": 0.6497695852534562,
      "eval_loss": 0.9557706117630005,
      "eval_runtime": 298.2159,
      "eval_samples_per_second": 0.728,
      "eval_steps_per_second": 0.184,
      "step": 2464
    },
    {
      "epoch": 22.0,
      "learning_rate": 4.319319319319319e-05,
      "loss": 0.7597,
      "step": 2470
    },
    {
      "epoch": 22.0,
      "learning_rate": 4.314314314314314e-05,
      "loss": 0.7568,
      "step": 2480
    },
    {
      "epoch": 22.0,
      "learning_rate": 4.30930930930931e-05,
      "loss": 0.5606,
      "step": 2490
    },
    {
      "epoch": 22.0,
      "learning_rate": 4.304304304304305e-05,
      "loss": 0.6645,
      "step": 2500
    },
    {
      "epoch": 22.0,
      "learning_rate": 4.2992992992993e-05,
      "loss": 0.7971,
      "step": 2510
    },
    {
      "epoch": 22.01,
      "learning_rate": 4.294294294294294e-05,
      "loss": 1.1073,
      "step": 2520
    },
    {
      "epoch": 22.01,
      "learning_rate": 4.289289289289289e-05,
      "loss": 1.0306,
      "step": 2530
    },
    {
      "epoch": 22.01,
      "learning_rate": 4.284284284284284e-05,
      "loss": 0.6571,
      "step": 2540
    },
    {
      "epoch": 22.01,
      "learning_rate": 4.27927927927928e-05,
      "loss": 1.0657,
      "step": 2550
    },
    {
      "epoch": 22.01,
      "learning_rate": 4.274274274274275e-05,
      "loss": 0.8202,
      "step": 2560
    },
    {
      "epoch": 22.01,
      "learning_rate": 4.269269269269269e-05,
      "loss": 0.6722,
      "step": 2570
    },
    {
      "epoch": 22.01,
      "eval_accuracy": 0.5714285714285714,
      "eval_loss": 1.1864163875579834,
      "eval_runtime": 297.8839,
      "eval_samples_per_second": 0.728,
      "eval_steps_per_second": 0.185,
      "step": 2576
    },
    {
      "epoch": 23.0,
      "learning_rate": 4.264264264264264e-05,
      "loss": 0.7994,
      "step": 2580
    },
    {
      "epoch": 23.0,
      "learning_rate": 4.259259259259259e-05,
      "loss": 0.7551,
      "step": 2590
    },
    {
      "epoch": 23.0,
      "learning_rate": 4.254254254254255e-05,
      "loss": 0.8671,
      "step": 2600
    },
    {
      "epoch": 23.0,
      "learning_rate": 4.24924924924925e-05,
      "loss": 0.771,
      "step": 2610
    },
    {
      "epoch": 23.0,
      "learning_rate": 4.244244244244244e-05,
      "loss": 0.6954,
      "step": 2620
    },
    {
      "epoch": 23.0,
      "learning_rate": 4.239239239239239e-05,
      "loss": 0.954,
      "step": 2630
    },
    {
      "epoch": 23.01,
      "learning_rate": 4.234234234234234e-05,
      "loss": 0.5557,
      "step": 2640
    },
    {
      "epoch": 23.01,
      "learning_rate": 4.229229229229229e-05,
      "loss": 0.5324,
      "step": 2650
    },
    {
      "epoch": 23.01,
      "learning_rate": 4.224224224224225e-05,
      "loss": 0.9089,
      "step": 2660
    },
    {
      "epoch": 23.01,
      "learning_rate": 4.219219219219219e-05,
      "loss": 0.8594,
      "step": 2670
    },
    {
      "epoch": 23.01,
      "learning_rate": 4.214214214214214e-05,
      "loss": 0.8441,
      "step": 2680
    },
    {
      "epoch": 23.01,
      "eval_accuracy": 0.7235023041474654,
      "eval_loss": 0.9690277576446533,
      "eval_runtime": 270.313,
      "eval_samples_per_second": 0.803,
      "eval_steps_per_second": 0.203,
      "step": 2688
    },
    {
      "epoch": 24.0,
      "learning_rate": 4.209209209209209e-05,
      "loss": 0.6466,
      "step": 2690
    },
    {
      "epoch": 24.0,
      "learning_rate": 4.204204204204204e-05,
      "loss": 0.8761,
      "step": 2700
    },
    {
      "epoch": 24.0,
      "learning_rate": 4.199199199199199e-05,
      "loss": 0.7452,
      "step": 2710
    },
    {
      "epoch": 24.0,
      "learning_rate": 4.194194194194194e-05,
      "loss": 0.7736,
      "step": 2720
    },
    {
      "epoch": 24.0,
      "learning_rate": 4.189189189189189e-05,
      "loss": 0.617,
      "step": 2730
    },
    {
      "epoch": 24.0,
      "learning_rate": 4.1841841841841843e-05,
      "loss": 0.7974,
      "step": 2740
    },
    {
      "epoch": 24.01,
      "learning_rate": 4.1791791791791793e-05,
      "loss": 0.8033,
      "step": 2750
    },
    {
      "epoch": 24.01,
      "learning_rate": 4.1741741741741744e-05,
      "loss": 0.8278,
      "step": 2760
    },
    {
      "epoch": 24.01,
      "learning_rate": 4.1691691691691694e-05,
      "loss": 0.8947,
      "step": 2770
    },
    {
      "epoch": 24.01,
      "learning_rate": 4.1641641641641644e-05,
      "loss": 0.9541,
      "step": 2780
    },
    {
      "epoch": 24.01,
      "learning_rate": 4.1591591591591594e-05,
      "loss": 0.656,
      "step": 2790
    },
    {
      "epoch": 24.01,
      "learning_rate": 4.1541541541541544e-05,
      "loss": 0.7971,
      "step": 2800
    },
    {
      "epoch": 24.01,
      "eval_accuracy": 0.6774193548387096,
      "eval_loss": 0.9349454045295715,
      "eval_runtime": 296.9226,
      "eval_samples_per_second": 0.731,
      "eval_steps_per_second": 0.185,
      "step": 2800
    },
    {
      "epoch": 25.0,
      "learning_rate": 4.1491491491491494e-05,
      "loss": 0.6172,
      "step": 2810
    },
    {
      "epoch": 25.0,
      "learning_rate": 4.1441441441441444e-05,
      "loss": 0.735,
      "step": 2820
    },
    {
      "epoch": 25.0,
      "learning_rate": 4.1391391391391394e-05,
      "loss": 0.6726,
      "step": 2830
    },
    {
      "epoch": 25.0,
      "learning_rate": 4.1341341341341344e-05,
      "loss": 0.6413,
      "step": 2840
    },
    {
      "epoch": 25.0,
      "learning_rate": 4.1291291291291294e-05,
      "loss": 0.936,
      "step": 2850
    },
    {
      "epoch": 25.01,
      "learning_rate": 4.124124124124124e-05,
      "loss": 0.6391,
      "step": 2860
    },
    {
      "epoch": 25.01,
      "learning_rate": 4.1191191191191194e-05,
      "loss": 0.4554,
      "step": 2870
    },
    {
      "epoch": 25.01,
      "learning_rate": 4.1141141141141144e-05,
      "loss": 0.8284,
      "step": 2880
    },
    {
      "epoch": 25.01,
      "learning_rate": 4.1091091091091094e-05,
      "loss": 0.5659,
      "step": 2890
    },
    {
      "epoch": 25.01,
      "learning_rate": 4.1041041041041045e-05,
      "loss": 1.0947,
      "step": 2900
    },
    {
      "epoch": 25.01,
      "learning_rate": 4.099099099099099e-05,
      "loss": 0.8296,
      "step": 2910
    },
    {
      "epoch": 25.01,
      "eval_accuracy": 0.45161290322580644,
      "eval_loss": 1.4573835134506226,
      "eval_runtime": 259.4536,
      "eval_samples_per_second": 0.836,
      "eval_steps_per_second": 0.212,
      "step": 2912
    },
    {
      "epoch": 26.0,
      "learning_rate": 4.0940940940940945e-05,
      "loss": 0.9093,
      "step": 2920
    },
    {
      "epoch": 26.0,
      "learning_rate": 4.0890890890890895e-05,
      "loss": 0.8202,
      "step": 2930
    },
    {
      "epoch": 26.0,
      "learning_rate": 4.0840840840840845e-05,
      "loss": 0.6358,
      "step": 2940
    },
    {
      "epoch": 26.0,
      "learning_rate": 4.0790790790790795e-05,
      "loss": 1.4038,
      "step": 2950
    },
    {
      "epoch": 26.0,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.8683,
      "step": 2960
    },
    {
      "epoch": 26.01,
      "learning_rate": 4.069069069069069e-05,
      "loss": 0.6112,
      "step": 2970
    },
    {
      "epoch": 26.01,
      "learning_rate": 4.0640640640640645e-05,
      "loss": 0.5557,
      "step": 2980
    },
    {
      "epoch": 26.01,
      "learning_rate": 4.0590590590590595e-05,
      "loss": 0.9164,
      "step": 2990
    },
    {
      "epoch": 26.01,
      "learning_rate": 4.0540540540540545e-05,
      "loss": 0.7501,
      "step": 3000
    },
    {
      "epoch": 26.01,
      "learning_rate": 4.049049049049049e-05,
      "loss": 0.5944,
      "step": 3010
    },
    {
      "epoch": 26.01,
      "learning_rate": 4.044044044044044e-05,
      "loss": 0.8613,
      "step": 3020
    },
    {
      "epoch": 26.01,
      "eval_accuracy": 0.7188940092165899,
      "eval_loss": 0.8688245415687561,
      "eval_runtime": 295.9383,
      "eval_samples_per_second": 0.733,
      "eval_steps_per_second": 0.186,
      "step": 3024
    },
    {
      "epoch": 27.0,
      "learning_rate": 4.0390390390390395e-05,
      "loss": 0.6532,
      "step": 3030
    },
    {
      "epoch": 27.0,
      "learning_rate": 4.0340340340340346e-05,
      "loss": 0.4861,
      "step": 3040
    },
    {
      "epoch": 27.0,
      "learning_rate": 4.0290290290290296e-05,
      "loss": 0.7092,
      "step": 3050
    },
    {
      "epoch": 27.0,
      "learning_rate": 4.024024024024024e-05,
      "loss": 0.757,
      "step": 3060
    },
    {
      "epoch": 27.0,
      "learning_rate": 4.019019019019019e-05,
      "loss": 0.9267,
      "step": 3070
    },
    {
      "epoch": 27.01,
      "learning_rate": 4.014014014014014e-05,
      "loss": 0.9786,
      "step": 3080
    },
    {
      "epoch": 27.01,
      "learning_rate": 4.0090090090090096e-05,
      "loss": 0.8151,
      "step": 3090
    },
    {
      "epoch": 27.01,
      "learning_rate": 4.0040040040040046e-05,
      "loss": 1.0989,
      "step": 3100
    },
    {
      "epoch": 27.01,
      "learning_rate": 3.998998998998999e-05,
      "loss": 0.5852,
      "step": 3110
    },
    {
      "epoch": 27.01,
      "learning_rate": 3.993993993993994e-05,
      "loss": 0.7241,
      "step": 3120
    },
    {
      "epoch": 27.01,
      "learning_rate": 3.988988988988989e-05,
      "loss": 0.5614,
      "step": 3130
    },
    {
      "epoch": 27.01,
      "eval_accuracy": 0.6082949308755761,
      "eval_loss": 1.2101435661315918,
      "eval_runtime": 300.7051,
      "eval_samples_per_second": 0.722,
      "eval_steps_per_second": 0.183,
      "step": 3136
    },
    {
      "epoch": 28.0,
      "learning_rate": 3.9839839839839846e-05,
      "loss": 0.9899,
      "step": 3140
    },
    {
      "epoch": 28.0,
      "learning_rate": 3.9789789789789796e-05,
      "loss": 0.5849,
      "step": 3150
    },
    {
      "epoch": 28.0,
      "learning_rate": 3.973973973973974e-05,
      "loss": 0.589,
      "step": 3160
    },
    {
      "epoch": 28.0,
      "learning_rate": 3.968968968968969e-05,
      "loss": 0.8114,
      "step": 3170
    },
    {
      "epoch": 28.0,
      "learning_rate": 3.963963963963964e-05,
      "loss": 0.821,
      "step": 3180
    },
    {
      "epoch": 28.0,
      "learning_rate": 3.958958958958959e-05,
      "loss": 0.7074,
      "step": 3190
    },
    {
      "epoch": 28.01,
      "learning_rate": 3.953953953953955e-05,
      "loss": 0.9248,
      "step": 3200
    },
    {
      "epoch": 28.01,
      "learning_rate": 3.948948948948949e-05,
      "loss": 0.7385,
      "step": 3210
    },
    {
      "epoch": 28.01,
      "learning_rate": 3.943943943943944e-05,
      "loss": 1.1212,
      "step": 3220
    },
    {
      "epoch": 28.01,
      "learning_rate": 3.938938938938939e-05,
      "loss": 0.8575,
      "step": 3230
    },
    {
      "epoch": 28.01,
      "learning_rate": 3.933933933933934e-05,
      "loss": 0.6971,
      "step": 3240
    },
    {
      "epoch": 28.01,
      "eval_accuracy": 0.46543778801843316,
      "eval_loss": 1.300562858581543,
      "eval_runtime": 265.8911,
      "eval_samples_per_second": 0.816,
      "eval_steps_per_second": 0.207,
      "step": 3248
    },
    {
      "epoch": 29.0,
      "learning_rate": 3.92892892892893e-05,
      "loss": 1.1448,
      "step": 3250
    },
    {
      "epoch": 29.0,
      "learning_rate": 3.923923923923924e-05,
      "loss": 0.4975,
      "step": 3260
    },
    {
      "epoch": 29.0,
      "learning_rate": 3.918918918918919e-05,
      "loss": 0.5939,
      "step": 3270
    },
    {
      "epoch": 29.0,
      "learning_rate": 3.913913913913914e-05,
      "loss": 0.842,
      "step": 3280
    },
    {
      "epoch": 29.0,
      "learning_rate": 3.908908908908909e-05,
      "loss": 0.7987,
      "step": 3290
    },
    {
      "epoch": 29.0,
      "learning_rate": 3.903903903903904e-05,
      "loss": 0.7503,
      "step": 3300
    },
    {
      "epoch": 29.01,
      "learning_rate": 3.898898898898899e-05,
      "loss": 0.7446,
      "step": 3310
    },
    {
      "epoch": 29.01,
      "learning_rate": 3.893893893893894e-05,
      "loss": 0.756,
      "step": 3320
    },
    {
      "epoch": 29.01,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.8646,
      "step": 3330
    },
    {
      "epoch": 29.01,
      "learning_rate": 3.883883883883884e-05,
      "loss": 0.822,
      "step": 3340
    },
    {
      "epoch": 29.01,
      "learning_rate": 3.878878878878879e-05,
      "loss": 0.7897,
      "step": 3350
    },
    {
      "epoch": 29.01,
      "learning_rate": 3.873873873873874e-05,
      "loss": 0.9642,
      "step": 3360
    },
    {
      "epoch": 29.01,
      "eval_accuracy": 0.631336405529954,
      "eval_loss": 0.957271158695221,
      "eval_runtime": 301.681,
      "eval_samples_per_second": 0.719,
      "eval_steps_per_second": 0.182,
      "step": 3360
    },
    {
      "epoch": 30.0,
      "learning_rate": 3.868868868868869e-05,
      "loss": 0.6378,
      "step": 3370
    },
    {
      "epoch": 30.0,
      "learning_rate": 3.863863863863864e-05,
      "loss": 0.8326,
      "step": 3380
    },
    {
      "epoch": 30.0,
      "learning_rate": 3.858858858858859e-05,
      "loss": 0.7346,
      "step": 3390
    },
    {
      "epoch": 30.0,
      "learning_rate": 3.8538538538538534e-05,
      "loss": 0.8719,
      "step": 3400
    },
    {
      "epoch": 30.0,
      "learning_rate": 3.848848848848849e-05,
      "loss": 0.4954,
      "step": 3410
    },
    {
      "epoch": 30.01,
      "learning_rate": 3.843843843843844e-05,
      "loss": 1.4229,
      "step": 3420
    },
    {
      "epoch": 30.01,
      "learning_rate": 3.838838838838839e-05,
      "loss": 0.8864,
      "step": 3430
    },
    {
      "epoch": 30.01,
      "learning_rate": 3.833833833833834e-05,
      "loss": 0.7032,
      "step": 3440
    },
    {
      "epoch": 30.01,
      "learning_rate": 3.8288288288288285e-05,
      "loss": 0.52,
      "step": 3450
    },
    {
      "epoch": 30.01,
      "learning_rate": 3.823823823823824e-05,
      "loss": 0.5118,
      "step": 3460
    },
    {
      "epoch": 30.01,
      "learning_rate": 3.818818818818819e-05,
      "loss": 0.836,
      "step": 3470
    },
    {
      "epoch": 30.01,
      "eval_accuracy": 0.6221198156682027,
      "eval_loss": 1.1268316507339478,
      "eval_runtime": 289.9073,
      "eval_samples_per_second": 0.749,
      "eval_steps_per_second": 0.19,
      "step": 3472
    },
    {
      "epoch": 31.0,
      "learning_rate": 3.813813813813814e-05,
      "loss": 0.6036,
      "step": 3480
    },
    {
      "epoch": 31.0,
      "learning_rate": 3.808808808808809e-05,
      "loss": 1.1195,
      "step": 3490
    },
    {
      "epoch": 31.0,
      "learning_rate": 3.8038038038038035e-05,
      "loss": 0.5791,
      "step": 3500
    },
    {
      "epoch": 31.0,
      "learning_rate": 3.7987987987987985e-05,
      "loss": 0.6039,
      "step": 3510
    },
    {
      "epoch": 31.0,
      "learning_rate": 3.793793793793794e-05,
      "loss": 0.8095,
      "step": 3520
    },
    {
      "epoch": 31.01,
      "learning_rate": 3.788788788788789e-05,
      "loss": 0.8178,
      "step": 3530
    },
    {
      "epoch": 31.01,
      "learning_rate": 3.783783783783784e-05,
      "loss": 1.014,
      "step": 3540
    },
    {
      "epoch": 31.01,
      "learning_rate": 3.7787787787787786e-05,
      "loss": 0.7561,
      "step": 3550
    },
    {
