{
  "best_metric": 0.08489208633093526,
  "best_model_checkpoint": "resnet-50-finetuned-pokemon-finetuned-pokemon/checkpoint-2486",
  "epoch": 99.34640522875817,
  "eval_steps": 500,
  "global_step": 3800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.26,
      "learning_rate": 1.3157894736842106e-06,
      "loss": 1.1424,
      "step": 10
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.631578947368421e-06,
      "loss": 1.1313,
      "step": 20
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.9473684210526315e-06,
      "loss": 1.1894,
      "step": 30
    },
    {
      "epoch": 0.99,
      "eval_accuracy": 0.013669064748201438,
      "eval_loss": 9.211542129516602,
      "eval_runtime": 5.8886,
      "eval_samples_per_second": 236.048,
      "eval_steps_per_second": 7.472,
      "step": 38
    },
    {
      "epoch": 1.05,
      "learning_rate": 5.263157894736842e-06,
      "loss": 1.1934,
      "step": 40
    },
    {
      "epoch": 1.31,
      "learning_rate": 6.578947368421053e-06,
      "loss": 1.1333,
      "step": 50
    },
    {
      "epoch": 1.57,
      "learning_rate": 7.894736842105263e-06,
      "loss": 1.0844,
      "step": 60
    },
    {
      "epoch": 1.83,
      "learning_rate": 9.210526315789474e-06,
      "loss": 1.1389,
      "step": 70
    },
    {
      "epoch": 1.99,
      "eval_accuracy": 0.012949640287769784,
      "eval_loss": 9.252067565917969,
      "eval_runtime": 5.8088,
      "eval_samples_per_second": 239.291,
      "eval_steps_per_second": 7.575,
      "step": 76
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.0526315789473684e-05,
      "loss": 1.1388,
      "step": 80
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.1842105263157895e-05,
      "loss": 1.1188,
      "step": 90
    },
    {
      "epoch": 2.61,
      "learning_rate": 1.3157894736842106e-05,
      "loss": 1.1029,
      "step": 100
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.4473684210526317e-05,
      "loss": 1.0432,
      "step": 110
    },
    {
      "epoch": 2.98,
      "eval_accuracy": 0.014388489208633094,
      "eval_loss": 9.476502418518066,
      "eval_runtime": 5.8873,
      "eval_samples_per_second": 236.102,
      "eval_steps_per_second": 7.474,
      "step": 114
    },
    {
      "epoch": 3.14,
      "learning_rate": 1.5789473684210526e-05,
      "loss": 1.124,
      "step": 120
    },
    {
      "epoch": 3.4,
      "learning_rate": 1.7105263157894737e-05,
      "loss": 1.062,
      "step": 130
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.8421052631578947e-05,
      "loss": 1.1051,
      "step": 140
    },
    {
      "epoch": 3.92,
      "learning_rate": 1.9736842105263158e-05,
      "loss": 1.0625,
      "step": 150
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.013669064748201438,
      "eval_loss": 9.7667818069458,
      "eval_runtime": 5.875,
      "eval_samples_per_second": 236.595,
      "eval_steps_per_second": 7.489,
      "step": 153
    },
    {
      "epoch": 4.18,
      "learning_rate": 2.105263157894737e-05,
      "loss": 1.1224,
      "step": 160
    },
    {
      "epoch": 4.44,
      "learning_rate": 2.236842105263158e-05,
      "loss": 1.0557,
      "step": 170
    },
    {
      "epoch": 4.71,
      "learning_rate": 2.368421052631579e-05,
      "loss": 1.0433,
      "step": 180
    },
    {
      "epoch": 4.97,
      "learning_rate": 2.5e-05,
      "loss": 1.0805,
      "step": 190
    },
    {
      "epoch": 4.99,
      "eval_accuracy": 0.013669064748201438,
      "eval_loss": 10.252641677856445,
      "eval_runtime": 5.9184,
      "eval_samples_per_second": 234.862,
      "eval_steps_per_second": 7.434,
      "step": 191
    },
    {
      "epoch": 5.23,
      "learning_rate": 2.6315789473684212e-05,
      "loss": 1.0663,
      "step": 200
    },
    {
      "epoch": 5.49,
      "learning_rate": 2.7631578947368426e-05,
      "loss": 1.0406,
      "step": 210
    },
    {
      "epoch": 5.75,
      "learning_rate": 2.8947368421052634e-05,
      "loss": 1.0353,
      "step": 220
    },
    {
      "epoch": 5.99,
      "eval_accuracy": 0.012949640287769784,
      "eval_loss": 10.323753356933594,
      "eval_runtime": 5.9005,
      "eval_samples_per_second": 235.572,
      "eval_steps_per_second": 7.457,
      "step": 229
    },
    {
      "epoch": 6.01,
      "learning_rate": 3.0263157894736844e-05,
      "loss": 1.0401,
      "step": 230
    },
    {
      "epoch": 6.27,
      "learning_rate": 3.157894736842105e-05,
      "loss": 1.0144,
      "step": 240
    },
    {
      "epoch": 6.54,
      "learning_rate": 3.289473684210527e-05,
      "loss": 1.0139,
      "step": 250
    },
    {
      "epoch": 6.8,
      "learning_rate": 3.421052631578947e-05,
      "loss": 0.9747,
      "step": 260
    },
    {
      "epoch": 6.98,
      "eval_accuracy": 0.016546762589928057,
      "eval_loss": 10.577895164489746,
      "eval_runtime": 6.0278,
      "eval_samples_per_second": 230.598,
      "eval_steps_per_second": 7.299,
      "step": 267
    },
    {
      "epoch": 7.06,
      "learning_rate": 3.5526315789473684e-05,
      "loss": 1.0199,
      "step": 270
    },
    {
      "epoch": 7.32,
      "learning_rate": 3.6842105263157895e-05,
      "loss": 0.9931,
      "step": 280
    },
    {
      "epoch": 7.58,
      "learning_rate": 3.815789473684211e-05,
      "loss": 0.9694,
      "step": 290
    },
    {
      "epoch": 7.84,
      "learning_rate": 3.9473684210526316e-05,
      "loss": 0.9708,
      "step": 300
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.017985611510791366,
      "eval_loss": 10.745814323425293,
      "eval_runtime": 6.0966,
      "eval_samples_per_second": 227.997,
      "eval_steps_per_second": 7.217,
      "step": 306
    },
    {
      "epoch": 8.1,
      "learning_rate": 4.078947368421053e-05,
      "loss": 0.9497,
      "step": 310
    },
    {
      "epoch": 8.37,
      "learning_rate": 4.210526315789474e-05,
      "loss": 0.8812,
      "step": 320
    },
    {
      "epoch": 8.63,
      "learning_rate": 4.342105263157895e-05,
      "loss": 0.8991,
      "step": 330
    },
    {
      "epoch": 8.89,
      "learning_rate": 4.473684210526316e-05,
      "loss": 0.8886,
      "step": 340
    },
    {
      "epoch": 8.99,
      "eval_accuracy": 0.019424460431654675,
      "eval_loss": 11.007186889648438,
      "eval_runtime": 6.1583,
      "eval_samples_per_second": 225.713,
      "eval_steps_per_second": 7.145,
      "step": 344
    },
    {
      "epoch": 9.15,
      "learning_rate": 4.605263157894737e-05,
      "loss": 0.8914,
      "step": 350
    },
    {
      "epoch": 9.41,
      "learning_rate": 4.736842105263158e-05,
      "loss": 0.8727,
      "step": 360
    },
    {
      "epoch": 9.67,
      "learning_rate": 4.868421052631579e-05,
      "loss": 0.8327,
      "step": 370
    },
    {
      "epoch": 9.93,
      "learning_rate": 5e-05,
      "loss": 0.8408,
      "step": 380
    },
    {
      "epoch": 9.99,
      "eval_accuracy": 0.022302158273381296,
      "eval_loss": 11.317117691040039,
      "eval_runtime": 6.1694,
      "eval_samples_per_second": 225.305,
      "eval_steps_per_second": 7.132,
      "step": 382
    },
    {
      "epoch": 10.2,
      "learning_rate": 4.985380116959065e-05,
      "loss": 0.8767,
      "step": 390
    },
    {
      "epoch": 10.46,
      "learning_rate": 4.970760233918128e-05,
      "loss": 0.801,
      "step": 400
    },
    {
      "epoch": 10.72,
      "learning_rate": 4.956140350877193e-05,
      "loss": 0.804,
      "step": 410
    },
    {
      "epoch": 10.98,
      "learning_rate": 4.941520467836258e-05,
      "loss": 0.802,
      "step": 420
    },
    {
      "epoch": 10.98,
      "eval_accuracy": 0.02446043165467626,
      "eval_loss": 11.554548263549805,
      "eval_runtime": 6.1954,
      "eval_samples_per_second": 224.36,
      "eval_steps_per_second": 7.102,
      "step": 420
    },
    {
      "epoch": 11.24,
      "learning_rate": 4.926900584795322e-05,
      "loss": 0.8308,
      "step": 430
    },
    {
      "epoch": 11.5,
      "learning_rate": 4.912280701754386e-05,
      "loss": 0.7814,
      "step": 440
    },
    {
      "epoch": 11.76,
      "learning_rate": 4.8976608187134504e-05,
      "loss": 0.7903,
      "step": 450
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.02877697841726619,
      "eval_loss": 11.77219295501709,
      "eval_runtime": 6.2667,
      "eval_samples_per_second": 221.808,
      "eval_steps_per_second": 7.021,
      "step": 459
    },
    {
      "epoch": 12.03,
      "learning_rate": 4.883040935672515e-05,
      "loss": 0.8129,
      "step": 460
    },
    {
      "epoch": 12.29,
      "learning_rate": 4.868421052631579e-05,
      "loss": 0.7376,
      "step": 470
    },
    {
      "epoch": 12.55,
      "learning_rate": 4.853801169590643e-05,
      "loss": 0.7731,
      "step": 480
