{
  "best_metric": 0.8536585365853658,
  "best_model_checkpoint": "MAE-CT-CPC-Dicotomized-v3/checkpoint-1215",
  "epoch": 49.00775,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.6975,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.6908,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 7.5e-07,
      "loss": 0.6764,
      "step": 30
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.6459,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.25e-06,
      "loss": 0.6431,
      "step": 50
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.5e-06,
      "loss": 0.6808,
      "step": 60
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.75e-06,
      "loss": 0.5509,
      "step": 70
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.5652,
      "step": 80
    },
    {
      "epoch": 0.02,
      "eval_accuracy": 0.6829268292682927,
      "eval_loss": 0.6259790658950806,
      "eval_runtime": 14.8753,
      "eval_samples_per_second": 2.756,
      "eval_steps_per_second": 0.739,
      "step": 81
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.25e-06,
      "loss": 0.4455,
      "step": 90
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.5e-06,
      "loss": 0.7811,
      "step": 100
    },
    {
      "epoch": 1.01,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.5599,
      "step": 110
    },
    {
      "epoch": 1.01,
      "learning_rate": 3e-06,
      "loss": 0.646,
      "step": 120
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.6918,
      "step": 130
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.5e-06,
      "loss": 0.599,
      "step": 140
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.6309,
      "step": 150
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.6109,
      "step": 160
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.6829268292682927,
      "eval_loss": 0.6052387356758118,
      "eval_runtime": 8.1219,
      "eval_samples_per_second": 5.048,
      "eval_steps_per_second": 1.354,
      "step": 162
    },
    {
      "epoch": 2.0,
      "learning_rate": 4.25e-06,
      "loss": 0.5684,
      "step": 170
    },
    {
      "epoch": 2.0,
      "learning_rate": 4.5e-06,
      "loss": 0.6318,
      "step": 180
    },
    {
      "epoch": 2.01,
      "learning_rate": 4.75e-06,
      "loss": 0.5066,
      "step": 190
    },
    {
      "epoch": 2.01,
      "learning_rate": 5e-06,
      "loss": 0.6018,
      "step": 200
    },
    {
      "epoch": 2.01,
      "learning_rate": 5.2500000000000006e-06,
      "loss": 0.6814,
      "step": 210
    },
    {
      "epoch": 2.01,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.5391,
      "step": 220
    },
    {
      "epoch": 2.02,
      "learning_rate": 5.75e-06,
      "loss": 0.544,
      "step": 230
    },
    {
      "epoch": 2.02,
      "learning_rate": 6e-06,
      "loss": 0.7142,
      "step": 240
    },
    {
      "epoch": 2.02,
      "eval_accuracy": 0.6829268292682927,
      "eval_loss": 0.6206896901130676,
      "eval_runtime": 7.8405,
      "eval_samples_per_second": 5.229,
      "eval_steps_per_second": 1.403,
      "step": 243
    },
    {
      "epoch": 3.0,
      "learning_rate": 6.25e-06,
      "loss": 0.632,
      "step": 250
    },
    {
      "epoch": 3.0,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.6202,
      "step": 260
    },
    {
      "epoch": 3.01,
      "learning_rate": 6.750000000000001e-06,
      "loss": 0.5775,
      "step": 270
    },
    {
      "epoch": 3.01,
      "learning_rate": 7e-06,
      "loss": 0.5195,
      "step": 280
    },
    {
      "epoch": 3.01,
      "learning_rate": 7.25e-06,
      "loss": 0.851,
      "step": 290
    },
    {
      "epoch": 3.01,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6404,
      "step": 300
    },
    {
      "epoch": 3.02,
      "learning_rate": 7.75e-06,
      "loss": 0.6425,
      "step": 310
    },
    {
      "epoch": 3.02,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.5037,
      "step": 320
    },
    {
      "epoch": 3.02,
      "eval_accuracy": 0.6829268292682927,
      "eval_loss": 0.6219127178192139,
      "eval_runtime": 7.8713,
      "eval_samples_per_second": 5.209,
      "eval_steps_per_second": 1.397,
      "step": 324
    },
    {
      "epoch": 4.0,
      "learning_rate": 8.25e-06,
      "loss": 0.4113,
      "step": 330
    },
    {
      "epoch": 4.0,
      "learning_rate": 8.5e-06,
      "loss": 0.6047,
      "step": 340
    },
    {
      "epoch": 4.01,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.5416,
      "step": 350
    },
    {
      "epoch": 4.01,
      "learning_rate": 9e-06,
      "loss": 0.845,
      "step": 360
    },
    {
      "epoch": 4.01,
      "learning_rate": 9.250000000000001e-06,