      "epoch": 31.01,
      "learning_rate": 3.7737737737737736e-05,
      "loss": 0.7431,
      "step": 3560
    },
    {
      "epoch": 31.01,
      "learning_rate": 3.768768768768769e-05,
      "loss": 0.4614,
      "step": 3570
    },
    {
      "epoch": 31.01,
      "learning_rate": 3.763763763763764e-05,
      "loss": 0.7166,
      "step": 3580
    },
    {
      "epoch": 31.01,
      "eval_accuracy": 0.5622119815668203,
      "eval_loss": 1.2384337186813354,
      "eval_runtime": 267.4392,
      "eval_samples_per_second": 0.811,
      "eval_steps_per_second": 0.206,
      "step": 3584
    },
    {
      "epoch": 32.0,
      "learning_rate": 3.758758758758759e-05,
      "loss": 0.5924,
      "step": 3590
    },
    {
      "epoch": 32.0,
      "learning_rate": 3.7537537537537536e-05,
      "loss": 0.8062,
      "step": 3600
    },
    {
      "epoch": 32.0,
      "learning_rate": 3.7487487487487486e-05,
      "loss": 0.6028,
      "step": 3610
    },
    {
      "epoch": 32.0,
      "learning_rate": 3.7437437437437436e-05,
      "loss": 0.7778,
      "step": 3620
    },
    {
      "epoch": 32.0,
      "learning_rate": 3.738738738738739e-05,
      "loss": 0.7601,
      "step": 3630
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.733733733733734e-05,
      "loss": 0.6678,
      "step": 3640
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.7287287287287286e-05,
      "loss": 0.8642,
      "step": 3650
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.7237237237237236e-05,
      "loss": 0.996,
      "step": 3660
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.7187187187187186e-05,
      "loss": 1.0843,
      "step": 3670
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.713713713713714e-05,
      "loss": 0.5206,
      "step": 3680
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.708708708708709e-05,
      "loss": 0.9302,
      "step": 3690
    },
    {
      "epoch": 32.01,
      "eval_accuracy": 0.5990783410138248,
      "eval_loss": 1.062028408050537,
      "eval_runtime": 313.3562,
      "eval_samples_per_second": 0.693,
      "eval_steps_per_second": 0.176,
      "step": 3696
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.9226,
      "step": 3700
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.698698698698699e-05,
      "loss": 0.9627,
      "step": 3710
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.693693693693694e-05,
      "loss": 0.8562,
      "step": 3720
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.688688688688689e-05,
      "loss": 0.6863,
      "step": 3730
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.6836836836836844e-05,
      "loss": 0.7831,
      "step": 3740
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.678678678678679e-05,
      "loss": 0.5242,
      "step": 3750
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.673673673673674e-05,
      "loss": 0.8605,
      "step": 3760
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.668668668668669e-05,
      "loss": 0.7769,
      "step": 3770
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.663663663663664e-05,
      "loss": 0.4738,
      "step": 3780
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.6586586586586594e-05,
      "loss": 0.6548,
      "step": 3790
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.653653653653654e-05,
      "loss": 0.7729,
      "step": 3800
    },
    {
      "epoch": 33.01,
      "eval_accuracy": 0.5622119815668203,
      "eval_loss": 1.325316309928894,
      "eval_runtime": 296.7797,
      "eval_samples_per_second": 0.731,
      "eval_steps_per_second": 0.185,
      "step": 3808
    },
    {
      "epoch": 34.0,
      "learning_rate": 3.648648648648649e-05,
      "loss": 0.5793,
      "step": 3810
    },
    {
      "epoch": 34.0,
      "learning_rate": 3.643643643643644e-05,
      "loss": 0.7713,
      "step": 3820
    },
    {
      "epoch": 34.0,
      "learning_rate": 3.638638638638639e-05,
      "loss": 0.8823,
      "step": 3830
    },
    {
      "epoch": 34.0,
      "learning_rate": 3.633633633633634e-05,
      "loss": 0.7171,
      "step": 3840
    },
    {
      "epoch": 34.0,
      "learning_rate": 3.628628628628629e-05,
      "loss": 1.0916,
      "step": 3850
    },
    {
      "epoch": 34.0,
      "learning_rate": 3.623623623623624e-05,
      "loss": 0.7068,
      "step": 3860
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.618618618618619e-05,
      "loss": 0.538,
      "step": 3870
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.613613613613614e-05,
      "loss": 0.6683,
      "step": 3880
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.608608608608609e-05,
      "loss": 0.6505,
      "step": 3890
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.603603603603604e-05,
      "loss": 0.8202,
      "step": 3900
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.598598598598599e-05,
      "loss": 0.7419,
      "step": 3910
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.593593593593594e-05,
      "loss": 0.8005,
      "step": 3920
    },
    {
      "epoch": 34.01,
      "eval_accuracy": 0.4930875576036866,
      "eval_loss": 1.4979034662246704,
      "eval_runtime": 265.0615,
      "eval_samples_per_second": 0.819,
      "eval_steps_per_second": 0.207,
      "step": 3920
    },
    {
      "epoch": 35.0,
      "learning_rate": 3.588588588588589e-05,
      "loss": 0.5319,
      "step": 3930
    },
    {
      "epoch": 35.0,
      "learning_rate": 3.583583583583583e-05,
      "loss": 0.7364,
      "step": 3940
    },
    {
      "epoch": 35.0,
      "learning_rate": 3.578578578578579e-05,
      "loss": 0.7636,
      "step": 3950
    },
    {
      "epoch": 35.0,
      "learning_rate": 3.573573573573574e-05,
      "loss": 0.8862,
      "step": 3960
    },
    {
      "epoch": 35.0,
      "learning_rate": 3.568568568568569e-05,
      "loss": 1.0223,
      "step": 3970
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.563563563563564e-05,
      "loss": 0.5304,
      "step": 3980
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.558558558558558e-05,
      "loss": 0.7228,
      "step": 3990
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.553553553553554e-05,
      "loss": 0.9252,
      "step": 4000
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.548548548548549e-05,
      "loss": 0.7089,
      "step": 4010
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.543543543543544e-05,
      "loss": 1.0647,
      "step": 4020
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.538538538538539e-05,
      "loss": 0.8025,
      "step": 4030
    },
    {
      "epoch": 35.01,
      "eval_accuracy": 0.5668202764976958,
      "eval_loss": 0.9785929918289185,
      "eval_runtime": 315.0409,
      "eval_samples_per_second": 0.689,
      "eval_steps_per_second": 0.175,
      "step": 4032
    },
    {
      "epoch": 36.0,
      "learning_rate": 3.533533533533533e-05,
      "loss": 0.5936,
      "step": 4040
    },
    {
      "epoch": 36.0,
      "learning_rate": 3.528528528528528e-05,
      "loss": 0.5975,
      "step": 4050
    },
    {
      "epoch": 36.0,
      "learning_rate": 3.523523523523524e-05,
      "loss": 0.8373,
      "step": 4060
    },
    {
      "epoch": 36.0,
      "learning_rate": 3.518518518518519e-05,
      "loss": 0.8668,
      "step": 4070
    },
    {
      "epoch": 36.0,
      "learning_rate": 3.513513513513514e-05,
      "loss": 0.6539,
      "step": 4080
    },
    {
      "epoch": 36.01,
      "learning_rate": 3.508508508508508e-05,
      "loss": 0.8509,
      "step": 4090
    },
    {
      "epoch": 36.01,
      "learning_rate": 3.503503503503503e-05,
      "loss": 0.8254,
      "step": 4100
    },
    {
      "epoch": 36.01,
      "learning_rate": 3.498498498498499e-05,
      "loss": 0.8809,
      "step": 4110
    },
    {
      "epoch": 36.01,
      "learning_rate": 3.493493493493494e-05,
      "loss": 0.6156,
      "step": 4120
    },
    {
      "epoch": 36.01,
      "learning_rate": 3.488488488488489e-05,
      "loss": 0.5421,
      "step": 4130
    },
    {
      "epoch": 36.01,
      "learning_rate": 3.483483483483483e-05,
      "loss": 0.881,
      "step": 4140
    },
    {
      "epoch": 36.01,
      "eval_accuracy": 0.6543778801843319,
      "eval_loss": 0.847679615020752,
      "eval_runtime": 313.0972,
      "eval_samples_per_second": 0.693,
      "eval_steps_per_second": 0.176,
      "step": 4144
    },
    {
      "epoch": 37.0,
      "learning_rate": 3.478478478478478e-05,
      "loss": 0.4785,
      "step": 4150
    },
    {
      "epoch": 37.0,
      "learning_rate": 3.473473473473473e-05,
      "loss": 0.5836,
      "step": 4160
    },
    {
      "epoch": 37.0,
      "learning_rate": 3.468468468468469e-05,
      "loss": 0.6737,
      "step": 4170
    },
    {
      "epoch": 37.0,
      "learning_rate": 3.463463463463464e-05,
      "loss": 0.5356,
      "step": 4180
    },
    {
      "epoch": 37.0,
      "learning_rate": 3.458458458458458e-05,
      "loss": 0.4705,
      "step": 4190
    },
    {
      "epoch": 37.01,
      "learning_rate": 3.453453453453453e-05,
      "loss": 0.3981,
      "step": 4200
    },
    {
      "epoch": 37.01,
      "learning_rate": 3.448448448448448e-05,
      "loss": 0.8925,
      "step": 4210
    },
    {
      "epoch": 37.01,
      "learning_rate": 3.443443443443444e-05,
      "loss": 0.7896,
      "step": 4220
    },
    {
      "epoch": 37.01,
      "learning_rate": 3.438438438438439e-05,
      "loss": 0.4501,
      "step": 4230
    },
    {
      "epoch": 37.01,
      "learning_rate": 3.4334334334334334e-05,
      "loss": 0.6288,
      "step": 4240
    },
    {
      "epoch": 37.01,
      "learning_rate": 3.4284284284284284e-05,
      "loss": 0.5343,
      "step": 4250
    },
    {
      "epoch": 37.01,
      "eval_accuracy": 0.6543778801843319,
      "eval_loss": 1.3106931447982788,
      "eval_runtime": 287.1044,
      "eval_samples_per_second": 0.756,
      "eval_steps_per_second": 0.192,
      "step": 4256
    },
    {
      "epoch": 38.0,
      "learning_rate": 3.4234234234234234e-05,
      "loss": 0.8355,
      "step": 4260
    },
    {
      "epoch": 38.0,
      "learning_rate": 3.4184184184184184e-05,
      "loss": 1.0089,
      "step": 4270
    },
    {
      "epoch": 38.0,
      "learning_rate": 3.413413413413414e-05,
      "loss": 0.3021,
      "step": 4280
    },
    {
      "epoch": 38.0,
      "learning_rate": 3.4084084084084084e-05,
      "loss": 1.0049,
      "step": 4290
    },
    {
      "epoch": 38.0,
      "learning_rate": 3.4034034034034034e-05,
      "loss": 0.8026,
      "step": 4300
    },
    {
      "epoch": 38.0,
      "learning_rate": 3.3983983983983984e-05,
      "loss": 1.0637,
      "step": 4310
    },
    {
      "epoch": 38.01,
      "learning_rate": 3.3933933933933934e-05,
      "loss": 0.5581,
      "step": 4320
    },
    {
      "epoch": 38.01,
      "learning_rate": 3.388388388388389e-05,
      "loss": 1.043,
      "step": 4330
    },
    {
      "epoch": 38.01,
      "learning_rate": 3.3833833833833834e-05,
      "loss": 0.7903,
      "step": 4340
    },
    {
      "epoch": 38.01,
      "learning_rate": 3.3783783783783784e-05,
      "loss": 0.5935,
      "step": 4350
    },
    {
      "epoch": 38.01,
      "learning_rate": 3.3733733733733734e-05,
      "loss": 0.5611,
      "step": 4360
    },
    {
      "epoch": 38.01,
      "eval_accuracy": 0.6866359447004609,
      "eval_loss": 0.9519971609115601,
      "eval_runtime": 272.965,
      "eval_samples_per_second": 0.795,
      "eval_steps_per_second": 0.201,
      "step": 4368
    },
    {
      "epoch": 39.0,
      "learning_rate": 3.3683683683683684e-05,
      "loss": 0.7632,
      "step": 4370
    },
    {
      "epoch": 39.0,
      "learning_rate": 3.3633633633633635e-05,
      "loss": 0.6651,
      "step": 4380
    },
    {
      "epoch": 39.0,
      "learning_rate": 3.3583583583583585e-05,
      "loss": 0.4778,
      "step": 4390
    },
    {
      "epoch": 39.0,
      "learning_rate": 3.3533533533533535e-05,
      "loss": 0.6968,
      "step": 4400
    },
    {
      "epoch": 39.0,
      "learning_rate": 3.3483483483483485e-05,
      "loss": 0.593,
      "step": 4410
    },
    {
      "epoch": 39.0,
      "learning_rate": 3.3433433433433435e-05,
      "loss": 0.6402,
      "step": 4420
    },
    {
      "epoch": 39.01,
      "learning_rate": 3.3383383383383385e-05,
      "loss": 0.8551,
      "step": 4430
    },
    {
      "epoch": 39.01,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.8956,
      "step": 4440
    },
    {
      "epoch": 39.01,
      "learning_rate": 3.3283283283283285e-05,
      "loss": 1.1142,
      "step": 4450
    },
    {
      "epoch": 39.01,
      "learning_rate": 3.3233233233233235e-05,
      "loss": 0.6806,
      "step": 4460
    },
    {
      "epoch": 39.01,
      "learning_rate": 3.3183183183183185e-05,
      "loss": 0.6406,
      "step": 4470
    },
    {
      "epoch": 39.01,
      "learning_rate": 3.3133133133133135e-05,
      "loss": 0.6824,
      "step": 4480
    },
    {
      "epoch": 39.01,
      "eval_accuracy": 0.728110599078341,
      "eval_loss": 0.790860652923584,
      "eval_runtime": 300.0837,
      "eval_samples_per_second": 0.723,
      "eval_steps_per_second": 0.183,
      "step": 4480
    },
    {
      "epoch": 40.0,
      "learning_rate": 3.3083083083083085e-05,
      "loss": 0.476,
      "step": 4490
    },
    {
      "epoch": 40.0,
      "learning_rate": 3.3033033033033035e-05,
      "loss": 0.8035,
      "step": 4500
    },
    {
      "epoch": 40.0,
      "learning_rate": 3.2982982982982985e-05,
      "loss": 0.4653,
      "step": 4510
    },
    {
      "epoch": 40.0,
      "learning_rate": 3.2932932932932935e-05,
      "loss": 0.735,
      "step": 4520
    },
    {
      "epoch": 40.0,
      "learning_rate": 3.2882882882882886e-05,
      "loss": 0.6043,
      "step": 4530
    },
    {
      "epoch": 40.01,
      "learning_rate": 3.2832832832832836e-05,
      "loss": 0.4144,
      "step": 4540
    },
    {
      "epoch": 40.01,
      "learning_rate": 3.2782782782782786e-05,
      "loss": 0.9656,
      "step": 4550
    },
    {
      "epoch": 40.01,
      "learning_rate": 3.2732732732732736e-05,
      "loss": 0.8462,
      "step": 4560
    },
    {
      "epoch": 40.01,
      "learning_rate": 3.2682682682682686e-05,
      "loss": 0.6785,
      "step": 4570
    },
    {
      "epoch": 40.01,
      "learning_rate": 3.263263263263263e-05,
      "loss": 0.7361,
      "step": 4580
    },
    {
      "epoch": 40.01,
      "learning_rate": 3.258258258258258e-05,
      "loss": 0.6146,
      "step": 4590
    },
    {
      "epoch": 40.01,
      "eval_accuracy": 0.6175115207373272,
      "eval_loss": 1.0886244773864746,
      "eval_runtime": 290.8478,
      "eval_samples_per_second": 0.746,
      "eval_steps_per_second": 0.189,
      "step": 4592
    },
    {
      "epoch": 41.0,
      "learning_rate": 3.2532532532532536e-05,
      "loss": 0.9568,
      "step": 4600
    },
    {
      "epoch": 41.0,
      "learning_rate": 3.2482482482482486e-05,
      "loss": 0.7078,
      "step": 4610
    },
    {
      "epoch": 41.0,
      "learning_rate": 3.2432432432432436e-05,
      "loss": 0.5041,
      "step": 4620
    },
    {
      "epoch": 41.0,
      "learning_rate": 3.238238238238238e-05,
      "loss": 0.636,
      "step": 4630
    },
    {
      "epoch": 41.0,
      "learning_rate": 3.233233233233233e-05,
      "loss": 0.7659,
      "step": 4640
    },
    {
      "epoch": 41.01,
      "learning_rate": 3.2282282282282286e-05,
      "loss": 0.3718,
      "step": 4650
    },
    {