    },
    {
      "epoch": 12.81,
      "learning_rate": 4.839181286549708e-05,
      "loss": 0.7553,
      "step": 490
    },
    {
      "epoch": 12.99,
      "eval_accuracy": 0.03525179856115108,
      "eval_loss": 11.983402252197266,
      "eval_runtime": 6.562,
      "eval_samples_per_second": 211.826,
      "eval_steps_per_second": 6.705,
      "step": 497
    },
    {
      "epoch": 13.07,
      "learning_rate": 4.824561403508772e-05,
      "loss": 0.744,
      "step": 500
    },
    {
      "epoch": 13.33,
      "learning_rate": 4.8099415204678366e-05,
      "loss": 0.7221,
      "step": 510
    },
    {
      "epoch": 13.59,
      "learning_rate": 4.7953216374269006e-05,
      "loss": 0.728,
      "step": 520
    },
    {
      "epoch": 13.86,
      "learning_rate": 4.780701754385965e-05,
      "loss": 0.7413,
      "step": 530
    },
    {
      "epoch": 13.99,
      "eval_accuracy": 0.04460431654676259,
      "eval_loss": 11.981474876403809,
      "eval_runtime": 6.1295,
      "eval_samples_per_second": 226.771,
      "eval_steps_per_second": 7.178,
      "step": 535
    },
    {
      "epoch": 14.12,
      "learning_rate": 4.7660818713450294e-05,
      "loss": 0.6931,
      "step": 540
    },
    {
      "epoch": 14.38,
      "learning_rate": 4.751461988304094e-05,
      "loss": 0.6729,
      "step": 550
    },
    {
      "epoch": 14.64,
      "learning_rate": 4.736842105263158e-05,
      "loss": 0.6931,
      "step": 560
    },
    {
      "epoch": 14.9,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.6272,
      "step": 570
    },
    {
      "epoch": 14.98,
      "eval_accuracy": 0.04964028776978417,
      "eval_loss": 12.08712387084961,
      "eval_runtime": 6.1408,
      "eval_samples_per_second": 226.354,
      "eval_steps_per_second": 7.165,
      "step": 573
    },
    {
      "epoch": 15.16,
      "learning_rate": 4.707602339181287e-05,
      "loss": 0.6859,
      "step": 580
    },
    {
      "epoch": 15.42,
      "learning_rate": 4.6929824561403515e-05,
      "loss": 0.6595,
      "step": 590
    },
    {
      "epoch": 15.69,
      "learning_rate": 4.678362573099415e-05,
      "loss": 0.6183,
      "step": 600
    },
    {
      "epoch": 15.95,
      "learning_rate": 4.6637426900584796e-05,
      "loss": 0.6944,
      "step": 610
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.058992805755395686,
      "eval_loss": 12.371334075927734,
      "eval_runtime": 6.0154,
      "eval_samples_per_second": 231.073,
      "eval_steps_per_second": 7.315,
      "step": 612
    },
    {
      "epoch": 16.21,
      "learning_rate": 4.649122807017544e-05,
      "loss": 0.6187,
      "step": 620
    },
    {
      "epoch": 16.47,
      "learning_rate": 4.634502923976608e-05,
      "loss": 0.6033,
      "step": 630
    },
    {
      "epoch": 16.73,
      "learning_rate": 4.619883040935672e-05,
      "loss": 0.6347,
      "step": 640
    },
    {
      "epoch": 16.99,
      "learning_rate": 4.605263157894737e-05,
      "loss": 0.6322,
      "step": 650
    },
    {
      "epoch": 16.99,
      "eval_accuracy": 0.05539568345323741,
      "eval_loss": 12.682552337646484,
      "eval_runtime": 5.924,
      "eval_samples_per_second": 234.639,
      "eval_steps_per_second": 7.427,
      "step": 650
    },
    {
      "epoch": 17.25,
      "learning_rate": 4.590643274853802e-05,
      "loss": 0.6833,
      "step": 660
    },
    {
      "epoch": 17.52,
      "learning_rate": 4.576023391812866e-05,
      "loss": 0.6246,
      "step": 670
    },
    {
      "epoch": 17.78,
      "learning_rate": 4.56140350877193e-05,
      "loss": 0.6131,
      "step": 680
    },
    {
      "epoch": 17.99,
      "eval_accuracy": 0.06115107913669065,
      "eval_loss": 12.481949806213379,
      "eval_runtime": 5.8554,
      "eval_samples_per_second": 237.388,
      "eval_steps_per_second": 7.514,
      "step": 688
    },
    {
      "epoch": 18.04,
      "learning_rate": 4.5467836257309945e-05,
      "loss": 0.5705,
      "step": 690
    },
    {
      "epoch": 18.3,
      "learning_rate": 4.5321637426900585e-05,
      "loss": 0.5652,
      "step": 700
    },
    {
      "epoch": 18.56,
      "learning_rate": 4.517543859649123e-05,
      "loss": 0.5502,
      "step": 710
    },
    {
      "epoch": 18.82,
      "learning_rate": 4.502923976608187e-05,
      "loss": 0.5916,
      "step": 720
    },
    {
      "epoch": 18.98,
      "eval_accuracy": 0.06474820143884892,
      "eval_loss": 12.624631881713867,
      "eval_runtime": 5.8769,
      "eval_samples_per_second": 236.519,
      "eval_steps_per_second": 7.487,
      "step": 726
    },
    {
      "epoch": 19.08,
      "learning_rate": 4.488304093567251e-05,
      "loss": 0.6098,
      "step": 730
    },
    {
      "epoch": 19.35,
      "learning_rate": 4.473684210526316e-05,
      "loss": 0.5802,
      "step": 740
    },
    {
      "epoch": 19.61,
      "learning_rate": 4.4590643274853806e-05,
      "loss": 0.4978,
      "step": 750
    },
    {
      "epoch": 19.87,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.5094,
      "step": 760
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.06690647482014389,
      "eval_loss": 12.664143562316895,
      "eval_runtime": 5.9348,
      "eval_samples_per_second": 234.211,
      "eval_steps_per_second": 7.414,
      "step": 765
    },
    {
      "epoch": 20.13,
      "learning_rate": 4.429824561403509e-05,
      "loss": 0.5806,
      "step": 770
    },
    {
      "epoch": 20.39,
      "learning_rate": 4.4152046783625734e-05,
      "loss": 0.5324,
      "step": 780
    },
    {
      "epoch": 20.65,
      "learning_rate": 4.400584795321638e-05,
      "loss": 0.5366,
      "step": 790
    },
    {
      "epoch": 20.92,
      "learning_rate": 4.3859649122807014e-05,
      "loss": 0.5201,
      "step": 800
    },
    {
      "epoch": 20.99,
      "eval_accuracy": 0.06618705035971223,
      "eval_loss": 12.886053085327148,
      "eval_runtime": 5.9484,
      "eval_samples_per_second": 233.676,
      "eval_steps_per_second": 7.397,
      "step": 803
    },
    {
      "epoch": 21.18,
      "learning_rate": 4.371345029239766e-05,
      "loss": 0.5913,
      "step": 810
    },
    {
      "epoch": 21.44,
      "learning_rate": 4.356725146198831e-05,
      "loss": 0.5353,
      "step": 820
    },
    {
      "epoch": 21.7,
      "learning_rate": 4.342105263157895e-05,
      "loss": 0.542,
      "step": 830
    },
    {
      "epoch": 21.96,
      "learning_rate": 4.327485380116959e-05,
      "loss": 0.4731,
      "step": 840
    },
    {
      "epoch": 21.99,
      "eval_accuracy": 0.06546762589928058,
      "eval_loss": 12.743083000183105,
      "eval_runtime": 5.8758,
      "eval_samples_per_second": 236.564,
      "eval_steps_per_second": 7.488,
      "step": 841
    },
    {
      "epoch": 22.22,
      "learning_rate": 4.3128654970760236e-05,
      "loss": 0.4812,
      "step": 850
    },
    {
      "epoch": 22.48,
      "learning_rate": 4.298245614035088e-05,
      "loss": 0.5204,
      "step": 860
    },
    {
      "epoch": 22.75,
      "learning_rate": 4.283625730994152e-05,
      "loss": 0.5132,
      "step": 870
    },
    {
      "epoch": 22.98,
      "eval_accuracy": 0.07050359712230216,
      "eval_loss": 12.778566360473633,
      "eval_runtime": 5.856,
      "eval_samples_per_second": 237.364,
      "eval_steps_per_second": 7.514,
      "step": 879
    },
    {
      "epoch": 23.01,
      "learning_rate": 4.269005847953216e-05,
      "loss": 0.5084,
      "step": 880
    },
    {
      "epoch": 23.27,
      "learning_rate": 4.254385964912281e-05,
      "loss": 0.5183,
      "step": 890
    },
    {
      "epoch": 23.53,
      "learning_rate": 4.239766081871345e-05,
      "loss": 0.5286,
      "step": 900
    },
    {
      "epoch": 23.79,
      "learning_rate": 4.22514619883041e-05,
      "loss": 0.5036,
      "step": 910
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.07266187050359713,
      "eval_loss": 12.998967170715332,
      "eval_runtime": 5.8521,
      "eval_samples_per_second": 237.522,
      "eval_steps_per_second": 7.519,
      "step": 918
    },
    {
      "epoch": 24.05,
      "learning_rate": 4.210526315789474e-05,
      "loss": 0.4826,
      "step": 920
    },
    {
      "epoch": 24.31,
      "learning_rate": 4.195906432748538e-05,
      "loss": 0.5105,
      "step": 930
    },
    {
      "epoch": 24.58,
      "learning_rate": 4.1812865497076025e-05,
      "loss": 0.5133,
      "step": 940
    },
    {
      "epoch": 24.84,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.4863,
      "step": 950
    },
    {
      "epoch": 24.99,
      "eval_accuracy": 0.07266187050359713,
      "eval_loss": 13.041873931884766,
      "eval_runtime": 7.3809,
      "eval_samples_per_second": 188.323,
      "eval_steps_per_second": 5.961,
      "step": 956
    },
    {
      "epoch": 25.1,