      "epoch": 41.01,
      "learning_rate": 3.2232232232232236e-05,
      "loss": 0.6691,
      "step": 4660
    },
    {
      "epoch": 41.01,
      "learning_rate": 3.2182182182182187e-05,
      "loss": 0.924,
      "step": 4670
    },
    {
      "epoch": 41.01,
      "learning_rate": 3.213213213213213e-05,
      "loss": 0.6134,
      "step": 4680
    },
    {
      "epoch": 41.01,
      "learning_rate": 3.208208208208208e-05,
      "loss": 0.6887,
      "step": 4690
    },
    {
      "epoch": 41.01,
      "learning_rate": 3.203203203203203e-05,
      "loss": 1.0098,
      "step": 4700
    },
    {
      "epoch": 41.01,
      "eval_accuracy": 0.631336405529954,
      "eval_loss": 1.0434428453445435,
      "eval_runtime": 294.0778,
      "eval_samples_per_second": 0.738,
      "eval_steps_per_second": 0.187,
      "step": 4704
    },
    {
      "epoch": 42.0,
      "learning_rate": 3.198198198198199e-05,
      "loss": 0.6864,
      "step": 4710
    },
    {
      "epoch": 42.0,
      "learning_rate": 3.193193193193194e-05,
      "loss": 0.5044,
      "step": 4720
    },
    {
      "epoch": 42.0,
      "learning_rate": 3.188188188188188e-05,
      "loss": 0.4966,
      "step": 4730
    },
    {
      "epoch": 42.0,
      "learning_rate": 3.183183183183183e-05,
      "loss": 0.7376,
      "step": 4740
    },
    {
      "epoch": 42.0,
      "learning_rate": 3.178178178178178e-05,
      "loss": 0.742,
      "step": 4750
    },
    {
      "epoch": 42.01,
      "learning_rate": 3.173173173173174e-05,
      "loss": 0.8865,
      "step": 4760
    },
    {
      "epoch": 42.01,
      "learning_rate": 3.168168168168169e-05,
      "loss": 0.7557,
      "step": 4770
    },
    {
      "epoch": 42.01,
      "learning_rate": 3.163163163163163e-05,
      "loss": 0.4744,
      "step": 4780
    },
    {
      "epoch": 42.01,
      "learning_rate": 3.158158158158158e-05,
      "loss": 0.964,
      "step": 4790
    },
    {
      "epoch": 42.01,
      "learning_rate": 3.153153153153153e-05,
      "loss": 0.5519,
      "step": 4800
    },
    {
      "epoch": 42.01,
      "learning_rate": 3.148148148148148e-05,
      "loss": 0.5555,
      "step": 4810
    },
    {
      "epoch": 42.01,
      "eval_accuracy": 0.6912442396313364,
      "eval_loss": 0.9602895379066467,
      "eval_runtime": 297.4448,
      "eval_samples_per_second": 0.73,
      "eval_steps_per_second": 0.185,
      "step": 4816
    },
    {
      "epoch": 43.0,
      "learning_rate": 3.143143143143144e-05,
      "loss": 0.8269,
      "step": 4820
    },
    {
      "epoch": 43.0,
      "learning_rate": 3.138138138138138e-05,
      "loss": 0.527,
      "step": 4830
    },
    {
      "epoch": 43.0,
      "learning_rate": 3.133133133133133e-05,
      "loss": 0.6083,
      "step": 4840
    },
    {
      "epoch": 43.0,
      "learning_rate": 3.128128128128128e-05,
      "loss": 0.821,
      "step": 4850
    },
    {
      "epoch": 43.0,
      "learning_rate": 3.123123123123123e-05,
      "loss": 0.7753,
      "step": 4860
    },
    {
      "epoch": 43.0,
      "learning_rate": 3.118118118118119e-05,
      "loss": 0.9982,
      "step": 4870
    },
    {
      "epoch": 43.01,
      "learning_rate": 3.113113113113113e-05,
      "loss": 1.0857,
      "step": 4880
    },
    {
      "epoch": 43.01,
      "learning_rate": 3.108108108108108e-05,
      "loss": 0.6124,
      "step": 4890
    },
    {
      "epoch": 43.01,
      "learning_rate": 3.103103103103103e-05,
      "loss": 0.5802,
      "step": 4900
    },
    {
      "epoch": 43.01,
      "learning_rate": 3.098098098098098e-05,
      "loss": 0.4761,
      "step": 4910
    },
    {
      "epoch": 43.01,
      "learning_rate": 3.093093093093093e-05,
      "loss": 0.4578,
      "step": 4920
    },
    {
      "epoch": 43.01,
      "eval_accuracy": 0.5944700460829493,
      "eval_loss": 1.2341152429580688,
      "eval_runtime": 262.4228,
      "eval_samples_per_second": 0.827,
      "eval_steps_per_second": 0.21,
      "step": 4928
    },
    {
      "epoch": 44.0,
      "learning_rate": 3.088088088088088e-05,
      "loss": 0.4575,
      "step": 4930
    },
    {
      "epoch": 44.0,
      "learning_rate": 3.083083083083083e-05,
      "loss": 0.5237,
      "step": 4940
    },
    {
      "epoch": 44.0,
      "learning_rate": 3.078078078078078e-05,
      "loss": 0.6241,
      "step": 4950
    },
    {
      "epoch": 44.0,
      "learning_rate": 3.073073073073073e-05,
      "loss": 0.415,
      "step": 4960
    },
    {
      "epoch": 44.0,
      "learning_rate": 3.068068068068068e-05,
      "loss": 0.6699,
      "step": 4970
    },
    {
      "epoch": 44.0,
      "learning_rate": 3.063063063063063e-05,
      "loss": 0.6522,
      "step": 4980
    },
    {
      "epoch": 44.01,
      "learning_rate": 3.058058058058058e-05,
      "loss": 0.5589,
      "step": 4990
    },
    {
      "epoch": 44.01,
      "learning_rate": 3.053053053053053e-05,
      "loss": 0.5531,
      "step": 5000
    },
    {
      "epoch": 44.01,
      "learning_rate": 3.0480480480480482e-05,
      "loss": 0.9139,
      "step": 5010
    },
    {
      "epoch": 44.01,
      "learning_rate": 3.0430430430430436e-05,
      "loss": 0.5568,
      "step": 5020
    },
    {
      "epoch": 44.01,
      "learning_rate": 3.038038038038038e-05,
      "loss": 0.3951,
      "step": 5030
    },
    {
      "epoch": 44.01,
      "learning_rate": 3.0330330330330332e-05,
      "loss": 0.5883,
      "step": 5040
    },
    {
      "epoch": 44.01,
      "eval_accuracy": 0.6359447004608295,
      "eval_loss": 1.255889654159546,
      "eval_runtime": 292.6708,
      "eval_samples_per_second": 0.741,
      "eval_steps_per_second": 0.188,
      "step": 5040
    },
    {
      "epoch": 45.0,
      "learning_rate": 3.0280280280280282e-05,
      "loss": 0.5707,
      "step": 5050
    },
    {
      "epoch": 45.0,
      "learning_rate": 3.0230230230230232e-05,
      "loss": 0.5603,
      "step": 5060
    },
    {
      "epoch": 45.0,
      "learning_rate": 3.0180180180180183e-05,
      "loss": 0.8447,
      "step": 5070
    },
    {
      "epoch": 45.0,
      "learning_rate": 3.013013013013013e-05,
      "loss": 0.8674,
      "step": 5080
    },
    {
      "epoch": 45.0,
      "learning_rate": 3.008008008008008e-05,
      "loss": 0.8243,
      "step": 5090
    },
    {
      "epoch": 45.01,
      "learning_rate": 3.0030030030030033e-05,
      "loss": 0.5537,
      "step": 5100
    },
    {
      "epoch": 45.01,
      "learning_rate": 2.9979979979979983e-05,
      "loss": 0.9285,
      "step": 5110
    },
    {
      "epoch": 45.01,
      "learning_rate": 2.9929929929929933e-05,
      "loss": 0.8988,
      "step": 5120
    },
    {
      "epoch": 45.01,
      "learning_rate": 2.987987987987988e-05,
      "loss": 0.8101,
      "step": 5130
    },
    {
      "epoch": 45.01,
      "learning_rate": 2.982982982982983e-05,
      "loss": 0.7402,
      "step": 5140
    },
    {
      "epoch": 45.01,
      "learning_rate": 2.9779779779779783e-05,
      "loss": 0.3579,
      "step": 5150
    },
    {
      "epoch": 45.01,
      "eval_accuracy": 0.5622119815668203,
      "eval_loss": 1.2458745241165161,
      "eval_runtime": 296.1464,
      "eval_samples_per_second": 0.733,
      "eval_steps_per_second": 0.186,
      "step": 5152
    },
    {
      "epoch": 46.0,
      "learning_rate": 2.9729729729729733e-05,
      "loss": 0.8051,
      "step": 5160
    },
    {
      "epoch": 46.0,
      "learning_rate": 2.967967967967968e-05,
      "loss": 0.5127,
      "step": 5170
    },
    {
      "epoch": 46.0,
      "learning_rate": 2.962962962962963e-05,
      "loss": 0.6875,
      "step": 5180
    },
    {
      "epoch": 46.0,
      "learning_rate": 2.957957957957958e-05,
      "loss": 0.9627,
      "step": 5190
    },
    {
      "epoch": 46.0,
      "learning_rate": 2.952952952952953e-05,
      "loss": 0.6171,
      "step": 5200
    },
    {
      "epoch": 46.01,
      "learning_rate": 2.9479479479479484e-05,
      "loss": 0.3914,
      "step": 5210
    },
    {
      "epoch": 46.01,
      "learning_rate": 2.9429429429429427e-05,
      "loss": 0.5702,
      "step": 5220
    },
    {
      "epoch": 46.01,
      "learning_rate": 2.937937937937938e-05,
      "loss": 0.7497,
      "step": 5230
    },
    {
      "epoch": 46.01,
      "learning_rate": 2.932932932932933e-05,
      "loss": 0.6988,
      "step": 5240
    },
    {
      "epoch": 46.01,
      "learning_rate": 2.927927927927928e-05,
      "loss": 0.2937,
      "step": 5250
    },
    {
      "epoch": 46.01,
      "learning_rate": 2.9229229229229234e-05,
      "loss": 0.7936,
      "step": 5260
    },
    {
      "epoch": 46.01,
      "eval_accuracy": 0.6082949308755761,
      "eval_loss": 1.2684736251831055,
      "eval_runtime": 266.3687,
      "eval_samples_per_second": 0.815,
      "eval_steps_per_second": 0.206,
      "step": 5264
    },
    {
      "epoch": 47.0,
      "learning_rate": 2.9179179179179177e-05,
      "loss": 1.048,
      "step": 5270
    },
    {
      "epoch": 47.0,
      "learning_rate": 2.912912912912913e-05,
      "loss": 0.8651,
      "step": 5280
    },
    {
      "epoch": 47.0,
      "learning_rate": 2.907907907907908e-05,
      "loss": 0.3689,
      "step": 5290
    },
    {
      "epoch": 47.0,
      "learning_rate": 2.902902902902903e-05,
      "loss": 0.6916,
      "step": 5300
    },
    {
      "epoch": 47.0,
      "learning_rate": 2.897897897897898e-05,
      "loss": 0.6369,
      "step": 5310
    },
    {
      "epoch": 47.01,
      "learning_rate": 2.8928928928928928e-05,
      "loss": 0.5628,
      "step": 5320
    },
    {
      "epoch": 47.01,
      "learning_rate": 2.8878878878878878e-05,
      "loss": 0.8365,
      "step": 5330
    },
    {
      "epoch": 47.01,
      "learning_rate": 2.882882882882883e-05,
      "loss": 0.7938,
      "step": 5340
    },
    {
      "epoch": 47.01,
      "learning_rate": 2.877877877877878e-05,
      "loss": 1.0672,
      "step": 5350
    },
    {
      "epoch": 47.01,
      "learning_rate": 2.872872872872873e-05,
      "loss": 0.646,
      "step": 5360
    },
    {
      "epoch": 47.01,
      "learning_rate": 2.8678678678678678e-05,
      "loss": 0.4331,
      "step": 5370
    },
    {
      "epoch": 47.01,
      "eval_accuracy": 0.7096774193548387,
      "eval_loss": 0.9117954969406128,
      "eval_runtime": 311.9878,
      "eval_samples_per_second": 0.696,
      "eval_steps_per_second": 0.176,
      "step": 5376
    },
    {
      "epoch": 48.0,
      "learning_rate": 2.8628628628628628e-05,
      "loss": 0.7897,
      "step": 5380
    },
    {
      "epoch": 48.0,
      "learning_rate": 2.857857857857858e-05,
      "loss": 0.451,
      "step": 5390
    },
    {
      "epoch": 48.0,
      "learning_rate": 2.852852852852853e-05,
      "loss": 0.6085,
      "step": 5400
    },
    {
      "epoch": 48.0,
      "learning_rate": 2.847847847847848e-05,
      "loss": 0.4599,
      "step": 5410
    },
    {
      "epoch": 48.0,
      "learning_rate": 2.8428428428428428e-05,
      "loss": 0.7639,
      "step": 5420
    },
    {
      "epoch": 48.0,
      "learning_rate": 2.8378378378378378e-05,
      "loss": 0.6014,
      "step": 5430
    },
    {
      "epoch": 48.01,
      "learning_rate": 2.832832832832833e-05,
      "loss": 0.5606,
      "step": 5440
    },
    {
      "epoch": 48.01,
      "learning_rate": 2.8278278278278282e-05,
      "loss": 0.4729,
      "step": 5450
    },
    {
      "epoch": 48.01,
      "learning_rate": 2.8228228228228232e-05,
      "loss": 0.7321,
      "step": 5460
    },
    {
      "epoch": 48.01,
      "learning_rate": 2.817817817817818e-05,
      "loss": 0.437,
      "step": 5470
    },
    {
      "epoch": 48.01,
      "learning_rate": 2.812812812812813e-05,
      "loss": 0.8989,
      "step": 5480
    },
    {
      "epoch": 48.01,
      "eval_accuracy": 0.5806451612903226,
      "eval_loss": 1.3405805826187134,
      "eval_runtime": 290.3011,
      "eval_samples_per_second": 0.747,
      "eval_steps_per_second": 0.189,
      "step": 5488
    },
    {
      "epoch": 49.0,
      "learning_rate": 2.807807807807808e-05,
      "loss": 0.3377,
      "step": 5490
    },
    {
      "epoch": 49.0,
      "learning_rate": 2.8028028028028032e-05,
      "loss": 0.4904,
      "step": 5500
    },
    {
      "epoch": 49.0,
      "learning_rate": 2.7977977977977982e-05,
      "loss": 0.4106,
      "step": 5510
    },
    {
      "epoch": 49.0,
      "learning_rate": 2.7927927927927926e-05,
      "loss": 0.9969,
      "step": 5520
    },
    {
      "epoch": 49.0,
      "learning_rate": 2.787787787787788e-05,
      "loss": 0.5752,
      "step": 5530
    },
    {
      "epoch": 49.0,
      "learning_rate": 2.782782782782783e-05,
      "loss": 0.7738,
      "step": 5540
    },
    {
      "epoch": 49.01,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.5356,
      "step": 5550
    },
    {
      "epoch": 49.01,
      "learning_rate": 2.7727727727727733e-05,
      "loss": 0.3357,
      "step": 5560
    },
    {
      "epoch": 49.01,
      "learning_rate": 2.7677677677677676e-05,
      "loss": 0.6553,
      "step": 5570
    },
    {
      "epoch": 49.01,
      "learning_rate": 2.762762762762763e-05,
      "loss": 0.78,
      "step": 5580
    },
    {
      "epoch": 49.01,
      "learning_rate": 2.757757757757758e-05,
      "loss": 0.85,
      "step": 5590
    },
    {
      "epoch": 49.01,
      "learning_rate": 2.752752752752753e-05,
      "loss": 0.7674,
      "step": 5600
    },
    {
      "epoch": 49.01,
      "eval_accuracy": 0.5483870967741935,
      "eval_loss": 1.52310049533844,
      "eval_runtime": 273.4347,
      "eval_samples_per_second": 0.794,
      "eval_steps_per_second": 0.201,
      "step": 5600
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.7477477477477483e-05,
      "loss": 0.7532,
      "step": 5610
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.7427427427427426e-05,
      "loss": 0.5015,
      "step": 5620
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.7377377377377376e-05,
      "loss": 0.8234,
      "step": 5630
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.732732732732733e-05,
      "loss": 0.4865,
      "step": 5640
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.727727727727728e-05,
      "loss": 0.8966,
      "step": 5650
    },
    {
      "epoch": 50.01,
      "learning_rate": 2.722722722722723e-05,
      "loss": 0.4941,
      "step": 5660
    },
    {
      "epoch": 50.01,
      "learning_rate": 2.7177177177177177e-05,
      "loss": 0.63,
      "step": 5670
    },
    {
      "epoch": 50.01,
      "learning_rate": 2.7127127127127127e-05,
      "loss": 0.9561,
      "step": 5680
    },
    {
      "epoch": 50.01,
      "learning_rate": 2.707707707707708e-05,
      "loss": 0.3966,
      "step": 5690
    },
    {
      "epoch": 50.01,
      "learning_rate": 2.702702702702703e-05,
      "loss": 0.388,
      "step": 5700
    },
    {
      "epoch": 50.01,
      "learning_rate": 2.697697697697698e-05,
      "loss": 0.8136,
      "step": 5710
    },
    {
      "epoch": 50.01,
      "eval_accuracy": 0.6221198156682027,
      "eval_loss": 1.221036672592163,
      "eval_runtime": 307.1022,
      "eval_samples_per_second": 0.707,
      "eval_steps_per_second": 0.179,
      "step": 5712
    },
    {
      "epoch": 51.0,
      "learning_rate": 2.6926926926926927e-05,
      "loss": 0.4861,
      "step": 5720
    },
    {
      "epoch": 51.0,
      "learning_rate": 2.6876876876876877e-05,
      "loss": 0.4606,
      "step": 5730
    },
    {
      "epoch": 51.0,
      "learning_rate": 2.6826826826826827e-05,
      "loss": 0.6793,