      "learning_rate": 4.152046783625731e-05,
      "loss": 0.5028,
      "step": 960
    },
    {
      "epoch": 25.36,
      "learning_rate": 4.137426900584795e-05,
      "loss": 0.5185,
      "step": 970
    },
    {
      "epoch": 25.62,
      "learning_rate": 4.12280701754386e-05,
      "loss": 0.484,
      "step": 980
    },
    {
      "epoch": 25.88,
      "learning_rate": 4.1081871345029247e-05,
      "loss": 0.4852,
      "step": 990
    },
    {
      "epoch": 25.99,
      "eval_accuracy": 0.07338129496402877,
      "eval_loss": 13.057332038879395,
      "eval_runtime": 6.099,
      "eval_samples_per_second": 227.908,
      "eval_steps_per_second": 7.214,
      "step": 994
    },
    {
      "epoch": 26.14,
      "learning_rate": 4.093567251461988e-05,
      "loss": 0.5227,
      "step": 1000
    },
    {
      "epoch": 26.41,
      "learning_rate": 4.078947368421053e-05,
      "loss": 0.4827,
      "step": 1010
    },
    {
      "epoch": 26.67,
      "learning_rate": 4.0643274853801174e-05,
      "loss": 0.5091,
      "step": 1020
    },
    {
      "epoch": 26.93,
      "learning_rate": 4.0497076023391814e-05,
      "loss": 0.4983,
      "step": 1030
    },
    {
      "epoch": 26.98,
      "eval_accuracy": 0.07194244604316546,
      "eval_loss": 13.131017684936523,
      "eval_runtime": 6.2465,
      "eval_samples_per_second": 222.525,
      "eval_steps_per_second": 7.044,
      "step": 1032
    },
    {
      "epoch": 27.19,
      "learning_rate": 4.0350877192982455e-05,
      "loss": 0.4918,
      "step": 1040
    },
    {
      "epoch": 27.45,
      "learning_rate": 4.02046783625731e-05,
      "loss": 0.4687,
      "step": 1050
    },
    {
      "epoch": 27.71,
      "learning_rate": 4.005847953216375e-05,
      "loss": 0.4687,
      "step": 1060
    },
    {
      "epoch": 27.97,
      "learning_rate": 3.991228070175439e-05,
      "loss": 0.459,
      "step": 1070
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.07482014388489208,
      "eval_loss": 13.068760871887207,
      "eval_runtime": 6.2511,
      "eval_samples_per_second": 222.36,
      "eval_steps_per_second": 7.039,
      "step": 1071
    },
    {
      "epoch": 28.24,
      "learning_rate": 3.976608187134503e-05,
      "loss": 0.4987,
      "step": 1080
    },
    {
      "epoch": 28.5,
      "learning_rate": 3.9619883040935676e-05,
      "loss": 0.4874,
      "step": 1090
    },
    {
      "epoch": 28.76,
      "learning_rate": 3.9473684210526316e-05,
      "loss": 0.4556,
      "step": 1100
    },
    {
      "epoch": 28.99,
      "eval_accuracy": 0.07482014388489208,
      "eval_loss": 13.41281509399414,
      "eval_runtime": 6.1781,
      "eval_samples_per_second": 224.989,
      "eval_steps_per_second": 7.122,
      "step": 1109
    },
    {
      "epoch": 29.02,
      "learning_rate": 3.932748538011696e-05,
      "loss": 0.47,
      "step": 1110
    },
    {
      "epoch": 29.28,
      "learning_rate": 3.9181286549707604e-05,
      "loss": 0.4709,
      "step": 1120
    },
    {
      "epoch": 29.54,
      "learning_rate": 3.9035087719298244e-05,
      "loss": 0.4499,
      "step": 1130
    },
    {
      "epoch": 29.8,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.4729,
      "step": 1140
    },
    {
      "epoch": 29.99,
      "eval_accuracy": 0.07410071942446043,
      "eval_loss": 13.353046417236328,
      "eval_runtime": 6.179,
      "eval_samples_per_second": 224.956,
      "eval_steps_per_second": 7.121,
      "step": 1147
    },
    {
      "epoch": 30.07,
      "learning_rate": 3.874269005847954e-05,
      "loss": 0.4329,
      "step": 1150
    },
    {
      "epoch": 30.33,
      "learning_rate": 3.859649122807018e-05,
      "loss": 0.4418,
      "step": 1160
    },
    {
      "epoch": 30.59,
      "learning_rate": 3.845029239766082e-05,
      "loss": 0.4428,
      "step": 1170
    },
    {
      "epoch": 30.85,
      "learning_rate": 3.8304093567251465e-05,
      "loss": 0.4659,
      "step": 1180
    },
    {
      "epoch": 30.98,
      "eval_accuracy": 0.07625899280575539,
      "eval_loss": 13.230795860290527,
      "eval_runtime": 6.1942,
      "eval_samples_per_second": 224.404,
      "eval_steps_per_second": 7.103,
      "step": 1185
    },
    {
      "epoch": 31.11,
      "learning_rate": 3.815789473684211e-05,
      "loss": 0.436,
      "step": 1190
    },
    {
      "epoch": 31.37,
      "learning_rate": 3.8011695906432746e-05,
      "loss": 0.4484,
      "step": 1200
    },
    {
      "epoch": 31.63,
      "learning_rate": 3.786549707602339e-05,
      "loss": 0.4099,
      "step": 1210
    },
    {
      "epoch": 31.9,
      "learning_rate": 3.771929824561404e-05,
      "loss": 0.4337,
      "step": 1220
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.07482014388489208,
      "eval_loss": 13.32636833190918,
      "eval_runtime": 6.1367,
      "eval_samples_per_second": 226.505,
      "eval_steps_per_second": 7.17,
      "step": 1224
    },
    {
      "epoch": 32.16,
      "learning_rate": 3.757309941520468e-05,
      "loss": 0.4737,
      "step": 1230
    },
    {
      "epoch": 32.42,
      "learning_rate": 3.742690058479532e-05,
      "loss": 0.4133,
      "step": 1240
    },
    {
      "epoch": 32.68,
      "learning_rate": 3.728070175438597e-05,
      "loss": 0.4175,
      "step": 1250
    },
    {
      "epoch": 32.94,
      "learning_rate": 3.713450292397661e-05,
      "loss": 0.456,
      "step": 1260
    },
    {
      "epoch": 32.99,
      "eval_accuracy": 0.07410071942446043,
      "eval_loss": 13.350556373596191,
      "eval_runtime": 6.1533,
      "eval_samples_per_second": 225.896,
      "eval_steps_per_second": 7.151,
      "step": 1262
    },
    {
      "epoch": 33.2,
      "learning_rate": 3.6988304093567254e-05,
      "loss": 0.4446,
      "step": 1270
    },
    {
      "epoch": 33.46,
      "learning_rate": 3.6842105263157895e-05,
      "loss": 0.4216,
      "step": 1280
    },
    {
      "epoch": 33.73,
      "learning_rate": 3.669590643274854e-05,
      "loss": 0.4576,
      "step": 1290
    },
    {
      "epoch": 33.99,
      "learning_rate": 3.654970760233918e-05,
      "loss": 0.4423,
      "step": 1300
    },
    {
      "epoch": 33.99,
      "eval_accuracy": 0.07841726618705036,
      "eval_loss": 13.360701560974121,
      "eval_runtime": 5.9537,
      "eval_samples_per_second": 233.467,
      "eval_steps_per_second": 7.39,
      "step": 1300
    },
    {
      "epoch": 34.25,
      "learning_rate": 3.640350877192983e-05,
      "loss": 0.4469,
      "step": 1310
    },
    {
      "epoch": 34.51,
      "learning_rate": 3.625730994152047e-05,
      "loss": 0.4226,
      "step": 1320
    },
    {
      "epoch": 34.77,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.4037,
      "step": 1330
    },
    {
      "epoch": 34.98,
      "eval_accuracy": 0.07338129496402877,
      "eval_loss": 13.252113342285156,
      "eval_runtime": 6.1434,
      "eval_samples_per_second": 226.26,
      "eval_steps_per_second": 7.162,
      "step": 1338
    },
    {
      "epoch": 35.03,
      "learning_rate": 3.5964912280701756e-05,
      "loss": 0.4277,
      "step": 1340
    },
    {
      "epoch": 35.29,
      "learning_rate": 3.5818713450292403e-05,
      "loss": 0.3596,
      "step": 1350
    },
    {
      "epoch": 35.56,
      "learning_rate": 3.5672514619883044e-05,
      "loss": 0.4047,
      "step": 1360
    },
    {
      "epoch": 35.82,
      "learning_rate": 3.5526315789473684e-05,
      "loss": 0.3891,
      "step": 1370
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.0776978417266187,
      "eval_loss": 13.370182991027832,
      "eval_runtime": 6.0896,
      "eval_samples_per_second": 228.258,
      "eval_steps_per_second": 7.225,
      "step": 1377
    },
    {
      "epoch": 36.08,
      "learning_rate": 3.538011695906433e-05,
      "loss": 0.4298,
      "step": 1380
    },
    {
      "epoch": 36.34,
      "learning_rate": 3.523391812865498e-05,
      "loss": 0.366,
      "step": 1390
    },
    {
      "epoch": 36.6,
      "learning_rate": 3.508771929824561e-05,
      "loss": 0.4485,
      "step": 1400
    },
    {
      "epoch": 36.86,
      "learning_rate": 3.494152046783626e-05,
      "loss": 0.3992,
      "step": 1410
    },
    {
      "epoch": 36.99,
      "eval_accuracy": 0.0776978417266187,
      "eval_loss": 13.476212501525879,
      "eval_runtime": 6.0276,
      "eval_samples_per_second": 230.606,
      "eval_steps_per_second": 7.3,
      "step": 1415
    },
    {
      "epoch": 37.12,
      "learning_rate": 3.4795321637426905e-05,
      "loss": 0.3786,
      "step": 1420
    },
    {
      "epoch": 37.39,
      "learning_rate": 3.4649122807017546e-05,
      "loss": 0.3805,
      "step": 1430
    },
    {
      "epoch": 37.65,
      "learning_rate": 3.4502923976608186e-05,
      "loss": 0.398,
      "step": 1440
    },
    {
      "epoch": 37.91,
      "learning_rate": 3.435672514619883e-05,
      "loss": 0.4014,
      "step": 1450
    },
    {
      "epoch": 37.99,
      "eval_accuracy": 0.07913669064748201,
      "eval_loss": 13.538166999816895,
      "eval_runtime": 5.9007,
      "eval_samples_per_second": 235.564,
      "eval_steps_per_second": 7.457,
      "step": 1453
    },
    {
      "epoch": 38.17,
      "learning_rate": 3.421052631578947e-05,
      "loss": 0.4176,
      "step": 1460
    },
    {
      "epoch": 38.43,
      "learning_rate": 3.406432748538012e-05,
      "loss": 0.3663,
      "step": 1470
    },
    {
      "epoch": 38.69,
      "learning_rate": 3.391812865497076e-05,
      "loss": 0.3658,
      "step": 1480
    },
    {
      "epoch": 38.95,
      "learning_rate": 3.377192982456141e-05,
      "loss": 0.3549,
      "step": 1490
    },
    {
      "epoch": 38.98,
      "eval_accuracy": 0.07913669064748201,
      "eval_loss": 13.555007934570312,
      "eval_runtime": 5.8979,
      "eval_samples_per_second": 235.677,
      "eval_steps_per_second": 7.46,
      "step": 1491
    },
    {
      "epoch": 39.22,
      "learning_rate": 3.362573099415205e-05,
      "loss": 0.4037,
      "step": 1500
    },
    {
      "epoch": 39.48,
      "learning_rate": 3.3479532163742695e-05,
      "loss": 0.3612,
      "step": 1510
    },
    {
      "epoch": 39.74,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.3879,
      "step": 1520
    },
    {
      "epoch": 40.0,
      "learning_rate": 3.3187134502923975e-05,
      "loss": 0.4048,
      "step": 1530
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.07985611510791367,
      "eval_loss": 13.640563011169434,
      "eval_runtime": 5.9909,
      "eval_samples_per_second": 232.02,
      "eval_steps_per_second": 7.345,
      "step": 1530
    },
    {
      "epoch": 40.26,
      "learning_rate": 3.304093567251462e-05,
      "loss": 0.3897,
      "step": 1540
    },
    {
      "epoch": 40.52,
      "learning_rate": 3.289473684210527e-05,
      "loss": 0.3554,
      "step": 1550
    },
    {
      "epoch": 40.78,
      "learning_rate": 3.274853801169591e-05,
      "loss": 0.3711,
      "step": 1560
    },
    {
      "epoch": 40.99,
      "eval_accuracy": 0.0776978417266187,
      "eval_loss": 13.51199722290039,
      "eval_runtime": 5.8868,
      "eval_samples_per_second": 236.12,
      "eval_steps_per_second": 7.474,
      "step": 1568
    },
    {
      "epoch": 41.05,
      "learning_rate": 3.260233918128655e-05,
      "loss": 0.4245,
      "step": 1570
    },
    {
      "epoch": 41.31,
      "learning_rate": 3.24561403508772e-05,
      "loss": 0.4058,
      "step": 1580
    },
    {
      "epoch": 41.57,
      "learning_rate": 3.230994152046784e-05,
      "loss": 0.3726,
      "step": 1590
    },
    {
      "epoch": 41.83,
      "learning_rate": 3.216374269005848e-05,
      "loss": 0.3834,
      "step": 1600
    },
    {
      "epoch": 41.99,
      "eval_accuracy": 0.07985611510791367,
      "eval_loss": 13.923029899597168,
      "eval_runtime": 5.9072,
      "eval_samples_per_second": 235.305,
      "eval_steps_per_second": 7.448,
      "step": 1606
    },
    {
      "epoch": 42.09,
      "learning_rate": 3.2017543859649124e-05,
      "loss": 0.4307,
      "step": 1610
    },
    {
      "epoch": 42.35,
      "learning_rate": 3.187134502923977e-05,
      "loss": 0.3423,
      "step": 1620
    },
    {
      "epoch": 42.61,
      "learning_rate": 3.172514619883041e-05,
      "loss": 0.3652,
      "step": 1630
    },
    {
      "epoch": 42.88,
      "learning_rate": 3.157894736842105e-05,
      "loss": 0.3475,
      "step": 1640
    },
    {
      "epoch": 42.98,
      "eval_accuracy": 0.07913669064748201,
      "eval_loss": 13.860240936279297,
      "eval_runtime": 5.9793,
      "eval_samples_per_second": 232.468,
      "eval_steps_per_second": 7.359,
      "step": 1644
    },
    {
      "epoch": 43.14,
      "learning_rate": 3.14327485380117e-05,
      "loss": 0.3636,
      "step": 1650
    },
    {
      "epoch": 43.4,
      "learning_rate": 3.128654970760234e-05,
      "loss": 0.3586,
      "step": 1660
    },
    {
      "epoch": 43.66,
      "learning_rate": 3.1140350877192986e-05,
      "loss": 0.3643,
      "step": 1670
    },
    {
      "epoch": 43.92,
      "learning_rate": 3.0994152046783626e-05,
      "loss": 0.3465,
      "step": 1680
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 13.69305419921875,
      "eval_runtime": 5.9121,
      "eval_samples_per_second": 235.109,
      "eval_steps_per_second": 7.442,
      "step": 1683
    },
    {
      "epoch": 44.18,
      "learning_rate": 3.084795321637427e-05,
      "loss": 0.412,
      "step": 1690
    },
    {
      "epoch": 44.44,
      "learning_rate": 3.0701754385964913e-05,
      "loss": 0.3521,
      "step": 1700
    },
    {
      "epoch": 44.71,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.3847,
      "step": 1710
    },
    {
      "epoch": 44.97,
      "learning_rate": 3.0409356725146197e-05,
      "loss": 0.3682,
      "step": 1720
    },
    {
      "epoch": 44.99,
      "eval_accuracy": 0.07841726618705036,
      "eval_loss": 13.777403831481934,
      "eval_runtime": 6.1334,
      "eval_samples_per_second": 226.629,
      "eval_steps_per_second": 7.174,
      "step": 1721
    },
    {
      "epoch": 45.23,
      "learning_rate": 3.0263157894736844e-05,
      "loss": 0.4067,
      "step": 1730
    },
    {
      "epoch": 45.49,
      "learning_rate": 3.0116959064327488e-05,
      "loss": 0.3724,
      "step": 1740
    },
    {
      "epoch": 45.75,
      "learning_rate": 2.997076023391813e-05,
      "loss": 0.3613,
      "step": 1750
    },
    {
      "epoch": 45.99,
      "eval_accuracy": 0.07913669064748201,
      "eval_loss": 14.02348518371582,
      "eval_runtime": 6.1306,
      "eval_samples_per_second": 226.73,
      "eval_steps_per_second": 7.177,
      "step": 1759
    },
    {
      "epoch": 46.01,
      "learning_rate": 2.9824561403508772e-05,
      "loss": 0.3766,
      "step": 1760
    },
    {
      "epoch": 46.27,
      "learning_rate": 2.9678362573099415e-05,
      "loss": 0.4555,
      "step": 1770
    },
    {
      "epoch": 46.54,
      "learning_rate": 2.9532163742690062e-05,
      "loss": 0.3324,
      "step": 1780
    },
    {
      "epoch": 46.8,
      "learning_rate": 2.9385964912280706e-05,
      "loss": 0.368,
      "step": 1790
    },
    {
      "epoch": 46.98,
      "eval_accuracy": 0.08129496402877698,
      "eval_loss": 13.92888355255127,
      "eval_runtime": 6.1887,
      "eval_samples_per_second": 224.602,
      "eval_steps_per_second": 7.11,
      "step": 1797
    },
    {
      "epoch": 47.06,
      "learning_rate": 2.9239766081871346e-05,
      "loss": 0.2888,
      "step": 1800
    },
    {
      "epoch": 47.32,
      "learning_rate": 2.909356725146199e-05,
      "loss": 0.3538,
      "step": 1810
    },
    {
      "epoch": 47.58,
      "learning_rate": 2.8947368421052634e-05,
      "loss": 0.3298,
      "step": 1820
    },
    {
      "epoch": 47.84,
      "learning_rate": 2.8801169590643277e-05,
      "loss": 0.3961,
      "step": 1830
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 14.254853248596191,
      "eval_runtime": 6.2277,
      "eval_samples_per_second": 223.197,
      "eval_steps_per_second": 7.065,
      "step": 1836
    },
    {
      "epoch": 48.1,
      "learning_rate": 2.8654970760233917e-05,
      "loss": 0.3433,
      "step": 1840
    },
    {
      "epoch": 48.37,
      "learning_rate": 2.850877192982456e-05,
      "loss": 0.366,
      "step": 1850
    },
    {
      "epoch": 48.63,
      "learning_rate": 2.8362573099415208e-05,
      "loss": 0.3278,
      "step": 1860
    },
    {
      "epoch": 48.89,
      "learning_rate": 2.821637426900585e-05,
      "loss": 0.365,
      "step": 1870
    },
    {
      "epoch": 48.99,
      "eval_accuracy": 0.08129496402877698,
      "eval_loss": 14.111425399780273,
      "eval_runtime": 6.2596,
      "eval_samples_per_second": 222.059,
      "eval_steps_per_second": 7.029,
      "step": 1874
    },
    {
      "epoch": 49.15,
      "learning_rate": 2.8070175438596492e-05,
      "loss": 0.3721,
      "step": 1880
    },
    {
      "epoch": 49.41,
      "learning_rate": 2.7923976608187135e-05,
      "loss": 0.3683,
      "step": 1890
    },
    {
      "epoch": 49.67,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.3506,
      "step": 1900
    },
    {
      "epoch": 49.93,
      "learning_rate": 2.7631578947368426e-05,
      "loss": 0.3259,
      "step": 1910
    },
    {
      "epoch": 49.99,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 13.97095012664795,
      "eval_runtime": 6.1784,
      "eval_samples_per_second": 224.977,
      "eval_steps_per_second": 7.122,
      "step": 1912
    },
    {
      "epoch": 50.2,