      "step": 5740
    },
    {
      "epoch": 51.0,
      "learning_rate": 2.677677677677678e-05,
      "loss": 0.5953,
      "step": 5750
    },
    {
      "epoch": 51.0,
      "learning_rate": 2.672672672672673e-05,
      "loss": 0.3818,
      "step": 5760
    },
    {
      "epoch": 51.01,
      "learning_rate": 2.6676676676676677e-05,
      "loss": 0.9076,
      "step": 5770
    },
    {
      "epoch": 51.01,
      "learning_rate": 2.6626626626626627e-05,
      "loss": 0.3619,
      "step": 5780
    },
    {
      "epoch": 51.01,
      "learning_rate": 2.6576576576576577e-05,
      "loss": 0.6886,
      "step": 5790
    },
    {
      "epoch": 51.01,
      "learning_rate": 2.652652652652653e-05,
      "loss": 0.321,
      "step": 5800
    },
    {
      "epoch": 51.01,
      "learning_rate": 2.6476476476476474e-05,
      "loss": 0.584,
      "step": 5810
    },
    {
      "epoch": 51.01,
      "learning_rate": 2.6426426426426428e-05,
      "loss": 0.6583,
      "step": 5820
    },
    {
      "epoch": 51.01,
      "eval_accuracy": 0.7050691244239631,
      "eval_loss": 0.9261677265167236,
      "eval_runtime": 261.3487,
      "eval_samples_per_second": 0.83,
      "eval_steps_per_second": 0.21,
      "step": 5824
    },
    {
      "epoch": 52.0,
      "learning_rate": 2.6376376376376378e-05,
      "loss": 0.2658,
      "step": 5830
    },
    {
      "epoch": 52.0,
      "learning_rate": 2.6326326326326328e-05,
      "loss": 0.4637,
      "step": 5840
    },
    {
      "epoch": 52.0,
      "learning_rate": 2.6276276276276278e-05,
      "loss": 0.5833,
      "step": 5850
    },
    {
      "epoch": 52.0,
      "learning_rate": 2.6226226226226224e-05,
      "loss": 0.5543,
      "step": 5860
    },
    {
      "epoch": 52.0,
      "learning_rate": 2.6176176176176175e-05,
      "loss": 0.875,
      "step": 5870
    },
    {
      "epoch": 52.01,
      "learning_rate": 2.6126126126126128e-05,
      "loss": 0.8624,
      "step": 5880
    },
    {
      "epoch": 52.01,
      "learning_rate": 2.6076076076076078e-05,
      "loss": 0.476,
      "step": 5890
    },
    {
      "epoch": 52.01,
      "learning_rate": 2.6026026026026028e-05,
      "loss": 0.7822,
      "step": 5900
    },
    {
      "epoch": 52.01,
      "learning_rate": 2.5975975975975975e-05,
      "loss": 0.5592,
      "step": 5910
    },
    {
      "epoch": 52.01,
      "learning_rate": 2.5925925925925925e-05,
      "loss": 0.4731,
      "step": 5920
    },
    {
      "epoch": 52.01,
      "learning_rate": 2.587587587587588e-05,
      "loss": 0.4305,
      "step": 5930
    },
    {
      "epoch": 52.01,
      "eval_accuracy": 0.695852534562212,
      "eval_loss": 1.0339142084121704,
      "eval_runtime": 276.0056,
      "eval_samples_per_second": 0.786,
      "eval_steps_per_second": 0.199,
      "step": 5936
    },
    {
      "epoch": 53.0,
      "learning_rate": 2.582582582582583e-05,
      "loss": 0.5639,
      "step": 5940
    },
    {
      "epoch": 53.0,
      "learning_rate": 2.577577577577578e-05,
      "loss": 0.9295,
      "step": 5950
    },
    {
      "epoch": 53.0,
      "learning_rate": 2.5725725725725725e-05,
      "loss": 1.2619,
      "step": 5960
    },
    {
      "epoch": 53.0,
      "learning_rate": 2.5675675675675675e-05,
      "loss": 0.4272,
      "step": 5970
    },
    {
      "epoch": 53.0,
      "learning_rate": 2.5625625625625625e-05,
      "loss": 0.7564,
      "step": 5980
    },
    {
      "epoch": 53.0,
      "learning_rate": 2.557557557557558e-05,
      "loss": 0.532,
      "step": 5990
    },
    {
      "epoch": 53.01,
      "learning_rate": 2.552552552552553e-05,
      "loss": 0.5171,
      "step": 6000
    },
    {
      "epoch": 53.01,
      "learning_rate": 2.5475475475475476e-05,
      "loss": 0.5332,
      "step": 6010
    },
    {
      "epoch": 53.01,
      "learning_rate": 2.5425425425425426e-05,
      "loss": 0.8714,
      "step": 6020
    },
    {
      "epoch": 53.01,
      "learning_rate": 2.5375375375375376e-05,
      "loss": 0.4191,
      "step": 6030
    },
    {
      "epoch": 53.01,
      "learning_rate": 2.532532532532533e-05,
      "loss": 0.7197,
      "step": 6040
    },
    {
      "epoch": 53.01,
      "eval_accuracy": 0.6682027649769585,
      "eval_loss": 1.1947866678237915,
      "eval_runtime": 300.5473,
      "eval_samples_per_second": 0.722,
      "eval_steps_per_second": 0.183,
      "step": 6048
    },
    {
      "epoch": 54.0,
      "learning_rate": 2.527527527527528e-05,
      "loss": 0.493,
      "step": 6050
    },
    {
      "epoch": 54.0,
      "learning_rate": 2.5225225225225222e-05,
      "loss": 0.4315,
      "step": 6060
    },
    {
      "epoch": 54.0,
      "learning_rate": 2.5175175175175176e-05,
      "loss": 0.5306,
      "step": 6070
    },
    {
      "epoch": 54.0,
      "learning_rate": 2.5125125125125126e-05,
      "loss": 0.7125,
      "step": 6080
    },
    {
      "epoch": 54.0,
      "learning_rate": 2.5075075075075076e-05,
      "loss": 0.4185,
      "step": 6090
    },
    {
      "epoch": 54.0,
      "learning_rate": 2.502502502502503e-05,
      "loss": 0.363,
      "step": 6100
    },
    {
      "epoch": 54.01,
      "learning_rate": 2.4974974974974976e-05,
      "loss": 0.6966,
      "step": 6110
    },
    {
      "epoch": 54.01,
      "learning_rate": 2.4924924924924926e-05,
      "loss": 0.4542,
      "step": 6120
    },
    {
      "epoch": 54.01,
      "learning_rate": 2.4874874874874876e-05,
      "loss": 0.4898,
      "step": 6130
    },
    {
      "epoch": 54.01,
      "learning_rate": 2.4824824824824826e-05,
      "loss": 0.4958,
      "step": 6140
    },
    {
      "epoch": 54.01,
      "learning_rate": 2.4774774774774777e-05,
      "loss": 0.4746,
      "step": 6150
    },
    {
      "epoch": 54.01,
      "learning_rate": 2.4724724724724727e-05,
      "loss": 0.7143,
      "step": 6160
    },
    {
      "epoch": 54.01,
      "eval_accuracy": 0.6774193548387096,
      "eval_loss": 1.185133934020996,
      "eval_runtime": 264.7801,
      "eval_samples_per_second": 0.82,
      "eval_steps_per_second": 0.208,
      "step": 6160
    },
    {
      "epoch": 55.0,
      "learning_rate": 2.4674674674674673e-05,
      "loss": 0.5054,
      "step": 6170
    },
    {
      "epoch": 55.0,
      "learning_rate": 2.4624624624624627e-05,
      "loss": 0.6785,
      "step": 6180
    },
    {
      "epoch": 55.0,
      "learning_rate": 2.4574574574574577e-05,
      "loss": 0.4588,
      "step": 6190
    },
    {
      "epoch": 55.0,
      "learning_rate": 2.4524524524524527e-05,
      "loss": 0.7944,
      "step": 6200
    },
    {
      "epoch": 55.0,
      "learning_rate": 2.4474474474474477e-05,
      "loss": 0.4709,
      "step": 6210
    },
    {
      "epoch": 55.01,
      "learning_rate": 2.4424424424424424e-05,
      "loss": 0.4091,
      "step": 6220
    },
    {
      "epoch": 55.01,
      "learning_rate": 2.4374374374374377e-05,
      "loss": 0.6705,
      "step": 6230
    },
    {
      "epoch": 55.01,
      "learning_rate": 2.4324324324324327e-05,
      "loss": 0.7386,
      "step": 6240
    },
    {
      "epoch": 55.01,
      "learning_rate": 2.4274274274274274e-05,
      "loss": 0.8226,
      "step": 6250
    },
    {
      "epoch": 55.01,
      "learning_rate": 2.4224224224224227e-05,
      "loss": 0.8602,
      "step": 6260
    },
    {
      "epoch": 55.01,
      "learning_rate": 2.4174174174174174e-05,
      "loss": 0.5441,
      "step": 6270
    },
    {
      "epoch": 55.01,
      "eval_accuracy": 0.663594470046083,
      "eval_loss": 1.0351015329360962,
      "eval_runtime": 306.6158,
      "eval_samples_per_second": 0.708,
      "eval_steps_per_second": 0.179,
      "step": 6272
    },
    {
      "epoch": 56.0,
      "learning_rate": 2.4124124124124124e-05,
      "loss": 0.5409,
      "step": 6280
    },
    {
      "epoch": 56.0,
      "learning_rate": 2.4074074074074074e-05,
      "loss": 0.4853,
      "step": 6290
    },
    {
      "epoch": 56.0,
      "learning_rate": 2.4024024024024024e-05,
      "loss": 0.7557,
      "step": 6300
    },
    {
      "epoch": 56.0,
      "learning_rate": 2.3973973973973978e-05,
      "loss": 0.4271,
      "step": 6310
    },
    {
      "epoch": 56.0,
      "learning_rate": 2.3923923923923924e-05,
      "loss": 0.8332,
      "step": 6320
    },
    {
      "epoch": 56.01,
      "learning_rate": 2.3873873873873874e-05,
      "loss": 0.7221,
      "step": 6330
    },
    {
      "epoch": 56.01,
      "learning_rate": 2.3823823823823824e-05,
      "loss": 0.1682,
      "step": 6340
    },
    {
      "epoch": 56.01,
      "learning_rate": 2.3773773773773775e-05,
      "loss": 0.6375,
      "step": 6350
    },
    {
      "epoch": 56.01,
      "learning_rate": 2.3723723723723725e-05,
      "loss": 0.3729,
      "step": 6360
    },
    {
      "epoch": 56.01,
      "learning_rate": 2.3673673673673675e-05,
      "loss": 0.5352,
      "step": 6370
    },
    {
      "epoch": 56.01,
      "learning_rate": 2.3623623623623625e-05,
      "loss": 0.6443,
      "step": 6380
    },
    {
      "epoch": 56.01,
      "eval_accuracy": 0.6866359447004609,
      "eval_loss": 1.0296823978424072,
      "eval_runtime": 299.1803,
      "eval_samples_per_second": 0.725,
      "eval_steps_per_second": 0.184,
      "step": 6384
    },
    {
      "epoch": 57.0,
      "learning_rate": 2.3573573573573575e-05,
      "loss": 0.2025,
      "step": 6390
    },
    {
      "epoch": 57.0,
      "learning_rate": 2.3523523523523525e-05,
      "loss": 0.4155,
      "step": 6400
    },
    {
      "epoch": 57.0,
      "learning_rate": 2.3473473473473475e-05,
      "loss": 0.8092,
      "step": 6410
    },
    {
      "epoch": 57.0,
      "learning_rate": 2.3423423423423425e-05,
      "loss": 0.5598,
      "step": 6420
    },
    {
      "epoch": 57.0,
      "learning_rate": 2.3373373373373375e-05,
      "loss": 0.7998,
      "step": 6430
    },
    {
      "epoch": 57.01,
      "learning_rate": 2.3323323323323322e-05,
      "loss": 0.3185,
      "step": 6440
    },
    {
      "epoch": 57.01,
      "learning_rate": 2.3273273273273275e-05,
      "loss": 0.4758,
      "step": 6450
    },
    {
      "epoch": 57.01,
      "learning_rate": 2.3223223223223225e-05,
      "loss": 0.7854,
      "step": 6460
    },
    {
      "epoch": 57.01,
      "learning_rate": 2.3173173173173175e-05,
      "loss": 0.9117,
      "step": 6470
    },
    {
      "epoch": 57.01,
      "learning_rate": 2.3123123123123125e-05,
      "loss": 0.2464,
      "step": 6480
    },
    {
      "epoch": 57.01,
      "learning_rate": 2.3073073073073072e-05,
      "loss": 0.7747,
      "step": 6490
    },
    {
      "epoch": 57.01,
      "eval_accuracy": 0.5990783410138248,
      "eval_loss": 1.517410159111023,
      "eval_runtime": 265.454,
      "eval_samples_per_second": 0.817,
      "eval_steps_per_second": 0.207,
      "step": 6496
    },
    {
      "epoch": 58.0,
      "learning_rate": 2.3023023023023026e-05,
      "loss": 0.9554,
      "step": 6500
    },
    {
      "epoch": 58.0,
      "learning_rate": 2.2972972972972976e-05,
      "loss": 0.5761,
      "step": 6510
    },
    {
      "epoch": 58.0,
      "learning_rate": 2.2922922922922922e-05,
      "loss": 0.6756,
      "step": 6520
    },
    {
      "epoch": 58.0,
      "learning_rate": 2.2872872872872876e-05,
      "loss": 0.5187,
      "step": 6530
    },
    {
      "epoch": 58.0,
      "learning_rate": 2.2822822822822822e-05,
      "loss": 0.5258,
      "step": 6540
    },
    {
      "epoch": 58.0,
      "learning_rate": 2.2772772772772773e-05,
      "loss": 0.569,
      "step": 6550
    },
    {
      "epoch": 58.01,
      "learning_rate": 2.2722722722722726e-05,
      "loss": 0.4751,
      "step": 6560
    },
    {
      "epoch": 58.01,
      "learning_rate": 2.2672672672672673e-05,
      "loss": 0.5917,
      "step": 6570
    },
    {
      "epoch": 58.01,
      "learning_rate": 2.2622622622622626e-05,
      "loss": 0.8047,
      "step": 6580
    },
    {
      "epoch": 58.01,
      "learning_rate": 2.2572572572572573e-05,
      "loss": 0.5942,
      "step": 6590
    },
    {
      "epoch": 58.01,
      "learning_rate": 2.2522522522522523e-05,
      "loss": 0.5943,
      "step": 6600
    },
    {
      "epoch": 58.01,
      "eval_accuracy": 0.6451612903225806,
      "eval_loss": 1.196057915687561,
      "eval_runtime": 308.0487,
      "eval_samples_per_second": 0.704,
      "eval_steps_per_second": 0.179,
      "step": 6608
    },
    {
      "epoch": 59.0,
      "learning_rate": 2.2472472472472473e-05,
      "loss": 0.6209,
      "step": 6610
    },
    {
      "epoch": 59.0,
      "learning_rate": 2.2422422422422423e-05,
      "loss": 0.3431,
      "step": 6620
    },
    {
      "epoch": 59.0,
      "learning_rate": 2.2372372372372373e-05,
      "loss": 0.2993,
      "step": 6630
    },
    {
      "epoch": 59.0,
      "learning_rate": 2.2322322322322323e-05,
      "loss": 0.3957,
      "step": 6640
    },
    {
      "epoch": 59.0,
      "learning_rate": 2.2272272272272273e-05,
      "loss": 0.6951,
      "step": 6650
    },
    {
      "epoch": 59.0,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.8616,
      "step": 6660
    },
    {
      "epoch": 59.01,
      "learning_rate": 2.2172172172172173e-05,
      "loss": 0.4296,
      "step": 6670
    },
    {
      "epoch": 59.01,
      "learning_rate": 2.2122122122122123e-05,
      "loss": 0.6608,
      "step": 6680
    },
    {
      "epoch": 59.01,
      "learning_rate": 2.2072072072072073e-05,
      "loss": 0.9616,
      "step": 6690
    },
    {
      "epoch": 59.01,
      "learning_rate": 2.2022022022022024e-05,
      "loss": 0.6165,
      "step": 6700
    },
    {
      "epoch": 59.01,
      "learning_rate": 2.197197197197197e-05,
      "loss": 0.5761,
      "step": 6710
    },
    {
      "epoch": 59.01,
      "learning_rate": 2.1921921921921924e-05,
      "loss": 0.5781,
      "step": 6720
    },
    {
      "epoch": 59.01,
      "eval_accuracy": 0.7142857142857143,
      "eval_loss": 1.2187379598617554,
      "eval_runtime": 300.7146,
      "eval_samples_per_second": 0.722,
      "eval_steps_per_second": 0.183,
      "step": 6720
    },
    {
      "epoch": 60.0,
      "learning_rate": 2.1871871871871874e-05,
      "loss": 0.8878,
      "step": 6730
    },
    {
      "epoch": 60.0,
      "learning_rate": 2.1821821821821824e-05,
      "loss": 0.3569,
      "step": 6740
    },
    {
      "epoch": 60.0,
      "learning_rate": 2.1771771771771774e-05,
      "loss": 0.6239,
      "step": 6750
    },
    {
      "epoch": 60.0,
      "learning_rate": 2.172172172172172e-05,
      "loss": 0.6266,
      "step": 6760
    },
    {
      "epoch": 60.0,
      "learning_rate": 2.1671671671671674e-05,
      "loss": 0.8228,
      "step": 6770
    },
    {
      "epoch": 60.01,
      "learning_rate": 2.1621621621621624e-05,
      "loss": 0.2555,
      "step": 6780
    },
    {
      "epoch": 60.01,
      "learning_rate": 2.157157157157157e-05,
      "loss": 0.7252,
      "step": 6790
    },
    {
      "epoch": 60.01,
      "learning_rate": 2.1521521521521524e-05,
      "loss": 0.8275,
      "step": 6800
    },
    {
      "epoch": 60.01,
      "learning_rate": 2.147147147147147e-05,
      "loss": 0.5406,
      "step": 6810
    },
    {
      "epoch": 60.01,
      "learning_rate": 2.142142142142142e-05,
      "loss": 0.3454,
      "step": 6820
    },
    {
      "epoch": 60.01,
      "learning_rate": 2.1371371371371374e-05,
      "loss": 0.6913,
      "step": 6830
    },
    {
      "epoch": 60.01,
      "eval_accuracy": 0.6728110599078341,
      "eval_loss": 1.159045696258545,
      "eval_runtime": 262.2513,
      "eval_samples_per_second": 0.827,
      "eval_steps_per_second": 0.21,
      "step": 6832
    },
    {
      "epoch": 61.0,
      "learning_rate": 2.132132132132132e-05,
      "loss": 0.3399,
      "step": 6840
    },
    {
      "epoch": 61.0,
      "learning_rate": 2.1271271271271275e-05,
      "loss": 0.5646,
      "step": 6850
    },
    {
      "epoch": 61.0,
      "learning_rate": 2.122122122122122e-05,
      "loss": 0.856,
      "step": 6860
    },
    {
      "epoch": 61.0,
      "learning_rate": 2.117117117117117e-05,
      "loss": 0.7805,
      "step": 6870
    },
    {
      "epoch": 61.0,
      "learning_rate": 2.1121121121121125e-05,
      "loss": 0.6905,
      "step": 6880
    },
    {
      "epoch": 61.01,
      "learning_rate": 2.107107107107107e-05,
      "loss": 0.7969,
      "step": 6890
    },
    {
      "epoch": 61.01,
      "learning_rate": 2.102102102102102e-05,
      "loss": 0.4824,
      "step": 6900
    },
    {
      "epoch": 61.01,
      "learning_rate": 2.097097097097097e-05,
      "loss": 0.7744,
      "step": 6910
    },
    {
      "epoch": 61.01,
      "learning_rate": 2.0920920920920922e-05,
      "loss": 0.5511,
      "step": 6920
    },
    {
      "epoch": 61.01,
      "learning_rate": 2.0870870870870872e-05,
      "loss": 0.4663,
      "step": 6930
    },
    {
      "epoch": 61.01,
      "learning_rate": 2.0820820820820822e-05,
      "loss": 0.6186,
      "step": 6940
    },
    {
      "epoch": 61.01,
      "eval_accuracy": 0.7235023041474654,
      "eval_loss": 1.049517035484314,
      "eval_runtime": 308.0162,
      "eval_samples_per_second": 0.705,
      "eval_steps_per_second": 0.179,
      "step": 6944
    },
    {
      "epoch": 62.0,
      "learning_rate": 2.0770770770770772e-05,
      "loss": 0.3485,
      "step": 6950
    },
    {
      "epoch": 62.0,
      "learning_rate": 2.0720720720720722e-05,
      "loss": 0.5408,
      "step": 6960
    },
    {
      "epoch": 62.0,
      "learning_rate": 2.0670670670670672e-05,
      "loss": 0.5365,
      "step": 6970
    },
    {
      "epoch": 62.0,
      "learning_rate": 2.062062062062062e-05,
      "loss": 0.2857,
      "step": 6980
    },
    {
      "epoch": 62.0,
      "learning_rate": 2.0570570570570572e-05,
      "loss": 0.4658,
      "step": 6990
    },
    {
      "epoch": 62.01,
      "learning_rate": 2.0520520520520522e-05,
      "loss": 0.8469,
      "step": 7000
    },
    {
      "epoch": 62.01,
      "learning_rate": 2.0470470470470472e-05,
      "loss": 0.7909,
      "step": 7010
    },
    {
      "epoch": 62.01,
      "learning_rate": 2.0420420420420422e-05,
      "loss": 0.514,
      "step": 7020
    },
    {
      "epoch": 62.01,
      "learning_rate": 2.037037037037037e-05,
      "loss": 0.4968,
      "step": 7030
    },
    {
      "epoch": 62.01,
      "learning_rate": 2.0320320320320323e-05,
      "loss": 0.2844,
      "step": 7040
    },
    {
      "epoch": 62.01,
      "learning_rate": 2.0270270270270273e-05,
      "loss": 0.5185,
      "step": 7050
    },
    {
      "epoch": 62.01,
      "eval_accuracy": 0.7050691244239631,
      "eval_loss": 0.9844396710395813,
      "eval_runtime": 302.6894,
      "eval_samples_per_second": 0.717,
      "eval_steps_per_second": 0.182,
      "step": 7056
    },
    {
      "epoch": 63.0,
      "learning_rate": 2.022022022022022e-05,
      "loss": 0.5965,
      "step": 7060
    },
    {
      "epoch": 63.0,
      "learning_rate": 2.0170170170170173e-05,
      "loss": 1.0196,
      "step": 7070
    },
    {
      "epoch": 63.0,
      "learning_rate": 2.012012012012012e-05,
      "loss": 0.3476,
      "step": 7080
    },
    {
      "epoch": 63.0,
      "learning_rate": 2.007007007007007e-05,
      "loss": 0.5091,
      "step": 7090
    },
    {
      "epoch": 63.0,
      "learning_rate": 2.0020020020020023e-05,
      "loss": 0.5697,
      "step": 7100
    },
    {
      "epoch": 63.0,
      "learning_rate": 1.996996996996997e-05,
      "loss": 0.4926,
      "step": 7110
    },
    {
      "epoch": 63.01,
      "learning_rate": 1.9919919919919923e-05,
      "loss": 0.7321,
      "step": 7120
    },
    {
      "epoch": 63.01,
      "learning_rate": 1.986986986986987e-05,
      "loss": 0.6439,
      "step": 7130
    },
    {
      "epoch": 63.01,
      "learning_rate": 1.981981981981982e-05,
      "loss": 0.4289,
      "step": 7140
    },
    {
      "epoch": 63.01,
      "learning_rate": 1.9769769769769773e-05,
      "loss": 0.7109,
      "step": 7150
    },
    {
      "epoch": 63.01,
      "learning_rate": 1.971971971971972e-05,
      "loss": 0.4077,
      "step": 7160
    },
    {
      "epoch": 63.01,
      "eval_accuracy": 0.631336405529954,
      "eval_loss": 1.319392204284668,
      "eval_runtime": 290.2861,
      "eval_samples_per_second": 0.748,
      "eval_steps_per_second": 0.189,
      "step": 7168
    },
    {
      "epoch": 64.0,
      "learning_rate": 1.966966966966967e-05,
      "loss": 0.662,
      "step": 7170
    },
    {
      "epoch": 64.0,
      "learning_rate": 1.961961961961962e-05,
      "loss": 0.3126,
      "step": 7180
    },
    {
      "epoch": 64.0,
      "learning_rate": 1.956956956956957e-05,
      "loss": 0.364,
      "step": 7190
    },
    {
      "epoch": 64.0,
      "learning_rate": 1.951951951951952e-05,
      "loss": 0.2242,
      "step": 7200
    },
    {
      "epoch": 64.0,
      "learning_rate": 1.946946946946947e-05,
      "loss": 0.5675,
      "step": 7210
    },
    {
      "epoch": 64.0,
      "learning_rate": 1.941941941941942e-05,
      "loss": 0.3905,
      "step": 7220
    },
    {
      "epoch": 64.01,
      "learning_rate": 1.936936936936937e-05,
      "loss": 0.616,
      "step": 7230
    },
    {
      "epoch": 64.01,
      "learning_rate": 1.931931931931932e-05,
      "loss": 0.7498,
      "step": 7240
    },
    {
      "epoch": 64.01,
      "learning_rate": 1.9269269269269267e-05,
      "loss": 0.4339,
      "step": 7250
    },
    {
      "epoch": 64.01,
      "learning_rate": 1.921921921921922e-05,
      "loss": 0.7677,
      "step": 7260
    },
    {
      "epoch": 64.01,
      "learning_rate": 1.916916916916917e-05,
      "loss": 0.7006,
      "step": 7270
    },
    {
      "epoch": 64.01,
      "learning_rate": 1.911911911911912e-05,
      "loss": 0.8217,
      "step": 7280
    },
    {
      "epoch": 64.01,
      "eval_accuracy": 0.663594470046083,
      "eval_loss": 1.2620233297348022,
      "eval_runtime": 273.0488,
      "eval_samples_per_second": 0.795,
      "eval_steps_per_second": 0.201,
      "step": 7280
    },
    {
      "epoch": 65.0,
      "learning_rate": 1.906906906906907e-05,
      "loss": 0.3433,
      "step": 7290
    },
    {
      "epoch": 65.0,
      "learning_rate": 1.9019019019019018e-05,
      "loss": 0.64,
      "step": 7300
    },
    {
      "epoch": 65.0,
      "learning_rate": 1.896896896896897e-05,
      "loss": 0.5905,
      "step": 7310
    },
    {
      "epoch": 65.0,
      "learning_rate": 1.891891891891892e-05,
      "loss": 0.4542,
      "step": 7320
    },
    {
      "epoch": 65.0,
      "learning_rate": 1.8868868868868868e-05,
      "loss": 0.3517,
      "step": 7330
    },
    {
      "epoch": 65.01,
      "learning_rate": 1.881881881881882e-05,
      "loss": 0.7221,
      "step": 7340
    },
    {
      "epoch": 65.01,
      "learning_rate": 1.8768768768768768e-05,
      "loss": 0.7196,
      "step": 7350
    },
    {
      "epoch": 65.01,
      "learning_rate": 1.8718718718718718e-05,
      "loss": 0.4093,
      "step": 7360
    },
    {
      "epoch": 65.01,
      "learning_rate": 1.866866866866867e-05,
      "loss": 0.748,
      "step": 7370
    },
    {
      "epoch": 65.01,
      "learning_rate": 1.8618618618618618e-05,
      "loss": 0.5039,
      "step": 7380
    },
    {
      "epoch": 65.01,
      "learning_rate": 1.856856856856857e-05,
      "loss": 0.5273,
      "step": 7390
    },
    {
      "epoch": 65.01,
      "eval_accuracy": 0.7373271889400922,
      "eval_loss": 1.039459228515625,
      "eval_runtime": 298.6976,
      "eval_samples_per_second": 0.726,
      "eval_steps_per_second": 0.184,
      "step": 7392
    },
    {
      "epoch": 66.0,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.4787,
      "step": 7400
    },
    {
      "epoch": 66.0,
      "learning_rate": 1.846846846846847e-05,
      "loss": 0.9991,
      "step": 7410
    },
    {
      "epoch": 66.0,
      "learning_rate": 1.8418418418418422e-05,
      "loss": 0.7719,
      "step": 7420
    },
    {
      "epoch": 66.0,
      "learning_rate": 1.836836836836837e-05,
      "loss": 0.4417,
      "step": 7430
    },
    {
      "epoch": 66.0,
      "learning_rate": 1.831831831831832e-05,
      "loss": 0.8676,
      "step": 7440
    },
    {
      "epoch": 66.01,
      "learning_rate": 1.826826826826827e-05,
      "loss": 0.2066,
      "step": 7450
    },
    {
      "epoch": 66.01,
      "learning_rate": 1.821821821821822e-05,
      "loss": 0.3548,
      "step": 7460
    },
    {
      "epoch": 66.01,
      "learning_rate": 1.816816816816817e-05,
      "loss": 0.6343,
      "step": 7470
    },
    {
      "epoch": 66.01,
      "learning_rate": 1.811811811811812e-05,
      "loss": 0.674,
      "step": 7480
    },
    {
      "epoch": 66.01,
      "learning_rate": 1.806806806806807e-05,
      "loss": 0.6832,
      "step": 7490
    },
    {
      "epoch": 66.01,
      "learning_rate": 1.801801801801802e-05,
      "loss": 0.9002,
      "step": 7500
    },
    {
      "epoch": 66.01,
      "eval_accuracy": 0.5806451612903226,
      "eval_loss": 1.5225118398666382,
      "eval_runtime": 287.6826,
      "eval_samples_per_second": 0.754,
      "eval_steps_per_second": 0.191,
      "step": 7504
    },
    {
      "epoch": 67.0,
      "learning_rate": 1.796796796796797e-05,
      "loss": 0.5458,
      "step": 7510
    },
    {
      "epoch": 67.0,
      "learning_rate": 1.7917917917917916e-05,
      "loss": 0.3226,
      "step": 7520
    },
    {
      "epoch": 67.0,
      "learning_rate": 1.786786786786787e-05,
      "loss": 0.3272,
      "step": 7530
    },
    {
      "epoch": 67.0,
      "learning_rate": 1.781781781781782e-05,
      "loss": 0.487,
      "step": 7540
    },
    {
      "epoch": 67.0,
      "learning_rate": 1.776776776776777e-05,
      "loss": 0.3854,
      "step": 7550
    },
    {
      "epoch": 67.01,
      "learning_rate": 1.771771771771772e-05,
      "loss": 0.6385,
      "step": 7560
    },
    {
      "epoch": 67.01,
      "learning_rate": 1.7667667667667666e-05,
      "loss": 0.4264,
      "step": 7570
    },
    {
      "epoch": 67.01,
      "learning_rate": 1.761761761761762e-05,
      "loss": 0.4963,
      "step": 7580
    },
    {
      "epoch": 67.01,
      "learning_rate": 1.756756756756757e-05,
      "loss": 0.7433,
      "step": 7590
    },
    {
      "epoch": 67.01,
      "learning_rate": 1.7517517517517516e-05,
      "loss": 0.9013,
      "step": 7600
    },
    {
      "epoch": 67.01,
      "learning_rate": 1.746746746746747e-05,
      "loss": 0.5763,
      "step": 7610
    },
    {
      "epoch": 67.01,
      "eval_accuracy": 0.6405529953917051,
      "eval_loss": 1.2559142112731934,
      "eval_runtime": 272.717,
      "eval_samples_per_second": 0.796,
      "eval_steps_per_second": 0.202,
      "step": 7616
    },
    {
      "epoch": 68.0,
      "learning_rate": 1.7417417417417416e-05,
      "loss": 0.2514,
      "step": 7620
    },
    {
      "epoch": 68.0,
      "learning_rate": 1.7367367367367366e-05,
      "loss": 0.3909,
      "step": 7630
    },
    {
      "epoch": 68.0,
      "learning_rate": 1.731731731731732e-05,
      "loss": 0.6764,
      "step": 7640
    },
    {
      "epoch": 68.0,
      "learning_rate": 1.7267267267267267e-05,
      "loss": 0.26,
      "step": 7650
    },
    {
      "epoch": 68.0,
      "learning_rate": 1.721721721721722e-05,
      "loss": 0.5319,
      "step": 7660
    },
    {
      "epoch": 68.0,
      "learning_rate": 1.7167167167167167e-05,
      "loss": 0.5178,
      "step": 7670
    },
    {
      "epoch": 68.01,
      "learning_rate": 1.7117117117117117e-05,
      "loss": 0.7583,
      "step": 7680
    },
    {
      "epoch": 68.01,
      "learning_rate": 1.706706706706707e-05,
      "loss": 0.359,
      "step": 7690
    },
    {
      "epoch": 68.01,
      "learning_rate": 1.7017017017017017e-05,
      "loss": 0.5274,
      "step": 7700
    },
    {
      "epoch": 68.01,
      "learning_rate": 1.6966966966966967e-05,
      "loss": 0.2864,
      "step": 7710
    },
    {
      "epoch": 68.01,
      "learning_rate": 1.6916916916916917e-05,
      "loss": 1.0535,
      "step": 7720
    },
    {
      "epoch": 68.01,
      "eval_accuracy": 0.6497695852534562,
      "eval_loss": 1.2645872831344604,
      "eval_runtime": 312.941,
      "eval_samples_per_second": 0.693,
      "eval_steps_per_second": 0.176,
      "step": 7728
    },
    {
      "epoch": 69.0,
      "learning_rate": 1.6866866866866867e-05,
      "loss": 0.6536,
      "step": 7730
    },
    {
      "epoch": 69.0,
      "learning_rate": 1.6816816816816817e-05,
      "loss": 0.2713,
      "step": 7740
    },
    {
      "epoch": 69.0,
      "learning_rate": 1.6766766766766767e-05,
      "loss": 0.4356,
      "step": 7750
    },
    {
      "epoch": 69.0,
      "learning_rate": 1.6716716716716717e-05,
      "loss": 0.4133,
      "step": 7760
    },
    {
      "epoch": 69.0,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.4187,
      "step": 7770
    },
    {
      "epoch": 69.0,
      "learning_rate": 1.6616616616616618e-05,
      "loss": 0.6511,
      "step": 7780
    },
    {
      "epoch": 69.01,
      "learning_rate": 1.6566566566566568e-05,
      "loss": 0.6337,
      "step": 7790
    },
    {
      "epoch": 69.01,
      "learning_rate": 1.6516516516516518e-05,
      "loss": 0.4122,
      "step": 7800
    },
    {
      "epoch": 69.01,
      "learning_rate": 1.6466466466466468e-05,
      "loss": 0.5112,
      "step": 7810
    },
    {
      "epoch": 69.01,
      "learning_rate": 1.6416416416416418e-05,
      "loss": 0.68,
      "step": 7820
    },
    {
      "epoch": 69.01,
      "learning_rate": 1.6366366366366368e-05,
      "loss": 0.8665,
      "step": 7830
    },
    {
      "epoch": 69.01,
      "learning_rate": 1.6316316316316315e-05,
      "loss": 1.0064,
      "step": 7840
    },
    {
      "epoch": 69.01,
      "eval_accuracy": 0.6866359447004609,
      "eval_loss": 1.1532840728759766,
      "eval_runtime": 291.6662,
      "eval_samples_per_second": 0.744,
      "eval_steps_per_second": 0.189,
      "step": 7840
    },
    {
      "epoch": 70.0,
      "learning_rate": 1.6266266266266268e-05,
      "loss": 0.42,
      "step": 7850
    },
    {
      "epoch": 70.0,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 0.475,
      "step": 7860
    },
    {
      "epoch": 70.0,
      "learning_rate": 1.6166166166166165e-05,
      "loss": 0.4704,
      "step": 7870
    },
    {
      "epoch": 70.0,
      "learning_rate": 1.6116116116116118e-05,
      "loss": 0.4484,
      "step": 7880
    },
    {
      "epoch": 70.0,
      "learning_rate": 1.6066066066066065e-05,
      "loss": 0.2969,
      "step": 7890
    },
    {
      "epoch": 70.01,
      "learning_rate": 1.6016016016016015e-05,
      "loss": 0.6184,
      "step": 7900
    },
    {
      "epoch": 70.01,
      "learning_rate": 1.596596596596597e-05,
      "loss": 0.5384,
      "step": 7910
    },
    {
      "epoch": 70.01,
      "learning_rate": 1.5915915915915915e-05,
      "loss": 0.4786,
      "step": 7920
    },
    {
      "epoch": 70.01,
      "learning_rate": 1.586586586586587e-05,