      "learning_rate": 2.7485380116959063e-05,
      "loss": 0.3456,
      "step": 1920
    },
    {
      "epoch": 50.46,
      "learning_rate": 2.733918128654971e-05,
      "loss": 0.3499,
      "step": 1930
    },
    {
      "epoch": 50.72,
      "learning_rate": 2.7192982456140354e-05,
      "loss": 0.3255,
      "step": 1940
    },
    {
      "epoch": 50.98,
      "learning_rate": 2.7046783625730997e-05,
      "loss": 0.2998,
      "step": 1950
    },
    {
      "epoch": 50.98,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 14.028827667236328,
      "eval_runtime": 6.2989,
      "eval_samples_per_second": 220.674,
      "eval_steps_per_second": 6.985,
      "step": 1950
    },
    {
      "epoch": 51.24,
      "learning_rate": 2.6900584795321637e-05,
      "loss": 0.3892,
      "step": 1960
    },
    {
      "epoch": 51.5,
      "learning_rate": 2.675438596491228e-05,
      "loss": 0.336,
      "step": 1970
    },
    {
      "epoch": 51.76,
      "learning_rate": 2.6608187134502928e-05,
      "loss": 0.3203,
      "step": 1980
    },
    {
      "epoch": 52.0,
      "eval_accuracy": 0.08129496402877698,
      "eval_loss": 13.939803123474121,
      "eval_runtime": 6.0691,
      "eval_samples_per_second": 229.028,
      "eval_steps_per_second": 7.25,
      "step": 1989
    },
    {
      "epoch": 52.03,
      "learning_rate": 2.6461988304093572e-05,
      "loss": 0.3512,
      "step": 1990
    },
    {
      "epoch": 52.29,
      "learning_rate": 2.6315789473684212e-05,
      "loss": 0.3178,
      "step": 2000
    },
    {
      "epoch": 52.55,
      "learning_rate": 2.6169590643274856e-05,
      "loss": 0.3224,
      "step": 2010
    },
    {
      "epoch": 52.81,
      "learning_rate": 2.60233918128655e-05,
      "loss": 0.3104,
      "step": 2020
    },
    {
      "epoch": 52.99,
      "eval_accuracy": 0.08201438848920864,
      "eval_loss": 14.025542259216309,
      "eval_runtime": 5.9734,
      "eval_samples_per_second": 232.696,
      "eval_steps_per_second": 7.366,
      "step": 2027
    },
    {
      "epoch": 53.07,
      "learning_rate": 2.5877192982456143e-05,
      "loss": 0.3159,
      "step": 2030
    },
    {
      "epoch": 53.33,
      "learning_rate": 2.5730994152046783e-05,
      "loss": 0.3305,
      "step": 2040
    },
    {
      "epoch": 53.59,
      "learning_rate": 2.5584795321637427e-05,
      "loss": 0.3085,
      "step": 2050
    },
    {
      "epoch": 53.86,
      "learning_rate": 2.5438596491228074e-05,
      "loss": 0.3232,
      "step": 2060
    },
    {
      "epoch": 53.99,
      "eval_accuracy": 0.08273381294964029,
      "eval_loss": 13.93545150756836,
      "eval_runtime": 6.1249,
      "eval_samples_per_second": 226.943,
      "eval_steps_per_second": 7.184,
      "step": 2065
    },
    {
      "epoch": 54.12,
      "learning_rate": 2.5292397660818717e-05,
      "loss": 0.3035,
      "step": 2070
    },
    {
      "epoch": 54.38,
      "learning_rate": 2.5146198830409358e-05,
      "loss": 0.3235,
      "step": 2080
    },
    {
      "epoch": 54.64,
      "learning_rate": 2.5e-05,
      "loss": 0.3359,
      "step": 2090
    },
    {
      "epoch": 54.9,
      "learning_rate": 2.485380116959064e-05,
      "loss": 0.3521,
      "step": 2100
    },
    {
      "epoch": 54.98,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 13.862702369689941,
      "eval_runtime": 5.97,
      "eval_samples_per_second": 232.832,
      "eval_steps_per_second": 7.37,
      "step": 2103
    },
    {
      "epoch": 55.16,
      "learning_rate": 2.470760233918129e-05,
      "loss": 0.3049,
      "step": 2110
    },
    {
      "epoch": 55.42,
      "learning_rate": 2.456140350877193e-05,
      "loss": 0.3011,
      "step": 2120
    },
    {
      "epoch": 55.69,
      "learning_rate": 2.4415204678362576e-05,
      "loss": 0.2945,
      "step": 2130
    },
    {
      "epoch": 55.95,
      "learning_rate": 2.4269005847953216e-05,
      "loss": 0.3322,
      "step": 2140
    },
    {
      "epoch": 56.0,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 14.01791000366211,
      "eval_runtime": 5.9414,
      "eval_samples_per_second": 233.951,
      "eval_steps_per_second": 7.406,
      "step": 2142
    },
    {
      "epoch": 56.21,
      "learning_rate": 2.412280701754386e-05,
      "loss": 0.2962,
      "step": 2150
    },
    {
      "epoch": 56.47,
      "learning_rate": 2.3976608187134503e-05,
      "loss": 0.3282,
      "step": 2160
    },
    {
      "epoch": 56.73,
      "learning_rate": 2.3830409356725147e-05,
      "loss": 0.2965,
      "step": 2170
    },
    {
      "epoch": 56.99,
      "learning_rate": 2.368421052631579e-05,
      "loss": 0.3129,
      "step": 2180
    },
    {
      "epoch": 56.99,
      "eval_accuracy": 0.08201438848920864,
      "eval_loss": 13.96402359008789,
      "eval_runtime": 5.8814,
      "eval_samples_per_second": 236.34,
      "eval_steps_per_second": 7.481,
      "step": 2180
    },
    {
      "epoch": 57.25,
      "learning_rate": 2.3538011695906434e-05,
      "loss": 0.3487,
      "step": 2190
    },
    {
      "epoch": 57.52,
      "learning_rate": 2.3391812865497074e-05,
      "loss": 0.3089,
      "step": 2200
    },
    {
      "epoch": 57.78,
      "learning_rate": 2.324561403508772e-05,
      "loss": 0.3159,
      "step": 2210
    },
    {
      "epoch": 57.99,
      "eval_accuracy": 0.07985611510791367,
      "eval_loss": 14.199702262878418,
      "eval_runtime": 6.0594,
      "eval_samples_per_second": 229.395,
      "eval_steps_per_second": 7.261,
      "step": 2218
    },
    {
      "epoch": 58.04,
      "learning_rate": 2.309941520467836e-05,
      "loss": 0.34,
      "step": 2220
    },
    {
      "epoch": 58.3,
      "learning_rate": 2.295321637426901e-05,
      "loss": 0.3108,
      "step": 2230
    },
    {
      "epoch": 58.56,
      "learning_rate": 2.280701754385965e-05,
      "loss": 0.3316,
      "step": 2240
    },
    {
      "epoch": 58.82,
      "learning_rate": 2.2660818713450292e-05,
      "loss": 0.3118,
      "step": 2250
    },
    {
      "epoch": 58.98,
      "eval_accuracy": 0.08201438848920864,
      "eval_loss": 14.163899421691895,
      "eval_runtime": 5.9877,
      "eval_samples_per_second": 232.144,
      "eval_steps_per_second": 7.348,
      "step": 2256
    },
    {
      "epoch": 59.08,
      "learning_rate": 2.2514619883040936e-05,
      "loss": 0.321,
      "step": 2260
    },
    {
      "epoch": 59.35,
      "learning_rate": 2.236842105263158e-05,
      "loss": 0.2972,
      "step": 2270
    },
    {
      "epoch": 59.61,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.2783,
      "step": 2280
    },
    {
      "epoch": 59.87,
      "learning_rate": 2.2076023391812867e-05,
      "loss": 0.3196,
      "step": 2290
    },
    {
      "epoch": 60.0,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 14.033405303955078,
      "eval_runtime": 6.077,
      "eval_samples_per_second": 228.732,
      "eval_steps_per_second": 7.24,
      "step": 2295
    },
    {
      "epoch": 60.13,
      "learning_rate": 2.1929824561403507e-05,
      "loss": 0.3396,
      "step": 2300
    },
    {
      "epoch": 60.39,
      "learning_rate": 2.1783625730994154e-05,
      "loss": 0.3065,
      "step": 2310
    },
    {
      "epoch": 60.65,
      "learning_rate": 2.1637426900584794e-05,
      "loss": 0.3169,
      "step": 2320
    },
    {
      "epoch": 60.92,
      "learning_rate": 2.149122807017544e-05,
      "loss": 0.301,
      "step": 2330
    },
    {
      "epoch": 60.99,
      "eval_accuracy": 0.08201438848920864,
      "eval_loss": 13.995382308959961,
      "eval_runtime": 6.1902,
      "eval_samples_per_second": 224.549,
      "eval_steps_per_second": 7.108,
      "step": 2333
    },
    {
      "epoch": 61.18,
      "learning_rate": 2.134502923976608e-05,
      "loss": 0.3011,
      "step": 2340
    },
    {
      "epoch": 61.44,
      "learning_rate": 2.1198830409356725e-05,
      "loss": 0.2741,
      "step": 2350
    },
    {
      "epoch": 61.7,
      "learning_rate": 2.105263157894737e-05,
      "loss": 0.2713,
      "step": 2360
    },
    {
      "epoch": 61.96,
      "learning_rate": 2.0906432748538013e-05,
      "loss": 0.3142,
      "step": 2370
    },
    {
      "epoch": 61.99,
      "eval_accuracy": 0.07985611510791367,
      "eval_loss": 14.143210411071777,
      "eval_runtime": 6.2246,
      "eval_samples_per_second": 223.307,
      "eval_steps_per_second": 7.069,
      "step": 2371
    },
    {
      "epoch": 62.22,
      "learning_rate": 2.0760233918128656e-05,
      "loss": 0.2731,
      "step": 2380
    },
    {
      "epoch": 62.48,
      "learning_rate": 2.06140350877193e-05,
      "loss": 0.2847,
      "step": 2390
    },
    {
      "epoch": 62.75,
      "learning_rate": 2.046783625730994e-05,
      "loss": 0.3192,
      "step": 2400
    },
    {
      "epoch": 62.98,