      "loss": 0.4313,
      "step": 7930
    },
    {
      "epoch": 70.01,
      "learning_rate": 1.5815815815815815e-05,
      "loss": 0.3226,
      "step": 7940
    },
    {
      "epoch": 70.01,
      "learning_rate": 1.5765765765765765e-05,
      "loss": 0.332,
      "step": 7950
    },
    {
      "epoch": 70.01,
      "eval_accuracy": 0.7004608294930875,
      "eval_loss": 1.0438106060028076,
      "eval_runtime": 269.5471,
      "eval_samples_per_second": 0.805,
      "eval_steps_per_second": 0.204,
      "step": 7952
    },
    {
      "epoch": 71.0,
      "learning_rate": 1.571571571571572e-05,
      "loss": 0.4512,
      "step": 7960
    },
    {
      "epoch": 71.0,
      "learning_rate": 1.5665665665665665e-05,
      "loss": 0.5144,
      "step": 7970
    },
    {
      "epoch": 71.0,
      "learning_rate": 1.5615615615615616e-05,
      "loss": 0.7567,
      "step": 7980
    },
    {
      "epoch": 71.0,
      "learning_rate": 1.5565565565565566e-05,
      "loss": 0.7875,
      "step": 7990
    },
    {
      "epoch": 71.0,
      "learning_rate": 1.5515515515515516e-05,
      "loss": 0.7107,
      "step": 8000
    },
    {
      "epoch": 71.01,
      "learning_rate": 1.5465465465465466e-05,
      "loss": 0.5407,
      "step": 8010
    },
    {
      "epoch": 71.01,
      "learning_rate": 1.5415415415415416e-05,
      "loss": 0.4464,
      "step": 8020
    },
    {
      "epoch": 71.01,
      "learning_rate": 1.5365365365365366e-05,
      "loss": 0.7124,
      "step": 8030
    },
    {
      "epoch": 71.01,
      "learning_rate": 1.5315315315315316e-05,
      "loss": 0.5239,
      "step": 8040
    },
    {
      "epoch": 71.01,
      "learning_rate": 1.5265265265265266e-05,
      "loss": 0.5302,
      "step": 8050
    },
    {
      "epoch": 71.01,
      "learning_rate": 1.5215215215215218e-05,
      "loss": 0.3978,
      "step": 8060
    },
    {
      "epoch": 71.01,
      "eval_accuracy": 0.7050691244239631,
      "eval_loss": 1.024812936782837,
      "eval_runtime": 313.3314,
      "eval_samples_per_second": 0.693,
      "eval_steps_per_second": 0.176,
      "step": 8064
    },
    {
      "epoch": 72.0,
      "learning_rate": 1.5165165165165166e-05,
      "loss": 0.4316,
      "step": 8070
    },
    {
      "epoch": 72.0,
      "learning_rate": 1.5115115115115116e-05,
      "loss": 0.7077,
      "step": 8080
    },
    {
      "epoch": 72.0,
      "learning_rate": 1.5065065065065065e-05,
      "loss": 0.7193,
      "step": 8090
    },
    {
      "epoch": 72.0,
      "learning_rate": 1.5015015015015016e-05,
      "loss": 0.8904,
      "step": 8100
    },
    {
      "epoch": 72.0,
      "learning_rate": 1.4964964964964966e-05,
      "loss": 0.4093,
      "step": 8110
    },
    {
      "epoch": 72.01,
      "learning_rate": 1.4914914914914915e-05,
      "loss": 0.478,
      "step": 8120
    },
    {
      "epoch": 72.01,
      "learning_rate": 1.4864864864864867e-05,
      "loss": 0.3221,
      "step": 8130
    },
    {
      "epoch": 72.01,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 0.2317,
      "step": 8140
    },
    {
      "epoch": 72.01,
      "learning_rate": 1.4764764764764765e-05,
      "loss": 0.1296,
      "step": 8150
    },
    {
      "epoch": 72.01,
      "learning_rate": 1.4714714714714713e-05,
      "loss": 0.1734,
      "step": 8160
    },
    {
      "epoch": 72.01,
      "learning_rate": 1.4664664664664665e-05,
      "loss": 0.4459,
      "step": 8170
    },
    {
      "epoch": 72.01,
      "eval_accuracy": 0.7465437788018433,
      "eval_loss": 1.0925841331481934,
      "eval_runtime": 293.4498,
      "eval_samples_per_second": 0.739,
      "eval_steps_per_second": 0.187,
      "step": 8176
    },
    {
      "epoch": 73.0,
      "learning_rate": 1.4614614614614617e-05,
      "loss": 0.6718,
      "step": 8180
    },
    {
      "epoch": 73.0,
      "learning_rate": 1.4564564564564565e-05,
      "loss": 0.5033,
      "step": 8190
    },
    {
      "epoch": 73.0,
      "learning_rate": 1.4514514514514515e-05,
      "loss": 0.6681,
      "step": 8200
    },
    {
      "epoch": 73.0,
      "learning_rate": 1.4464464464464464e-05,
      "loss": 0.4871,
      "step": 8210
    },
    {
      "epoch": 73.0,
      "learning_rate": 1.4414414414414416e-05,
      "loss": 0.9771,
      "step": 8220
    },
    {
      "epoch": 73.0,
      "learning_rate": 1.4364364364364366e-05,
      "loss": 0.7966,
      "step": 8230
    },
    {
      "epoch": 73.01,
      "learning_rate": 1.4314314314314314e-05,
      "loss": 0.4869,
      "step": 8240
    },
    {
      "epoch": 73.01,
      "learning_rate": 1.4264264264264266e-05,
      "loss": 0.3706,
      "step": 8250
    },
    {
      "epoch": 73.01,
      "learning_rate": 1.4214214214214214e-05,
      "loss": 0.3956,
      "step": 8260
    },
    {
      "epoch": 73.01,
      "learning_rate": 1.4164164164164164e-05,
      "loss": 0.4966,
      "step": 8270
    },
    {
      "epoch": 73.01,
      "learning_rate": 1.4114114114114116e-05,
      "loss": 0.511,
      "step": 8280
    },
    {
      "epoch": 73.01,
      "eval_accuracy": 0.7142857142857143,
      "eval_loss": 1.1232726573944092,
      "eval_runtime": 272.043,
      "eval_samples_per_second": 0.798,
      "eval_steps_per_second": 0.202,
      "step": 8288
    },
    {
      "epoch": 74.0,
      "learning_rate": 1.4064064064064064e-05,
      "loss": 0.4828,
      "step": 8290
    },
    {
      "epoch": 74.0,
      "learning_rate": 1.4014014014014016e-05,
      "loss": 0.1156,
      "step": 8300
    },
    {
      "epoch": 74.0,
      "learning_rate": 1.3963963963963963e-05,
      "loss": 0.5592,
      "step": 8310
    },
    {
      "epoch": 74.0,
      "learning_rate": 1.3913913913913915e-05,
      "loss": 0.4964,
      "step": 8320
    },
    {
      "epoch": 74.0,
      "learning_rate": 1.3863863863863866e-05,
      "loss": 0.5616,
      "step": 8330
    },
    {
      "epoch": 74.0,
      "learning_rate": 1.3813813813813815e-05,
      "loss": 0.2516,
      "step": 8340
    },
    {
      "epoch": 74.01,
      "learning_rate": 1.3763763763763765e-05,
      "loss": 0.747,
      "step": 8350
    },
    {
      "epoch": 74.01,
      "learning_rate": 1.3713713713713713e-05,
      "loss": 0.5726,
      "step": 8360
    },
    {
      "epoch": 74.01,
      "learning_rate": 1.3663663663663665e-05,
      "loss": 0.6994,
      "step": 8370
    },
    {
      "epoch": 74.01,
      "learning_rate": 1.3613613613613615e-05,
      "loss": 0.3082,
      "step": 8380
    },
    {
      "epoch": 74.01,
      "learning_rate": 1.3563563563563563e-05,
      "loss": 0.2266,
      "step": 8390
    },
    {
      "epoch": 74.01,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 0.7933,
      "step": 8400
    },
    {
      "epoch": 74.01,
      "eval_accuracy": 0.7188940092165899,
      "eval_loss": 1.153515100479126,
      "eval_runtime": 309.6639,
      "eval_samples_per_second": 0.701,
      "eval_steps_per_second": 0.178,
      "step": 8400
    },
    {
      "epoch": 75.0,
      "learning_rate": 1.3463463463463463e-05,
      "loss": 0.4028,
      "step": 8410
    },
    {
      "epoch": 75.0,
      "learning_rate": 1.3413413413413414e-05,
      "loss": 0.423,
      "step": 8420
    },
    {
      "epoch": 75.0,
      "learning_rate": 1.3363363363363365e-05,
      "loss": 0.6182,
      "step": 8430
    },
    {
      "epoch": 75.0,
      "learning_rate": 1.3313313313313314e-05,
      "loss": 0.3103,
      "step": 8440
    },
    {
      "epoch": 75.0,
      "learning_rate": 1.3263263263263265e-05,
      "loss": 0.8707,
      "step": 8450
    },
    {
      "epoch": 75.01,
      "learning_rate": 1.3213213213213214e-05,
      "loss": 0.2099,
      "step": 8460
    },
    {
      "epoch": 75.01,
      "learning_rate": 1.3163163163163164e-05,
      "loss": 0.6191,
      "step": 8470
    },
    {
      "epoch": 75.01,
      "learning_rate": 1.3113113113113112e-05,
      "loss": 0.4439,
      "step": 8480
    },
    {
      "epoch": 75.01,
      "learning_rate": 1.3063063063063064e-05,
      "loss": 0.574,
      "step": 8490
    },
    {
      "epoch": 75.01,
      "learning_rate": 1.3013013013013014e-05,
      "loss": 0.3623,
      "step": 8500
    },
    {
      "epoch": 75.01,
      "learning_rate": 1.2962962962962962e-05,
      "loss": 0.3739,
      "step": 8510
    },
    {
      "epoch": 75.01,
      "eval_accuracy": 0.6912442396313364,
      "eval_loss": 1.3055968284606934,
      "eval_runtime": 291.9614,
      "eval_samples_per_second": 0.743,
      "eval_steps_per_second": 0.188,
      "step": 8512
    },
    {
      "epoch": 76.0,
      "learning_rate": 1.2912912912912914e-05,
      "loss": 0.2046,
      "step": 8520
    },
    {
      "epoch": 76.0,
      "learning_rate": 1.2862862862862863e-05,
      "loss": 0.6326,
      "step": 8530
    },
    {
      "epoch": 76.0,
      "learning_rate": 1.2812812812812813e-05,
      "loss": 0.5719,
      "step": 8540
    },
    {
      "epoch": 76.0,
      "learning_rate": 1.2762762762762764e-05,
      "loss": 0.5504,
      "step": 8550
    },
    {
      "epoch": 76.0,
      "learning_rate": 1.2712712712712713e-05,
      "loss": 0.4249,
      "step": 8560
    },
    {
      "epoch": 76.01,
      "learning_rate": 1.2662662662662665e-05,
      "loss": 0.7839,
      "step": 8570
    },
    {
      "epoch": 76.01,
      "learning_rate": 1.2612612612612611e-05,
      "loss": 0.3668,
      "step": 8580
    },
    {
      "epoch": 76.01,
      "learning_rate": 1.2562562562562563e-05,
      "loss": 0.5527,
      "step": 8590
    },
    {
      "epoch": 76.01,
      "learning_rate": 1.2512512512512515e-05,
      "loss": 0.708,
      "step": 8600
    },
    {
      "epoch": 76.01,
      "learning_rate": 1.2462462462462463e-05,
      "loss": 0.2421,
      "step": 8610
    },
    {
      "epoch": 76.01,
      "learning_rate": 1.2412412412412413e-05,
      "loss": 0.6976,
      "step": 8620
    },
    {
      "epoch": 76.01,
      "eval_accuracy": 0.6682027649769585,
      "eval_loss": 1.315895438194275,
      "eval_runtime": 271.4047,
      "eval_samples_per_second": 0.8,
      "eval_steps_per_second": 0.203,
      "step": 8624
    },
    {
      "epoch": 77.0,
      "learning_rate": 1.2362362362362363e-05,
      "loss": 0.4569,
      "step": 8630
    },
    {
      "epoch": 77.0,
      "learning_rate": 1.2312312312312313e-05,
      "loss": 0.7489,
      "step": 8640
    },
    {
      "epoch": 77.0,
      "learning_rate": 1.2262262262262263e-05,
      "loss": 0.3409,
      "step": 8650
    },
    {
      "epoch": 77.0,
      "learning_rate": 1.2212212212212212e-05,
      "loss": 0.6826,
      "step": 8660
    },
    {
      "epoch": 77.0,
      "learning_rate": 1.2162162162162164e-05,
      "loss": 0.2785,
      "step": 8670
    },
    {
      "epoch": 77.01,
      "learning_rate": 1.2112112112112114e-05,
      "loss": 0.4416,
      "step": 8680
    },
    {
      "epoch": 77.01,
      "learning_rate": 1.2062062062062062e-05,
      "loss": 0.7004,
      "step": 8690
    },
    {
      "epoch": 77.01,
      "learning_rate": 1.2012012012012012e-05,
      "loss": 0.4488,
      "step": 8700
    },
    {
      "epoch": 77.01,
      "learning_rate": 1.1961961961961962e-05,
      "loss": 0.3466,
      "step": 8710
    },
    {
      "epoch": 77.01,
      "learning_rate": 1.1911911911911912e-05,
      "loss": 0.5827,
      "step": 8720
    },
    {
      "epoch": 77.01,
      "learning_rate": 1.1861861861861862e-05,
      "loss": 0.5453,
      "step": 8730
    },
    {
      "epoch": 77.01,
      "eval_accuracy": 0.6359447004608295,
      "eval_loss": 1.4541015625,
      "eval_runtime": 312.7897,
      "eval_samples_per_second": 0.694,
      "eval_steps_per_second": 0.176,
      "step": 8736
    },
    {
      "epoch": 78.0,
      "learning_rate": 1.1811811811811812e-05,
      "loss": 0.4305,
      "step": 8740
    },
    {
      "epoch": 78.0,
      "learning_rate": 1.1761761761761762e-05,
      "loss": 0.6358,
      "step": 8750
    },
    {
      "epoch": 78.0,
      "learning_rate": 1.1711711711711713e-05,
      "loss": 0.1943,
      "step": 8760
    },
    {
      "epoch": 78.0,
      "learning_rate": 1.1661661661661661e-05,
      "loss": 0.3346,
      "step": 8770
    },
    {
      "epoch": 78.0,
      "learning_rate": 1.1611611611611613e-05,
      "loss": 0.2824,
      "step": 8780
    },
    {
      "epoch": 78.0,
      "learning_rate": 1.1561561561561563e-05,
      "loss": 0.5726,
      "step": 8790
    },
    {
      "epoch": 78.01,
      "learning_rate": 1.1511511511511513e-05,
      "loss": 0.4458,
      "step": 8800
    },
    {
      "epoch": 78.01,
      "learning_rate": 1.1461461461461461e-05,
      "loss": 0.408,
      "step": 8810
    },
    {
      "epoch": 78.01,
      "learning_rate": 1.1411411411411411e-05,
      "loss": 0.4169,
      "step": 8820
    },
    {
      "epoch": 78.01,
      "learning_rate": 1.1361361361361363e-05,
      "loss": 0.3489,
      "step": 8830
    },
    {
      "epoch": 78.01,
      "learning_rate": 1.1311311311311313e-05,
      "loss": 0.2915,
      "step": 8840
    },
    {
      "epoch": 78.01,
      "eval_accuracy": 0.7050691244239631,
      "eval_loss": 1.260127067565918,
      "eval_runtime": 293.4278,
      "eval_samples_per_second": 0.74,
      "eval_steps_per_second": 0.187,
      "step": 8848
    },
    {
      "epoch": 79.0,
      "learning_rate": 1.1261261261261261e-05,
      "loss": 0.4307,
      "step": 8850
    },
    {
      "epoch": 79.0,
      "learning_rate": 1.1211211211211212e-05,
      "loss": 0.2094,
      "step": 8860
    },
    {
      "epoch": 79.0,
      "learning_rate": 1.1161161161161162e-05,
      "loss": 0.6485,
      "step": 8870
    },
    {
      "epoch": 79.0,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.1876,
      "step": 8880
    },
    {
      "epoch": 79.0,
      "learning_rate": 1.1061061061061062e-05,
      "loss": 0.5738,
      "step": 8890
    },
    {
      "epoch": 79.0,
      "learning_rate": 1.1011011011011012e-05,
      "loss": 0.4072,
      "step": 8900
    },
    {
      "epoch": 79.01,
      "learning_rate": 1.0960960960960962e-05,
      "loss": 0.7778,
      "step": 8910
    },
    {
      "epoch": 79.01,
      "learning_rate": 1.0910910910910912e-05,
      "loss": 0.5295,
      "step": 8920
    },
    {
      "epoch": 79.01,
      "learning_rate": 1.086086086086086e-05,
      "loss": 0.7981,
      "step": 8930
    },
    {
      "epoch": 79.01,
      "learning_rate": 1.0810810810810812e-05,
      "loss": 0.5851,
      "step": 8940
    },
    {
      "epoch": 79.01,
      "learning_rate": 1.0760760760760762e-05,
      "loss": 0.4999,
      "step": 8950
    },
    {
      "epoch": 79.01,
      "learning_rate": 1.071071071071071e-05,
      "loss": 0.6552,
      "step": 8960
    },
    {
      "epoch": 79.01,
      "eval_accuracy": 0.6543778801843319,
      "eval_loss": 1.5337839126586914,
      "eval_runtime": 268.8948,
      "eval_samples_per_second": 0.807,
      "eval_steps_per_second": 0.205,
      "step": 8960
    },
    {
      "epoch": 80.0,
      "learning_rate": 1.066066066066066e-05,
      "loss": 0.5533,
      "step": 8970
    },
    {
      "epoch": 80.0,
      "learning_rate": 1.061061061061061e-05,
      "loss": 0.2836,
      "step": 8980
    },
    {
      "epoch": 80.0,
      "learning_rate": 1.0560560560560562e-05,
      "loss": 0.3853,
      "step": 8990
    },
    {
      "epoch": 80.0,
      "learning_rate": 1.051051051051051e-05,
      "loss": 0.6543,
      "step": 9000
    },
    {
      "epoch": 80.0,
      "learning_rate": 1.0460460460460461e-05,
      "loss": 0.608,
      "step": 9010
    },
    {
      "epoch": 80.01,
      "learning_rate": 1.0410410410410411e-05,
      "loss": 0.7773,
      "step": 9020
    },
    {
      "epoch": 80.01,
      "learning_rate": 1.0360360360360361e-05,
      "loss": 0.6478,
      "step": 9030
    },
    {
      "epoch": 80.01,
      "learning_rate": 1.031031031031031e-05,
      "loss": 0.6136,
      "step": 9040
    },
    {
      "epoch": 80.01,
      "learning_rate": 1.0260260260260261e-05,
      "loss": 0.6951,
      "step": 9050
    },
    {
      "epoch": 80.01,
      "learning_rate": 1.0210210210210211e-05,
      "loss": 0.4127,
      "step": 9060
    },
    {
      "epoch": 80.01,
      "learning_rate": 1.0160160160160161e-05,
      "loss": 0.5067,
      "step": 9070
    },
    {
      "epoch": 80.01,
      "eval_accuracy": 0.6036866359447005,
      "eval_loss": 1.6629668474197388,
      "eval_runtime": 316.3804,
      "eval_samples_per_second": 0.686,
      "eval_steps_per_second": 0.174,
      "step": 9072
    },
    {
      "epoch": 81.0,
      "learning_rate": 1.011011011011011e-05,
      "loss": 0.7643,
      "step": 9080
    },
    {
      "epoch": 81.0,
      "learning_rate": 1.006006006006006e-05,
      "loss": 0.6841,
      "step": 9090
    },
    {
      "epoch": 81.0,
      "learning_rate": 1.0010010010010011e-05,
      "loss": 0.447,
      "step": 9100
    },
    {
      "epoch": 81.0,
      "learning_rate": 9.959959959959962e-06,
      "loss": 0.2101,
      "step": 9110
    },
    {
      "epoch": 81.0,
      "learning_rate": 9.90990990990991e-06,
      "loss": 0.3463,
      "step": 9120
    },
    {
      "epoch": 81.01,
      "learning_rate": 9.85985985985986e-06,
      "loss": 0.5148,
      "step": 9130
    },
    {
      "epoch": 81.01,
      "learning_rate": 9.80980980980981e-06,
      "loss": 0.3503,
      "step": 9140
    },
    {
      "epoch": 81.01,
      "learning_rate": 9.75975975975976e-06,
      "loss": 0.3074,
      "step": 9150
    },
    {
      "epoch": 81.01,
      "learning_rate": 9.70970970970971e-06,
      "loss": 0.1355,
      "step": 9160
    },
    {
      "epoch": 81.01,
      "learning_rate": 9.65965965965966e-06,
      "loss": 0.5866,
      "step": 9170
    },
    {
      "epoch": 81.01,
      "learning_rate": 9.60960960960961e-06,
      "loss": 0.5134,
      "step": 9180
    },
    {
      "epoch": 81.01,
      "eval_accuracy": 0.6405529953917051,
      "eval_loss": 1.4740076065063477,
      "eval_runtime": 303.9606,
      "eval_samples_per_second": 0.714,
      "eval_steps_per_second": 0.181,
      "step": 9184
    },
    {
      "epoch": 82.0,
      "learning_rate": 9.55955955955956e-06,
      "loss": 0.6976,
      "step": 9190
    },
    {
      "epoch": 82.0,
      "learning_rate": 9.509509509509509e-06,
      "loss": 0.552,
      "step": 9200
    },
    {
      "epoch": 82.0,
      "learning_rate": 9.45945945945946e-06,
      "loss": 0.3769,
      "step": 9210
    },
    {
      "epoch": 82.0,
      "learning_rate": 9.40940940940941e-06,
      "loss": 0.6531,
      "step": 9220
    },
    {
      "epoch": 82.0,
      "learning_rate": 9.359359359359359e-06,
      "loss": 0.4509,
      "step": 9230
    },
    {
      "epoch": 82.01,
      "learning_rate": 9.309309309309309e-06,
      "loss": 0.4027,
      "step": 9240
    },
    {
      "epoch": 82.01,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.4775,
      "step": 9250
    },
    {
      "epoch": 82.01,
      "learning_rate": 9.209209209209211e-06,
      "loss": 0.4307,
      "step": 9260
    },
    {
      "epoch": 82.01,
      "learning_rate": 9.15915915915916e-06,
      "loss": 0.7579,
      "step": 9270
    },
    {
      "epoch": 82.01,
      "learning_rate": 9.10910910910911e-06,
      "loss": 0.324,
      "step": 9280
    },
    {
      "epoch": 82.01,
      "learning_rate": 9.05905905905906e-06,
      "loss": 0.7271,
      "step": 9290
    },
    {
      "epoch": 82.01,
      "eval_accuracy": 0.7096774193548387,
      "eval_loss": 1.217063069343567,
      "eval_runtime": 268.0359,
      "eval_samples_per_second": 0.81,
      "eval_steps_per_second": 0.205,
      "step": 9296
    },
    {
      "epoch": 83.0,
      "learning_rate": 9.00900900900901e-06,
      "loss": 0.4672,
      "step": 9300
    },
    {
      "epoch": 83.0,
      "learning_rate": 8.958958958958958e-06,
      "loss": 0.3574,
      "step": 9310
    },
    {
      "epoch": 83.0,
      "learning_rate": 8.90890890890891e-06,
      "loss": 0.5802,
      "step": 9320
    },
    {
      "epoch": 83.0,
      "learning_rate": 8.85885885885886e-06,
      "loss": 0.7018,
      "step": 9330
    },
    {
      "epoch": 83.0,
      "learning_rate": 8.80880880880881e-06,
      "loss": 0.3151,
      "step": 9340
    },
    {
      "epoch": 83.0,
      "learning_rate": 8.758758758758758e-06,
      "loss": 0.4251,
      "step": 9350
    },
    {
      "epoch": 83.01,
      "learning_rate": 8.708708708708708e-06,
      "loss": 0.5884,
      "step": 9360
    },
    {
      "epoch": 83.01,
      "learning_rate": 8.65865865865866e-06,
      "loss": 0.7406,
      "step": 9370
    },
    {
      "epoch": 83.01,
      "learning_rate": 8.60860860860861e-06,
      "loss": 0.6026,
      "step": 9380
    },
    {
      "epoch": 83.01,
      "learning_rate": 8.558558558558558e-06,
      "loss": 0.3444,
      "step": 9390
    },
    {
      "epoch": 83.01,
      "learning_rate": 8.508508508508508e-06,
      "loss": 0.719,
      "step": 9400
    },
    {
      "epoch": 83.01,
      "eval_accuracy": 0.6405529953917051,
      "eval_loss": 1.3652501106262207,
      "eval_runtime": 305.7045,
      "eval_samples_per_second": 0.71,
      "eval_steps_per_second": 0.18,
      "step": 9408
    },
    {
      "epoch": 84.0,
      "learning_rate": 8.458458458458459e-06,
      "loss": 0.6204,
      "step": 9410
    },
    {
      "epoch": 84.0,
      "learning_rate": 8.408408408408409e-06,
      "loss": 0.3308,
      "step": 9420
    },
    {
      "epoch": 84.0,
      "learning_rate": 8.358358358358359e-06,
      "loss": 0.8213,
      "step": 9430
    },
    {
      "epoch": 84.0,
      "learning_rate": 8.308308308308309e-06,
      "loss": 0.1882,
      "step": 9440
    },
    {
      "epoch": 84.0,
      "learning_rate": 8.258258258258259e-06,
      "loss": 0.4721,
      "step": 9450
    },
    {
      "epoch": 84.0,
      "learning_rate": 8.208208208208209e-06,
      "loss": 0.3433,
      "step": 9460
    },
    {
      "epoch": 84.01,
      "learning_rate": 8.158158158158157e-06,
      "loss": 0.6425,
      "step": 9470
    },
    {
      "epoch": 84.01,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.4091,
      "step": 9480
    },
    {
      "epoch": 84.01,
      "learning_rate": 8.058058058058059e-06,
      "loss": 0.4593,
      "step": 9490
    },
    {
      "epoch": 84.01,
      "learning_rate": 8.008008008008007e-06,
      "loss": 0.8545,
      "step": 9500
    },
    {
      "epoch": 84.01,
      "learning_rate": 7.957957957957958e-06,
      "loss": 0.4124,
      "step": 9510
    },
    {
      "epoch": 84.01,
      "learning_rate": 7.907907907907908e-06,
      "loss": 0.1955,
      "step": 9520
    },
    {
      "epoch": 84.01,
      "eval_accuracy": 0.6543778801843319,
      "eval_loss": 1.469608187675476,
      "eval_runtime": 293.1438,
      "eval_samples_per_second": 0.74,
      "eval_steps_per_second": 0.188,
      "step": 9520
    },
    {
      "epoch": 85.0,
      "learning_rate": 7.85785785785786e-06,
      "loss": 0.8601,
      "step": 9530
    },
    {
      "epoch": 85.0,
      "learning_rate": 7.807807807807808e-06,
      "loss": 0.2809,
      "step": 9540
    },
    {
      "epoch": 85.0,
      "learning_rate": 7.757757757757758e-06,
      "loss": 0.5543,
      "step": 9550
    },
    {
      "epoch": 85.0,
      "learning_rate": 7.707707707707708e-06,
      "loss": 0.2644,
      "step": 9560
    },
    {
      "epoch": 85.0,
      "learning_rate": 7.657657657657658e-06,
      "loss": 0.8726,
      "step": 9570
    },
    {
      "epoch": 85.01,
      "learning_rate": 7.607607607607609e-06,
      "loss": 0.4903,
      "step": 9580
    },
    {
      "epoch": 85.01,
      "learning_rate": 7.557557557557558e-06,
      "loss": 0.402,
      "step": 9590
    },
    {
      "epoch": 85.01,
      "learning_rate": 7.507507507507508e-06,
      "loss": 0.8645,
      "step": 9600
    },
    {
      "epoch": 85.01,
      "learning_rate": 7.457457457457457e-06,
      "loss": 0.4412,
      "step": 9610
    },
    {
      "epoch": 85.01,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.5641,
      "step": 9620
    },
    {
      "epoch": 85.01,
      "learning_rate": 7.357357357357357e-06,
      "loss": 0.5761,
      "step": 9630
    },
    {
      "epoch": 85.01,
      "eval_accuracy": 0.663594470046083,
      "eval_loss": 1.3333886861801147,
      "eval_runtime": 270.1924,
      "eval_samples_per_second": 0.803,
      "eval_steps_per_second": 0.204,
      "step": 9632
    },
    {
      "epoch": 86.0,
      "learning_rate": 7.3073073073073085e-06,
      "loss": 0.2616,
      "step": 9640
    },
    {
      "epoch": 86.0,
      "learning_rate": 7.257257257257258e-06,
      "loss": 0.5752,
      "step": 9650
    },
    {
      "epoch": 86.0,
      "learning_rate": 7.207207207207208e-06,
      "loss": 0.5452,
      "step": 9660
    },
    {
      "epoch": 86.0,
      "learning_rate": 7.157157157157157e-06,
      "loss": 0.5021,
      "step": 9670
    },
    {
      "epoch": 86.0,
      "learning_rate": 7.107107107107107e-06,
      "loss": 0.2643,
      "step": 9680
    },
    {
      "epoch": 86.01,
      "learning_rate": 7.057057057057058e-06,
      "loss": 0.3674,
      "step": 9690
    },
    {
      "epoch": 86.01,
      "learning_rate": 7.007007007007008e-06,
      "loss": 0.6156,
      "step": 9700
    },
    {
      "epoch": 86.01,
      "learning_rate": 6.956956956956957e-06,
      "loss": 0.4581,
      "step": 9710
    },
    {
      "epoch": 86.01,
      "learning_rate": 6.906906906906907e-06,
      "loss": 0.5333,
      "step": 9720
    },
    {
      "epoch": 86.01,
      "learning_rate": 6.8568568568568566e-06,
      "loss": 0.5513,
      "step": 9730
    },
    {
      "epoch": 86.01,
      "learning_rate": 6.8068068068068075e-06,
      "loss": 0.7094,
      "step": 9740
    },
    {
      "epoch": 86.01,
      "eval_accuracy": 0.6912442396313364,
      "eval_loss": 1.2673174142837524,
      "eval_runtime": 310.39,
      "eval_samples_per_second": 0.699,
      "eval_steps_per_second": 0.177,
      "step": 9744
    },
    {
      "epoch": 87.0,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 0.2795,
      "step": 9750
    },
    {
      "epoch": 87.0,
      "learning_rate": 6.706706706706707e-06,
      "loss": 0.6359,
      "step": 9760
    },
    {
      "epoch": 87.0,
      "learning_rate": 6.656656656656657e-06,
      "loss": 0.3073,
      "step": 9770
    },
    {
      "epoch": 87.0,
      "learning_rate": 6.606606606606607e-06,
      "loss": 0.3954,
      "step": 9780
    },
    {
      "epoch": 87.0,
      "learning_rate": 6.556556556556556e-06,
      "loss": 0.4322,
      "step": 9790
    },
    {
      "epoch": 87.01,
      "learning_rate": 6.506506506506507e-06,
      "loss": 0.3841,
      "step": 9800
    },
    {
      "epoch": 87.01,
      "learning_rate": 6.456456456456457e-06,
      "loss": 0.2992,
      "step": 9810
    },
    {
      "epoch": 87.01,
      "learning_rate": 6.406406406406406e-06,
      "loss": 0.6195,
      "step": 9820
    },
    {
      "epoch": 87.01,
      "learning_rate": 6.356356356356356e-06,
      "loss": 0.6608,
      "step": 9830
    },
    {
      "epoch": 87.01,
      "learning_rate": 6.306306306306306e-06,
      "loss": 0.5034,
      "step": 9840
    },
    {
      "epoch": 87.01,
      "learning_rate": 6.256256256256257e-06,
      "loss": 0.5186,
      "step": 9850
    },
    {
      "epoch": 87.01,
      "eval_accuracy": 0.6866359447004609,
      "eval_loss": 1.3147472143173218,
      "eval_runtime": 301.6939,
      "eval_samples_per_second": 0.719,
      "eval_steps_per_second": 0.182,
      "step": 9856
    },
    {
      "epoch": 88.0,
      "learning_rate": 6.206206206206207e-06,
      "loss": 0.2595,
      "step": 9860
    },
    {
      "epoch": 88.0,
      "learning_rate": 6.156156156156157e-06,
      "loss": 0.6629,
      "step": 9870
    },
    {
      "epoch": 88.0,
      "learning_rate": 6.106106106106106e-06,
      "loss": 0.6307,
      "step": 9880
    },
    {
      "epoch": 88.0,
      "learning_rate": 6.056056056056057e-06,
      "loss": 0.596,
      "step": 9890
    },
    {
      "epoch": 88.0,
      "learning_rate": 6.006006006006006e-06,
      "loss": 0.465,
      "step": 9900
    },
    {
      "epoch": 88.0,
      "learning_rate": 5.955955955955956e-06,
      "loss": 0.4708,
      "step": 9910
    },
    {
      "epoch": 88.01,
      "learning_rate": 5.905905905905906e-06,
      "loss": 0.4007,
      "step": 9920
    },
    {
      "epoch": 88.01,
      "learning_rate": 5.855855855855856e-06,
      "loss": 0.6553,
      "step": 9930
    },
    {
      "epoch": 88.01,
      "learning_rate": 5.805805805805806e-06,
      "loss": 0.5481,
      "step": 9940
    },
    {
      "epoch": 88.01,
      "learning_rate": 5.755755755755756e-06,
      "loss": 0.1136,
      "step": 9950
    },
    {
      "epoch": 88.01,
      "learning_rate": 5.705705705705706e-06,
      "loss": 0.6876,
      "step": 9960
    },
    {
      "epoch": 88.01,
      "eval_accuracy": 0.7050691244239631,
      "eval_loss": 1.2622275352478027,
      "eval_runtime": 266.213,
      "eval_samples_per_second": 0.815,
      "eval_steps_per_second": 0.207,
      "step": 9968
    },
    {
      "epoch": 89.0,
      "learning_rate": 5.6556556556556565e-06,
      "loss": 0.9633,
      "step": 9970
    },
    {
      "epoch": 89.0,
      "learning_rate": 5.605605605605606e-06,
      "loss": 0.3353,
      "step": 9980
    },
    {
      "epoch": 89.0,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.3828,
      "step": 9990
    },
    {
      "epoch": 89.0,
      "learning_rate": 5.505505505505506e-06,
      "loss": 0.5327,
      "step": 10000
    },
    {
      "epoch": 89.0,
      "learning_rate": 5.455455455455456e-06,
      "loss": 0.4945,
      "step": 10010
    },
    {
      "epoch": 89.0,
      "learning_rate": 5.405405405405406e-06,
      "loss": 0.4838,
      "step": 10020
    },
    {
      "epoch": 89.01,
      "learning_rate": 5.355355355355355e-06,
      "loss": 0.3726,
      "step": 10030
    },
    {
      "epoch": 89.01,
      "learning_rate": 5.305305305305305e-06,
      "loss": 0.4962,
      "step": 10040
    },
    {
      "epoch": 89.01,
      "learning_rate": 5.255255255255255e-06,
      "loss": 0.33,
      "step": 10050
    },
    {
      "epoch": 89.01,
      "learning_rate": 5.2052052052052055e-06,
      "loss": 0.2023,
      "step": 10060
    },
    {
      "epoch": 89.01,
      "learning_rate": 5.155155155155155e-06,
      "loss": 0.1939,
      "step": 10070
    },
    {
      "epoch": 89.01,
      "learning_rate": 5.105105105105106e-06,
      "loss": 0.4912,
      "step": 10080
    },
    {
      "epoch": 89.01,
      "eval_accuracy": 0.7188940092165899,
      "eval_loss": 1.3053789138793945,
      "eval_runtime": 300.2496,
      "eval_samples_per_second": 0.723,
      "eval_steps_per_second": 0.183,
      "step": 10080
    },
    {
      "epoch": 90.0,
      "learning_rate": 5.055055055055055e-06,
      "loss": 0.6622,
      "step": 10090
    },
    {
      "epoch": 90.0,
      "learning_rate": 5.005005005005006e-06,