      "eval_accuracy": 0.07841726618705036,
      "eval_loss": 14.026898384094238,
      "eval_runtime": 6.2695,
      "eval_samples_per_second": 221.708,
      "eval_steps_per_second": 7.018,
      "step": 2409
    },
    {
      "epoch": 63.01,
      "learning_rate": 2.0321637426900587e-05,
      "loss": 0.3586,
      "step": 2410
    },
    {
      "epoch": 63.27,
      "learning_rate": 2.0175438596491227e-05,
      "loss": 0.3255,
      "step": 2420
    },
    {
      "epoch": 63.53,
      "learning_rate": 2.0029239766081874e-05,
      "loss": 0.2821,
      "step": 2430
    },
    {
      "epoch": 63.79,
      "learning_rate": 1.9883040935672515e-05,
      "loss": 0.3342,
      "step": 2440
    },
    {
      "epoch": 64.0,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 14.044992446899414,
      "eval_runtime": 6.2504,
      "eval_samples_per_second": 222.386,
      "eval_steps_per_second": 7.04,
      "step": 2448
    },
    {
      "epoch": 64.05,
      "learning_rate": 1.9736842105263158e-05,
      "loss": 0.3163,
      "step": 2450
    },
    {
      "epoch": 64.31,
      "learning_rate": 1.9590643274853802e-05,
      "loss": 0.3105,
      "step": 2460
    },
    {
      "epoch": 64.58,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.32,
      "step": 2470
    },
    {
      "epoch": 64.84,
      "learning_rate": 1.929824561403509e-05,
      "loss": 0.3045,
      "step": 2480
    },
    {
      "epoch": 64.99,
      "eval_accuracy": 0.08489208633093526,
      "eval_loss": 14.174620628356934,
      "eval_runtime": 6.2508,
      "eval_samples_per_second": 222.372,
      "eval_steps_per_second": 7.039,
      "step": 2486
    },
    {
      "epoch": 65.1,
      "learning_rate": 1.9152046783625733e-05,
      "loss": 0.3044,
      "step": 2490
    },
    {
      "epoch": 65.36,
      "learning_rate": 1.9005847953216373e-05,
      "loss": 0.3328,
      "step": 2500
    },
    {
      "epoch": 65.62,
      "learning_rate": 1.885964912280702e-05,
      "loss": 0.2961,
      "step": 2510
    },
    {
      "epoch": 65.88,
      "learning_rate": 1.871345029239766e-05,
      "loss": 0.2991,
      "step": 2520
    },
    {
      "epoch": 65.99,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 14.319219589233398,
      "eval_runtime": 6.2401,
      "eval_samples_per_second": 222.753,
      "eval_steps_per_second": 7.051,
      "step": 2524
    },
    {
      "epoch": 66.14,
      "learning_rate": 1.8567251461988304e-05,
      "loss": 0.3191,
      "step": 2530
    },
    {
      "epoch": 66.41,
      "learning_rate": 1.8421052631578947e-05,
      "loss": 0.3111,
      "step": 2540
    },
    {
      "epoch": 66.67,
      "learning_rate": 1.827485380116959e-05,
      "loss": 0.3201,
      "step": 2550
    },
    {
      "epoch": 66.93,
      "learning_rate": 1.8128654970760235e-05,
      "loss": 0.3228,
      "step": 2560
    },
    {
      "epoch": 66.98,
      "eval_accuracy": 0.07841726618705036,
      "eval_loss": 14.178192138671875,
      "eval_runtime": 6.2603,
      "eval_samples_per_second": 222.035,
      "eval_steps_per_second": 7.028,
      "step": 2562
    },
    {
      "epoch": 67.19,
      "learning_rate": 1.7982456140350878e-05,
      "loss": 0.2959,
      "step": 2570
    },
    {
      "epoch": 67.45,
      "learning_rate": 1.7836257309941522e-05,
      "loss": 0.2927,
      "step": 2580
    },
    {
      "epoch": 67.71,
      "learning_rate": 1.7690058479532165e-05,
      "loss": 0.2991,
      "step": 2590
    },
    {
      "epoch": 67.97,
      "learning_rate": 1.7543859649122806e-05,
      "loss": 0.2711,
      "step": 2600
    },
    {
      "epoch": 68.0,
      "eval_accuracy": 0.08489208633093526,
      "eval_loss": 14.426069259643555,
      "eval_runtime": 6.1828,
      "eval_samples_per_second": 224.816,
      "eval_steps_per_second": 7.116,
      "step": 2601
    },
    {
      "epoch": 68.24,
      "learning_rate": 1.7397660818713453e-05,
      "loss": 0.2786,
      "step": 2610
    },
    {
      "epoch": 68.5,
      "learning_rate": 1.7251461988304093e-05,
      "loss": 0.2746,
      "step": 2620
    },
    {
      "epoch": 68.76,
      "learning_rate": 1.7105263157894737e-05,
      "loss": 0.2473,
      "step": 2630
    },
    {
      "epoch": 68.99,
      "eval_accuracy": 0.08273381294964029,
      "eval_loss": 14.23031997680664,
      "eval_runtime": 6.0443,
      "eval_samples_per_second": 229.97,
      "eval_steps_per_second": 7.28,
      "step": 2639
    },
    {
      "epoch": 69.02,
      "learning_rate": 1.695906432748538e-05,
      "loss": 0.3484,
      "step": 2640
    },
    {
      "epoch": 69.28,
      "learning_rate": 1.6812865497076024e-05,
      "loss": 0.2993,
      "step": 2650
    },
    {
      "epoch": 69.54,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.3016,
      "step": 2660
    },
    {
      "epoch": 69.8,
      "learning_rate": 1.652046783625731e-05,
      "loss": 0.3287,
      "step": 2670
    },
    {
      "epoch": 69.99,
      "eval_accuracy": 0.08273381294964029,
      "eval_loss": 14.274971961975098,
      "eval_runtime": 5.9713,
      "eval_samples_per_second": 232.779,
      "eval_steps_per_second": 7.369,
      "step": 2677
    },
    {
      "epoch": 70.07,
      "learning_rate": 1.6374269005847955e-05,
      "loss": 0.2662,
      "step": 2680
    },
    {
      "epoch": 70.33,
      "learning_rate": 1.62280701754386e-05,
      "loss": 0.3035,
      "step": 2690
    },
    {
      "epoch": 70.59,
      "learning_rate": 1.608187134502924e-05,
      "loss": 0.2923,
      "step": 2700
    },
    {
      "epoch": 70.85,
      "learning_rate": 1.5935672514619886e-05,
      "loss": 0.2673,
      "step": 2710
    },
    {
      "epoch": 70.98,
      "eval_accuracy": 0.08201438848920864,
      "eval_loss": 14.23031234741211,
      "eval_runtime": 5.945,
      "eval_samples_per_second": 233.809,
      "eval_steps_per_second": 7.401,
      "step": 2715
    },
    {
      "epoch": 71.11,
      "learning_rate": 1.5789473684210526e-05,
      "loss": 0.2952,
      "step": 2720
    },
    {
      "epoch": 71.37,
      "learning_rate": 1.564327485380117e-05,
      "loss": 0.3177,
      "step": 2730
    },
    {
      "epoch": 71.63,
      "learning_rate": 1.5497076023391813e-05,
      "loss": 0.2803,
      "step": 2740
    },
    {
      "epoch": 71.9,
      "learning_rate": 1.5350877192982457e-05,
      "loss": 0.2843,
      "step": 2750
    },
    {
      "epoch": 72.0,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 14.408571243286133,
      "eval_runtime": 5.9636,
      "eval_samples_per_second": 233.082,
      "eval_steps_per_second": 7.378,
      "step": 2754
    },
    {
      "epoch": 72.16,
      "learning_rate": 1.5204678362573099e-05,
      "loss": 0.3007,
      "step": 2760
    },
    {
      "epoch": 72.42,
      "learning_rate": 1.5058479532163744e-05,
      "loss": 0.3098,
      "step": 2770
    },
    {
      "epoch": 72.68,
      "learning_rate": 1.4912280701754386e-05,
      "loss": 0.272,
      "step": 2780
    },
    {
      "epoch": 72.94,
      "learning_rate": 1.4766081871345031e-05,
      "loss": 0.3099,
      "step": 2790
    },
    {
      "epoch": 72.99,
      "eval_accuracy": 0.08273381294964029,
      "eval_loss": 14.518421173095703,
      "eval_runtime": 5.9749,
      "eval_samples_per_second": 232.641,
      "eval_steps_per_second": 7.364,
      "step": 2792
    },
    {
      "epoch": 73.2,
      "learning_rate": 1.4619883040935673e-05,
      "loss": 0.2839,
      "step": 2800
    },
    {
      "epoch": 73.46,
      "learning_rate": 1.4473684210526317e-05,
      "loss": 0.3161,
      "step": 2810
    },
    {
      "epoch": 73.73,
      "learning_rate": 1.4327485380116959e-05,
      "loss": 0.3241,
      "step": 2820
    },
    {
      "epoch": 73.99,
      "learning_rate": 1.4181286549707604e-05,
      "loss": 0.3102,
      "step": 2830
    },
    {
      "epoch": 73.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.276801109313965,
      "eval_runtime": 5.8868,
      "eval_samples_per_second": 236.121,
      "eval_steps_per_second": 7.474,
      "step": 2830
    },
    {
      "epoch": 74.25,
      "learning_rate": 1.4035087719298246e-05,
      "loss": 0.3081,
      "step": 2840
    },
    {
      "epoch": 74.51,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.2935,
      "step": 2850
    },
    {
      "epoch": 74.77,
      "learning_rate": 1.3742690058479531e-05,
      "loss": 0.2911,
      "step": 2860
    },
    {
      "epoch": 74.98,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.100983619689941,
      "eval_runtime": 6.0333,
      "eval_samples_per_second": 230.387,
      "eval_steps_per_second": 7.293,
      "step": 2868
    },
    {
      "epoch": 75.03,
      "learning_rate": 1.3596491228070177e-05,
      "loss": 0.3481,
      "step": 2870
    },
    {