      "loss": 0.5072,
      "step": 10100
    },
    {
      "epoch": 90.0,
      "learning_rate": 4.954954954954955e-06,
      "loss": 0.4493,
      "step": 10110
    },
    {
      "epoch": 90.0,
      "learning_rate": 4.904904904904905e-06,
      "loss": 0.2136,
      "step": 10120
    },
    {
      "epoch": 90.0,
      "learning_rate": 4.854854854854855e-06,
      "loss": 0.5732,
      "step": 10130
    },
    {
      "epoch": 90.01,
      "learning_rate": 4.804804804804805e-06,
      "loss": 0.3335,
      "step": 10140
    },
    {
      "epoch": 90.01,
      "learning_rate": 4.754754754754754e-06,
      "loss": 0.321,
      "step": 10150
    },
    {
      "epoch": 90.01,
      "learning_rate": 4.704704704704705e-06,
      "loss": 0.2949,
      "step": 10160
    },
    {
      "epoch": 90.01,
      "learning_rate": 4.6546546546546545e-06,
      "loss": 0.2822,
      "step": 10170
    },
    {
      "epoch": 90.01,
      "learning_rate": 4.6046046046046055e-06,
      "loss": 0.3274,
      "step": 10180
    },
    {
      "epoch": 90.01,
      "learning_rate": 4.554554554554555e-06,
      "loss": 0.194,
      "step": 10190
    },
    {
      "epoch": 90.01,
      "eval_accuracy": 0.695852534562212,
      "eval_loss": 1.324397087097168,
      "eval_runtime": 305.0227,
      "eval_samples_per_second": 0.711,
      "eval_steps_per_second": 0.18,
      "step": 10192
    },
    {
      "epoch": 91.0,
      "learning_rate": 4.504504504504505e-06,
      "loss": 0.574,
      "step": 10200
    },
    {
      "epoch": 91.0,
      "learning_rate": 4.454454454454455e-06,
      "loss": 0.4654,
      "step": 10210
    },
    {
      "epoch": 91.0,
      "learning_rate": 4.404404404404405e-06,
      "loss": 0.3711,
      "step": 10220
    },
    {
      "epoch": 91.0,
      "learning_rate": 4.354354354354354e-06,
      "loss": 0.7136,
      "step": 10230
    },
    {
      "epoch": 91.0,
      "learning_rate": 4.304304304304305e-06,
      "loss": 0.5138,
      "step": 10240
    },
    {
      "epoch": 91.01,
      "learning_rate": 4.254254254254254e-06,
      "loss": 0.4856,
      "step": 10250
    },
    {
      "epoch": 91.01,
      "learning_rate": 4.204204204204204e-06,
      "loss": 0.3505,
      "step": 10260
    },
    {
      "epoch": 91.01,
      "learning_rate": 4.154154154154154e-06,
      "loss": 0.6113,
      "step": 10270
    },
    {
      "epoch": 91.01,
      "learning_rate": 4.1041041041041045e-06,
      "loss": 0.4895,
      "step": 10280
    },
    {
      "epoch": 91.01,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 0.3489,
      "step": 10290
    },
    {
      "epoch": 91.01,
      "learning_rate": 4.004004004004004e-06,
      "loss": 0.6916,
      "step": 10300
    },
    {
      "epoch": 91.01,
      "eval_accuracy": 0.7327188940092166,
      "eval_loss": 1.180008888244629,
      "eval_runtime": 291.0996,
      "eval_samples_per_second": 0.745,
      "eval_steps_per_second": 0.189,
      "step": 10304
    },
    {
      "epoch": 92.0,
      "learning_rate": 3.953953953953954e-06,
      "loss": 0.4355,
      "step": 10310
    },
    {
      "epoch": 92.0,
      "learning_rate": 3.903903903903904e-06,
      "loss": 0.3321,
      "step": 10320
    },
    {
      "epoch": 92.0,
      "learning_rate": 3.853853853853854e-06,
      "loss": 0.479,
      "step": 10330
    },
    {
      "epoch": 92.0,
      "learning_rate": 3.8038038038038044e-06,
      "loss": 0.5512,
      "step": 10340
    },
    {
      "epoch": 92.0,
      "learning_rate": 3.753753753753754e-06,
      "loss": 0.6373,
      "step": 10350
    },
    {
      "epoch": 92.01,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.5632,
      "step": 10360
    },
    {
      "epoch": 92.01,
      "learning_rate": 3.6536536536536542e-06,
      "loss": 0.4257,
      "step": 10370
    },
    {
      "epoch": 92.01,
      "learning_rate": 3.603603603603604e-06,
      "loss": 0.3636,
      "step": 10380
    },
    {
      "epoch": 92.01,
      "learning_rate": 3.5535535535535535e-06,
      "loss": 0.4397,
      "step": 10390
    },
    {
      "epoch": 92.01,
      "learning_rate": 3.503503503503504e-06,
      "loss": 0.2298,
      "step": 10400
    },
    {
      "epoch": 92.01,
      "learning_rate": 3.4534534534534537e-06,
      "loss": 0.5735,
      "step": 10410
    },
    {
      "epoch": 92.01,
      "eval_accuracy": 0.7419354838709677,
      "eval_loss": 1.1056485176086426,
      "eval_runtime": 270.7239,
      "eval_samples_per_second": 0.802,
      "eval_steps_per_second": 0.203,
      "step": 10416
    },
    {
      "epoch": 93.0,
      "learning_rate": 3.4034034034034037e-06,
      "loss": 0.3679,
      "step": 10420
    },
    {
      "epoch": 93.0,
      "learning_rate": 3.3533533533533534e-06,
      "loss": 0.393,
      "step": 10430
    },
    {
      "epoch": 93.0,
      "learning_rate": 3.3033033033033035e-06,
      "loss": 0.7959,
      "step": 10440
    },
    {
      "epoch": 93.0,
      "learning_rate": 3.2532532532532535e-06,
      "loss": 0.9955,
      "step": 10450
    },
    {
      "epoch": 93.0,
      "learning_rate": 3.203203203203203e-06,
      "loss": 0.1878,
      "step": 10460
    },
    {
      "epoch": 93.0,
      "learning_rate": 3.153153153153153e-06,
      "loss": 0.5854,
      "step": 10470
    },
    {
      "epoch": 93.01,
      "learning_rate": 3.1031031031031033e-06,
      "loss": 0.65,
      "step": 10480
    },
    {
      "epoch": 93.01,
      "learning_rate": 3.053053053053053e-06,
      "loss": 0.4863,
      "step": 10490
    },
    {
      "epoch": 93.01,
      "learning_rate": 3.003003003003003e-06,
      "loss": 0.5301,
      "step": 10500
    },
    {
      "epoch": 93.01,
      "learning_rate": 2.952952952952953e-06,
      "loss": 0.7256,
      "step": 10510
    },
    {
      "epoch": 93.01,
      "learning_rate": 2.902902902902903e-06,
      "loss": 0.2122,
      "step": 10520
    },
    {
      "epoch": 93.01,
      "eval_accuracy": 0.728110599078341,
      "eval_loss": 1.1069555282592773,
      "eval_runtime": 303.0272,
      "eval_samples_per_second": 0.716,
      "eval_steps_per_second": 0.182,
      "step": 10528
    },
    {
      "epoch": 94.0,
      "learning_rate": 2.852852852852853e-06,
      "loss": 0.3414,
      "step": 10530
    },
    {
      "epoch": 94.0,
      "learning_rate": 2.802802802802803e-06,
      "loss": 0.1727,
      "step": 10540
    },
    {
      "epoch": 94.0,
      "learning_rate": 2.752752752752753e-06,
      "loss": 0.4684,
      "step": 10550
    },
    {
      "epoch": 94.0,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.2606,
      "step": 10560
    },
    {
      "epoch": 94.0,
      "learning_rate": 2.6526526526526527e-06,
      "loss": 0.6592,
      "step": 10570
    },
    {
      "epoch": 94.0,
      "learning_rate": 2.6026026026026027e-06,
      "loss": 0.2962,
      "step": 10580
    },
    {
      "epoch": 94.01,
      "learning_rate": 2.552552552552553e-06,
      "loss": 0.4132,
      "step": 10590
    },
    {
      "epoch": 94.01,
      "learning_rate": 2.502502502502503e-06,
      "loss": 0.3587,
      "step": 10600
    },
    {
      "epoch": 94.01,
      "learning_rate": 2.4524524524524525e-06,
      "loss": 0.4078,
      "step": 10610
    },
    {
      "epoch": 94.01,
      "learning_rate": 2.4024024024024026e-06,
      "loss": 0.2856,
      "step": 10620
    },
    {
      "epoch": 94.01,
      "learning_rate": 2.3523523523523527e-06,
      "loss": 0.4584,
      "step": 10630
    },
    {
      "epoch": 94.01,
      "learning_rate": 2.3023023023023027e-06,
      "loss": 0.1434,
      "step": 10640
    },
    {
      "epoch": 94.01,
      "eval_accuracy": 0.7096774193548387,
      "eval_loss": 1.177558422088623,
      "eval_runtime": 260.0267,
      "eval_samples_per_second": 0.835,
      "eval_steps_per_second": 0.212,
      "step": 10640
    },
    {
      "epoch": 95.0,
      "learning_rate": 2.2522522522522524e-06,
      "loss": 0.278,
      "step": 10650
    },
    {
      "epoch": 95.0,
      "learning_rate": 2.2022022022022024e-06,
      "loss": 0.3673,
      "step": 10660
    },
    {
      "epoch": 95.0,
      "learning_rate": 2.1521521521521525e-06,
      "loss": 0.6963,
      "step": 10670
    },
    {
      "epoch": 95.0,
      "learning_rate": 2.102102102102102e-06,
      "loss": 0.679,
      "step": 10680
    },
    {
      "epoch": 95.0,
      "learning_rate": 2.0520520520520522e-06,
      "loss": 0.4865,
      "step": 10690
    },
    {
      "epoch": 95.01,
      "learning_rate": 2.002002002002002e-06,
      "loss": 0.2773,
      "step": 10700
    },
    {
      "epoch": 95.01,
      "learning_rate": 1.951951951951952e-06,
      "loss": 0.4471,
      "step": 10710
    },
    {
      "epoch": 95.01,
      "learning_rate": 1.9019019019019022e-06,
      "loss": 0.4114,
      "step": 10720
    },
    {
      "epoch": 95.01,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.1724,
      "step": 10730
    },
    {
      "epoch": 95.01,
      "learning_rate": 1.801801801801802e-06,
      "loss": 0.3154,
      "step": 10740
    },
    {
      "epoch": 95.01,
      "learning_rate": 1.751751751751752e-06,
      "loss": 0.4681,
      "step": 10750
    },
    {
      "epoch": 95.01,
      "eval_accuracy": 0.7327188940092166,
      "eval_loss": 1.1504991054534912,
      "eval_runtime": 284.182,
      "eval_samples_per_second": 0.764,
      "eval_steps_per_second": 0.194,
      "step": 10752
    },
    {
      "epoch": 96.0,
      "learning_rate": 1.7017017017017019e-06,
      "loss": 0.6933,
      "step": 10760
    },
    {
      "epoch": 96.0,
      "learning_rate": 1.6516516516516517e-06,
      "loss": 0.2963,
      "step": 10770
    },
    {
      "epoch": 96.0,
      "learning_rate": 1.6016016016016016e-06,
      "loss": 0.6026,
      "step": 10780
    },
    {
      "epoch": 96.0,
      "learning_rate": 1.5515515515515517e-06,
      "loss": 0.3648,
      "step": 10790
    },
    {
      "epoch": 96.0,
      "learning_rate": 1.5015015015015015e-06,
      "loss": 0.4307,
      "step": 10800
    },
    {
      "epoch": 96.01,
      "learning_rate": 1.4514514514514516e-06,
      "loss": 0.523,
      "step": 10810
    },
    {
      "epoch": 96.01,
      "learning_rate": 1.4014014014014014e-06,
      "loss": 0.2837,
      "step": 10820
    },
    {
      "epoch": 96.01,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 0.3047,
      "step": 10830
    },
    {
      "epoch": 96.01,
      "learning_rate": 1.3013013013013014e-06,
      "loss": 0.2717,
      "step": 10840
    },
    {
      "epoch": 96.01,
      "learning_rate": 1.2512512512512514e-06,
      "loss": 0.406,
      "step": 10850
    },
    {
      "epoch": 96.01,
      "learning_rate": 1.2012012012012013e-06,
      "loss": 0.2856,
      "step": 10860
    },
    {
      "epoch": 96.01,
      "eval_accuracy": 0.7235023041474654,
      "eval_loss": 1.120323896408081,
      "eval_runtime": 301.8966,
      "eval_samples_per_second": 0.719,
      "eval_steps_per_second": 0.182,
      "step": 10864
    },
    {
      "epoch": 97.0,
      "learning_rate": 1.1511511511511514e-06,
      "loss": 0.6381,
      "step": 10870
    },
    {
      "epoch": 97.0,
      "learning_rate": 1.1011011011011012e-06,
      "loss": 0.48,
      "step": 10880
    },
    {
      "epoch": 97.0,
      "learning_rate": 1.051051051051051e-06,
      "loss": 0.3974,
      "step": 10890
    },
    {
      "epoch": 97.0,
      "learning_rate": 1.001001001001001e-06,
      "loss": 0.6672,
      "step": 10900
    },
    {
      "epoch": 97.0,
      "learning_rate": 9.509509509509511e-07,
      "loss": 0.4186,
      "step": 10910
    },
    {
      "epoch": 97.01,
      "learning_rate": 9.00900900900901e-07,
      "loss": 0.6856,
      "step": 10920
    },
    {
      "epoch": 97.01,
      "learning_rate": 8.508508508508509e-07,
      "loss": 0.6567,
      "step": 10930
    },
    {
      "epoch": 97.01,
      "learning_rate": 8.008008008008008e-07,
      "loss": 0.3078,
      "step": 10940
    },
    {
      "epoch": 97.01,
      "learning_rate": 7.507507507507508e-07,
      "loss": 0.5726,
      "step": 10950
    },
    {
      "epoch": 97.01,
      "learning_rate": 7.007007007007007e-07,
      "loss": 0.3046,
      "step": 10960
    },
    {
      "epoch": 97.01,
      "learning_rate": 6.506506506506507e-07,
      "loss": 0.6509,
      "step": 10970
    },
    {
      "epoch": 97.01,
      "eval_accuracy": 0.7188940092165899,
      "eval_loss": 1.1502245664596558,
      "eval_runtime": 262.4802,
      "eval_samples_per_second": 0.827,
      "eval_steps_per_second": 0.21,
      "step": 10976
    },
    {
      "epoch": 98.0,
      "learning_rate": 6.006006006006006e-07,
      "loss": 0.6324,
      "step": 10980
    },
    {
      "epoch": 98.0,
      "learning_rate": 5.505505505505506e-07,
      "loss": 0.5201,
      "step": 10990
    },
    {
      "epoch": 98.0,
      "learning_rate": 5.005005005005005e-07,
      "loss": 0.1996,
      "step": 11000
    },
    {
      "epoch": 98.0,
      "learning_rate": 4.504504504504505e-07,
      "loss": 0.3182,
      "step": 11010
    },
    {
      "epoch": 98.0,
      "learning_rate": 4.004004004004004e-07,
      "loss": 0.3381,
      "step": 11020
    },
    {
      "epoch": 98.0,
      "learning_rate": 3.5035035035035036e-07,
      "loss": 0.1659,
      "step": 11030
    },
    {
      "epoch": 98.01,
      "learning_rate": 3.003003003003003e-07,
      "loss": 0.2857,
      "step": 11040
    },
    {
      "epoch": 98.01,
      "learning_rate": 2.5025025025025023e-07,
      "loss": 0.3733,
      "step": 11050
    },
    {
      "epoch": 98.01,
      "learning_rate": 2.002002002002002e-07,
      "loss": 0.2705,
      "step": 11060
    },
    {
      "epoch": 98.01,
      "learning_rate": 1.5015015015015016e-07,
      "loss": 0.8079,
      "step": 11070
    },
    {
      "epoch": 98.01,
      "learning_rate": 1.001001001001001e-07,
      "loss": 0.1881,
      "step": 11080
    },
    {
      "epoch": 98.01,
      "eval_accuracy": 0.7188940092165899,
      "eval_loss": 1.1473973989486694,
      "eval_runtime": 305.6996,
      "eval_samples_per_second": 0.71,
      "eval_steps_per_second": 0.18,
      "step": 11088
    },
    {
      "epoch": 99.0,
      "learning_rate": 5.005005005005005e-08,
      "loss": 0.5458,
      "step": 11090
    },
    {
      "epoch": 99.0,
      "learning_rate": 0.0,
      "loss": 0.5577,
      "step": 11100
    },
    {
      "epoch": 99.0,
      "eval_accuracy": 0.7188940092165899,
      "eval_loss": 1.1473000049591064,
      "eval_runtime": 266.2449,
      "eval_samples_per_second": 0.815,
      "eval_steps_per_second": 0.207,
      "step": 11100
    },
    {
      "epoch": 99.0,
      "step": 11100,
      "total_flos": 5.520338427328414e+19,
      "train_loss": 0.7127831653753917,
      "train_runtime": 101770.1386,
      "train_samples_per_second": 0.436,
      "train_steps_per_second": 0.109
    },
    {
      "epoch": 99.0,
      "eval_accuracy": 0.6712962962962963,
      "eval_loss": 0.7699549794197083,
      "eval_runtime": 274.3143,
      "eval_samples_per_second": 0.787,
      "eval_steps_per_second": 0.197,
      "step": 11100
    },
    {
      "epoch": 99.0,
      "eval_accuracy": 0.6712962962962963,
      "eval_loss": 0.7699549794197083,
      "eval_runtime": 277.4968,
      "eval_samples_per_second": 0.778,
      "eval_steps_per_second": 0.195,
      "step": 11100
    }
  ],
  "logging_steps": 10,
  "max_steps": 11100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "total_flos": 5.520338427328414e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}