      "epoch": 75.29,
      "learning_rate": 1.3450292397660819e-05,
      "loss": 0.298,
      "step": 2880
    },
    {
      "epoch": 75.56,
      "learning_rate": 1.3304093567251464e-05,
      "loss": 0.282,
      "step": 2890
    },
    {
      "epoch": 75.82,
      "learning_rate": 1.3157894736842106e-05,
      "loss": 0.2927,
      "step": 2900
    },
    {
      "epoch": 76.0,
      "eval_accuracy": 0.08129496402877698,
      "eval_loss": 14.461828231811523,
      "eval_runtime": 6.3066,
      "eval_samples_per_second": 220.403,
      "eval_steps_per_second": 6.977,
      "step": 2907
    },
    {
      "epoch": 76.08,
      "learning_rate": 1.301169590643275e-05,
      "loss": 0.2929,
      "step": 2910
    },
    {
      "epoch": 76.34,
      "learning_rate": 1.2865497076023392e-05,
      "loss": 0.3012,
      "step": 2920
    },
    {
      "epoch": 76.6,
      "learning_rate": 1.2719298245614037e-05,
      "loss": 0.2694,
      "step": 2930
    },
    {
      "epoch": 76.86,
      "learning_rate": 1.2573099415204679e-05,
      "loss": 0.2967,
      "step": 2940
    },
    {
      "epoch": 76.99,
      "eval_accuracy": 0.08201438848920864,
      "eval_loss": 14.358075141906738,
      "eval_runtime": 6.119,
      "eval_samples_per_second": 227.162,
      "eval_steps_per_second": 7.191,
      "step": 2945
    },
    {
      "epoch": 77.12,
      "learning_rate": 1.242690058479532e-05,
      "loss": 0.2748,
      "step": 2950
    },
    {
      "epoch": 77.39,
      "learning_rate": 1.2280701754385964e-05,
      "loss": 0.281,
      "step": 2960
    },
    {
      "epoch": 77.65,
      "learning_rate": 1.2134502923976608e-05,
      "loss": 0.2897,
      "step": 2970
    },
    {
      "epoch": 77.91,
      "learning_rate": 1.1988304093567252e-05,
      "loss": 0.2446,
      "step": 2980
    },
    {
      "epoch": 77.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.456214904785156,
      "eval_runtime": 6.1336,
      "eval_samples_per_second": 226.62,
      "eval_steps_per_second": 7.174,
      "step": 2983
    },
    {
      "epoch": 78.17,
      "learning_rate": 1.1842105263157895e-05,
      "loss": 0.3009,
      "step": 2990
    },
    {
      "epoch": 78.43,
      "learning_rate": 1.1695906432748537e-05,
      "loss": 0.2901,
      "step": 3000
    },
    {
      "epoch": 78.69,
      "learning_rate": 1.154970760233918e-05,
      "loss": 0.2656,
      "step": 3010
    },
    {
      "epoch": 78.95,
      "learning_rate": 1.1403508771929824e-05,
      "loss": 0.3035,
      "step": 3020
    },
    {
      "epoch": 78.98,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.268139839172363,
      "eval_runtime": 6.2555,
      "eval_samples_per_second": 222.203,
      "eval_steps_per_second": 7.034,
      "step": 3021
    },
    {
      "epoch": 79.22,
      "learning_rate": 1.1257309941520468e-05,
      "loss": 0.2897,
      "step": 3030
    },
    {
      "epoch": 79.48,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.3152,
      "step": 3040
    },
    {
      "epoch": 79.74,
      "learning_rate": 1.0964912280701754e-05,
      "loss": 0.3034,
      "step": 3050
    },
    {
      "epoch": 80.0,
      "learning_rate": 1.0818713450292397e-05,
      "loss": 0.2989,
      "step": 3060
    },
    {
      "epoch": 80.0,
      "eval_accuracy": 0.08273381294964029,
      "eval_loss": 14.276779174804688,
      "eval_runtime": 6.208,
      "eval_samples_per_second": 223.906,
      "eval_steps_per_second": 7.088,
      "step": 3060
    },
    {
      "epoch": 80.26,
      "learning_rate": 1.067251461988304e-05,
      "loss": 0.3268,
      "step": 3070
    },
    {
      "epoch": 80.52,
      "learning_rate": 1.0526315789473684e-05,
      "loss": 0.2827,
      "step": 3080
    },
    {
      "epoch": 80.78,
      "learning_rate": 1.0380116959064328e-05,
      "loss": 0.2486,
      "step": 3090
    },
    {
      "epoch": 80.99,
      "eval_accuracy": 0.08201438848920864,
      "eval_loss": 14.424173355102539,
      "eval_runtime": 6.2731,
      "eval_samples_per_second": 221.581,
      "eval_steps_per_second": 7.014,
      "step": 3098
    },
    {
      "epoch": 81.05,
      "learning_rate": 1.023391812865497e-05,
      "loss": 0.2711,
      "step": 3100
    },
    {
      "epoch": 81.31,
      "learning_rate": 1.0087719298245614e-05,
      "loss": 0.2918,
      "step": 3110
    },
    {
      "epoch": 81.57,
      "learning_rate": 9.941520467836257e-06,
      "loss": 0.3038,
      "step": 3120
    },
    {
      "epoch": 81.83,
      "learning_rate": 9.795321637426901e-06,
      "loss": 0.2622,
      "step": 3130
    },
    {
      "epoch": 81.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.381049156188965,
      "eval_runtime": 6.1831,
      "eval_samples_per_second": 224.805,
      "eval_steps_per_second": 7.116,
      "step": 3136
    },
    {
      "epoch": 82.09,
      "learning_rate": 9.649122807017545e-06,
      "loss": 0.295,
      "step": 3140
    },
    {
      "epoch": 82.35,
      "learning_rate": 9.502923976608186e-06,
      "loss": 0.2911,
      "step": 3150
    },
    {
      "epoch": 82.61,
      "learning_rate": 9.35672514619883e-06,
      "loss": 0.2637,
      "step": 3160
    },
    {
      "epoch": 82.88,
      "learning_rate": 9.210526315789474e-06,
      "loss": 0.2892,
      "step": 3170
    },
    {
      "epoch": 82.98,
      "eval_accuracy": 0.08273381294964029,
      "eval_loss": 14.46367073059082,
      "eval_runtime": 5.9176,
      "eval_samples_per_second": 234.894,
      "eval_steps_per_second": 7.435,
      "step": 3174
    },
    {
      "epoch": 83.14,
      "learning_rate": 9.064327485380117e-06,
      "loss": 0.3097,
      "step": 3180
    },
    {
      "epoch": 83.4,
      "learning_rate": 8.918128654970761e-06,
      "loss": 0.2677,
      "step": 3190
    },
    {
      "epoch": 83.66,
      "learning_rate": 8.771929824561403e-06,
      "loss": 0.2416,
      "step": 3200
    },
    {
      "epoch": 83.92,
      "learning_rate": 8.625730994152046e-06,
      "loss": 0.2668,
      "step": 3210
    },
    {
      "epoch": 84.0,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.459735870361328,
      "eval_runtime": 5.9202,
      "eval_samples_per_second": 234.79,
      "eval_steps_per_second": 7.432,
      "step": 3213
    },
    {
      "epoch": 84.18,
      "learning_rate": 8.47953216374269e-06,
      "loss": 0.286,
      "step": 3220
    },
    {
      "epoch": 84.44,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.2743,
      "step": 3230
    },
    {
      "epoch": 84.71,
      "learning_rate": 8.187134502923977e-06,
      "loss": 0.2709,
      "step": 3240
    },
    {
      "epoch": 84.97,
      "learning_rate": 8.04093567251462e-06,
      "loss": 0.2527,
      "step": 3250
    },
    {
      "epoch": 84.99,
      "eval_accuracy": 0.08201438848920864,
      "eval_loss": 14.30977725982666,
      "eval_runtime": 5.9216,
      "eval_samples_per_second": 234.734,
      "eval_steps_per_second": 7.43,
      "step": 3251
    },
    {
      "epoch": 85.23,
      "learning_rate": 7.894736842105263e-06,
      "loss": 0.3011,
      "step": 3260
    },
    {
      "epoch": 85.49,
      "learning_rate": 7.748538011695907e-06,
      "loss": 0.2647,
      "step": 3270
    },
    {
      "epoch": 85.75,
      "learning_rate": 7.602339181286549e-06,
      "loss": 0.2636,
      "step": 3280
    },
    {
      "epoch": 85.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.37409782409668,
      "eval_runtime": 5.948,
      "eval_samples_per_second": 233.693,
      "eval_steps_per_second": 7.397,
      "step": 3289
    },
    {
      "epoch": 86.01,
      "learning_rate": 7.456140350877193e-06,
      "loss": 0.2907,
      "step": 3290
    },
    {
      "epoch": 86.27,
      "learning_rate": 7.3099415204678366e-06,
      "loss": 0.2792,
      "step": 3300
    },
    {
      "epoch": 86.54,
      "learning_rate": 7.163742690058479e-06,
      "loss": 0.3177,
      "step": 3310
    },
    {
      "epoch": 86.8,
      "learning_rate": 7.017543859649123e-06,
      "loss": 0.247,
      "step": 3320
    },
    {
      "epoch": 86.98,
      "eval_accuracy": 0.0841726618705036,
      "eval_loss": 14.536933898925781,
      "eval_runtime": 5.8678,
      "eval_samples_per_second": 236.885,
      "eval_steps_per_second": 7.499,
      "step": 3327
    },
    {
      "epoch": 87.06,
      "learning_rate": 6.871345029239766e-06,
      "loss": 0.2813,
      "step": 3330
    },
    {
      "epoch": 87.32,
      "learning_rate": 6.725146198830409e-06,
      "loss": 0.2539,
      "step": 3340
    },
    {
      "epoch": 87.58,
      "learning_rate": 6.578947368421053e-06,
      "loss": 0.2976,
      "step": 3350
    },
    {
      "epoch": 87.84,
      "learning_rate": 6.432748538011696e-06,
      "loss": 0.2693,
      "step": 3360
    },
    {
      "epoch": 88.0,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.403931617736816,
      "eval_runtime": 5.9739,
      "eval_samples_per_second": 232.68,
      "eval_steps_per_second": 7.365,
      "step": 3366
    },
    {
      "epoch": 88.1,
      "learning_rate": 6.286549707602339e-06,
      "loss": 0.3025,
      "step": 3370
    },
    {
      "epoch": 88.37,
      "learning_rate": 6.140350877192982e-06,
      "loss": 0.2483,
      "step": 3380
    },
    {
      "epoch": 88.63,
      "learning_rate": 5.994152046783626e-06,
      "loss": 0.2826,
      "step": 3390
    },
    {
      "epoch": 88.89,
      "learning_rate": 5.8479532163742686e-06,
      "loss": 0.2692,
      "step": 3400
    },
    {
      "epoch": 88.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.616063117980957,
      "eval_runtime": 5.9896,
      "eval_samples_per_second": 232.068,
      "eval_steps_per_second": 7.346,
      "step": 3404
    },
    {
      "epoch": 89.15,
      "learning_rate": 5.701754385964912e-06,
      "loss": 0.294,
      "step": 3410
    },
    {
      "epoch": 89.41,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.3021,
      "step": 3420
    },
    {
      "epoch": 89.67,
      "learning_rate": 5.409356725146199e-06,
      "loss": 0.2924,
      "step": 3430
    },
    {
      "epoch": 89.93,
      "learning_rate": 5.263157894736842e-06,
      "loss": 0.28,
      "step": 3440
    },
    {
      "epoch": 89.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.524378776550293,
      "eval_runtime": 5.9099,
      "eval_samples_per_second": 235.197,
      "eval_steps_per_second": 7.445,
      "step": 3442
    },
    {
      "epoch": 90.2,
      "learning_rate": 5.116959064327485e-06,
      "loss": 0.3174,
      "step": 3450
    },
    {
      "epoch": 90.46,
      "learning_rate": 4.970760233918129e-06,
      "loss": 0.2556,
      "step": 3460
    },
    {
      "epoch": 90.72,
      "learning_rate": 4.824561403508772e-06,
      "loss": 0.2672,
      "step": 3470
    },
    {
      "epoch": 90.98,
      "learning_rate": 4.678362573099415e-06,
      "loss": 0.2535,
      "step": 3480
    },
    {
      "epoch": 90.98,
      "eval_accuracy": 0.0841726618705036,
      "eval_loss": 14.40617847442627,
      "eval_runtime": 5.9735,
      "eval_samples_per_second": 232.695,
      "eval_steps_per_second": 7.366,
      "step": 3480
    },
    {
      "epoch": 91.24,
      "learning_rate": 4.532163742690059e-06,
      "loss": 0.3035,
      "step": 3490
    },
    {
      "epoch": 91.5,
      "learning_rate": 4.3859649122807014e-06,
      "loss": 0.2757,
      "step": 3500
    },
    {
      "epoch": 91.76,
      "learning_rate": 4.239766081871345e-06,
      "loss": 0.2887,
      "step": 3510
    },
    {
      "epoch": 92.0,
      "eval_accuracy": 0.08057553956834532,
      "eval_loss": 14.41130256652832,
      "eval_runtime": 5.9481,
      "eval_samples_per_second": 233.689,
      "eval_steps_per_second": 7.397,
      "step": 3519
    },
    {
      "epoch": 92.03,
      "learning_rate": 4.093567251461989e-06,
      "loss": 0.2733,
      "step": 3520
    },
    {
      "epoch": 92.29,
      "learning_rate": 3.9473684210526315e-06,
      "loss": 0.2809,
      "step": 3530
    },
    {
      "epoch": 92.55,
      "learning_rate": 3.8011695906432747e-06,
      "loss": 0.2788,
      "step": 3540
    },
    {
      "epoch": 92.81,
      "learning_rate": 3.6549707602339183e-06,
      "loss": 0.257,
      "step": 3550
    },
    {
      "epoch": 92.99,
      "eval_accuracy": 0.0841726618705036,
      "eval_loss": 14.344242095947266,
      "eval_runtime": 5.9071,
      "eval_samples_per_second": 235.31,
      "eval_steps_per_second": 7.449,
      "step": 3557
    },
    {
      "epoch": 93.07,
      "learning_rate": 3.5087719298245615e-06,
      "loss": 0.3108,
      "step": 3560
    },
    {
      "epoch": 93.33,
      "learning_rate": 3.3625730994152047e-06,
      "loss": 0.2408,
      "step": 3570
    },
    {
      "epoch": 93.59,
      "learning_rate": 3.216374269005848e-06,
      "loss": 0.2597,
      "step": 3580
    },
    {
      "epoch": 93.86,
      "learning_rate": 3.070175438596491e-06,
      "loss": 0.2627,
      "step": 3590
    },
    {
      "epoch": 93.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.469304084777832,
      "eval_runtime": 6.0553,
      "eval_samples_per_second": 229.553,
      "eval_steps_per_second": 7.266,
      "step": 3595
    },
    {
      "epoch": 94.12,
      "learning_rate": 2.9239766081871343e-06,
      "loss": 0.2628,
      "step": 3600
    },
    {
      "epoch": 94.38,
      "learning_rate": 2.777777777777778e-06,
      "loss": 0.283,
      "step": 3610
    },
    {
      "epoch": 94.64,
      "learning_rate": 2.631578947368421e-06,
      "loss": 0.27,
      "step": 3620
    },
    {
      "epoch": 94.9,
      "learning_rate": 2.4853801169590643e-06,
      "loss": 0.2804,
      "step": 3630
    },
    {
      "epoch": 94.98,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.322315216064453,
      "eval_runtime": 6.0881,
      "eval_samples_per_second": 228.315,
      "eval_steps_per_second": 7.227,
      "step": 3633
    },
    {
      "epoch": 95.16,
      "learning_rate": 2.3391812865497075e-06,
      "loss": 0.3117,
      "step": 3640
    },
    {
      "epoch": 95.42,
      "learning_rate": 2.1929824561403507e-06,
      "loss": 0.2948,
      "step": 3650
    },
    {
      "epoch": 95.69,
      "learning_rate": 2.0467836257309943e-06,
      "loss": 0.2926,
      "step": 3660
    },
    {
      "epoch": 95.95,
      "learning_rate": 1.9005847953216373e-06,
      "loss": 0.2529,
      "step": 3670
    },
    {
      "epoch": 96.0,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.384366989135742,
      "eval_runtime": 6.1884,
      "eval_samples_per_second": 224.614,
      "eval_steps_per_second": 7.11,
      "step": 3672
    },
    {
      "epoch": 96.21,
      "learning_rate": 1.7543859649122807e-06,
      "loss": 0.2694,
      "step": 3680
    },
    {
      "epoch": 96.47,
      "learning_rate": 1.608187134502924e-06,
      "loss": 0.2869,
      "step": 3690
    },
    {
      "epoch": 96.73,
      "learning_rate": 1.4619883040935671e-06,
      "loss": 0.2634,
      "step": 3700
    },
    {
      "epoch": 96.99,
      "learning_rate": 1.3157894736842106e-06,
      "loss": 0.2327,
      "step": 3710
    },
    {
      "epoch": 96.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.428401947021484,
      "eval_runtime": 6.2696,
      "eval_samples_per_second": 221.706,
      "eval_steps_per_second": 7.018,
      "step": 3710
    },
    {
      "epoch": 97.25,
      "learning_rate": 1.1695906432748538e-06,
      "loss": 0.2786,
      "step": 3720
    },
    {
      "epoch": 97.52,
      "learning_rate": 1.0233918128654972e-06,
      "loss": 0.2893,
      "step": 3730
    },
    {
      "epoch": 97.78,
      "learning_rate": 8.771929824561404e-07,
      "loss": 0.2643,
      "step": 3740
    },
    {
      "epoch": 97.99,
      "eval_accuracy": 0.08345323741007195,
      "eval_loss": 14.556747436523438,
      "eval_runtime": 6.288,
      "eval_samples_per_second": 221.057,
      "eval_steps_per_second": 6.997,
      "step": 3748
    },
    {
      "epoch": 98.04,
      "learning_rate": 7.309941520467836e-07,
      "loss": 0.2726,
      "step": 3750
    },
    {
      "epoch": 98.3,
      "learning_rate": 5.847953216374269e-07,
      "loss": 0.2169,
      "step": 3760
    },
    {
      "epoch": 98.56,
      "learning_rate": 4.385964912280702e-07,
      "loss": 0.2801,
      "step": 3770
    },
    {
      "epoch": 98.82,
      "learning_rate": 2.9239766081871344e-07,
      "loss": 0.284,
      "step": 3780
    },
    {
      "epoch": 98.98,
      "eval_accuracy": 0.08129496402877698,
      "eval_loss": 14.673846244812012,
      "eval_runtime": 6.2773,
      "eval_samples_per_second": 221.432,
      "eval_steps_per_second": 7.009,
      "step": 3786
    },
    {
      "epoch": 99.08,
      "learning_rate": 1.4619883040935672e-07,
      "loss": 0.2767,
      "step": 3790
    },
    {
      "epoch": 99.35,
      "learning_rate": 0.0,
      "loss": 0.2503,
      "step": 3800
    },
    {
      "epoch": 99.35,
      "eval_accuracy": 0.0841726618705036,
      "eval_loss": 14.536262512207031,
      "eval_runtime": 6.1763,
      "eval_samples_per_second": 225.053,
      "eval_steps_per_second": 7.124,
      "step": 3800
    },
    {
      "epoch": 99.35,
      "step": 3800,
      "total_flos": 1.0404624773781246e+19,
      "train_loss": 0.44781778210087825,
      "train_runtime": 6473.2407,
      "train_samples_per_second": 75.217,
      "train_steps_per_second": 0.587
    }
  ],
  "logging_steps": 10,
  "max_steps": 3800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 500,
  "total_flos": 1.0404624773781246e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}