      "loss": 0.9139,
      "step": 370
    },
    {
      "epoch": 4.01,
      "learning_rate": 9.5e-06,
      "loss": 0.6619,
      "step": 380
    },
    {
      "epoch": 4.02,
      "learning_rate": 9.75e-06,
      "loss": 0.6842,
      "step": 390
    },
    {
      "epoch": 4.02,
      "learning_rate": 1e-05,
      "loss": 0.5177,
      "step": 400
    },
    {
      "epoch": 4.02,
      "eval_accuracy": 0.6829268292682927,
      "eval_loss": 0.6455994844436646,
      "eval_runtime": 8.1279,
      "eval_samples_per_second": 5.044,
      "eval_steps_per_second": 1.353,
      "step": 405
    },
    {
      "epoch": 5.0,
      "learning_rate": 9.972222222222224e-06,
      "loss": 0.5696,
      "step": 410
    },
    {
      "epoch": 5.0,
      "learning_rate": 9.944444444444445e-06,
      "loss": 0.5995,
      "step": 420
    },
    {
      "epoch": 5.01,
      "learning_rate": 9.916666666666668e-06,
      "loss": 0.548,
      "step": 430
    },
    {
      "epoch": 5.01,
      "learning_rate": 9.88888888888889e-06,
      "loss": 0.5892,
      "step": 440
    },
    {
      "epoch": 5.01,
      "learning_rate": 9.861111111111112e-06,
      "loss": 0.6163,
      "step": 450
    },
    {
      "epoch": 5.01,
      "learning_rate": 9.833333333333333e-06,
      "loss": 0.623,
      "step": 460
    },
    {
      "epoch": 5.02,
      "learning_rate": 9.805555555555556e-06,
      "loss": 0.5926,
      "step": 470
    },
    {
      "epoch": 5.02,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.5363,
      "step": 480
    },
    {
      "epoch": 5.02,
      "eval_accuracy": 0.6829268292682927,
      "eval_loss": 0.5790859460830688,
      "eval_runtime": 7.8665,
      "eval_samples_per_second": 5.212,
      "eval_steps_per_second": 1.398,
      "step": 486
    },
    {
      "epoch": 6.0,
      "learning_rate": 9.75e-06,
      "loss": 0.5994,
      "step": 490
    },
    {
      "epoch": 6.0,
      "learning_rate": 9.722222222222223e-06,
      "loss": 0.3985,
      "step": 500
    },
    {
      "epoch": 6.01,
      "learning_rate": 9.694444444444446e-06,
      "loss": 0.4727,
      "step": 510
    },
    {
      "epoch": 6.01,
      "learning_rate": 9.666666666666667e-06,
      "loss": 0.4279,
      "step": 520
    },
    {
      "epoch": 6.01,
      "learning_rate": 9.63888888888889e-06,
      "loss": 0.7107,
      "step": 530
    },
    {
      "epoch": 6.01,
      "learning_rate": 9.611111111111112e-06,
      "loss": 0.602,
      "step": 540
    },
    {
      "epoch": 6.02,
      "learning_rate": 9.583333333333335e-06,
      "loss": 0.7659,
      "step": 550
    },
    {
      "epoch": 6.02,
      "learning_rate": 9.555555555555556e-06,
      "loss": 0.6797,
      "step": 560
    },
    {
      "epoch": 6.02,
      "eval_accuracy": 0.6829268292682927,
      "eval_loss": 0.6294647455215454,
      "eval_runtime": 7.9567,
      "eval_samples_per_second": 5.153,
      "eval_steps_per_second": 1.382,
      "step": 567
    },
    {
      "epoch": 7.0,
      "learning_rate": 9.527777777777778e-06,
      "loss": 0.6453,
      "step": 570
    },
    {
      "epoch": 7.0,
      "learning_rate": 9.5e-06,
      "loss": 0.5694,
      "step": 580
    },
    {
      "epoch": 7.01,
      "learning_rate": 9.472222222222223e-06,
      "loss": 0.4864,
      "step": 590
    },
    {
      "epoch": 7.01,
      "learning_rate": 9.444444444444445e-06,
      "loss": 0.6264,
      "step": 600
    },
    {
      "epoch": 7.01,
      "learning_rate": 9.416666666666667e-06,
      "loss": 0.4552,
      "step": 610
    },
    {
      "epoch": 7.01,
      "learning_rate": 9.38888888888889e-06,
      "loss": 0.5971,
      "step": 620
    },
    {
      "epoch": 7.02,
      "learning_rate": 9.361111111111111e-06,
      "loss": 0.4903,
      "step": 630
    },
    {
      "epoch": 7.02,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.557,
      "step": 640
    },
    {
      "epoch": 7.02,
      "eval_accuracy": 0.7073170731707317,
      "eval_loss": 0.5181677937507629,
      "eval_runtime": 8.0768,
      "eval_samples_per_second": 5.076,
      "eval_steps_per_second": 1.362,
      "step": 648
    },
    {
      "epoch": 8.0,
      "learning_rate": 9.305555555555557e-06,
      "loss": 0.7093,
      "step": 650
    },
    {
      "epoch": 8.0,
      "learning_rate": 9.277777777777778e-06,
      "loss": 0.5308,
      "step": 660
    },
    {
      "epoch": 8.01,
      "learning_rate": 9.250000000000001e-06,
      "loss": 0.4379,
      "step": 670
    },
    {
      "epoch": 8.01,
      "learning_rate": 9.222222222222224e-06,
      "loss": 0.3714,
      "step": 680
    },
    {
      "epoch": 8.01,
      "learning_rate": 9.194444444444445e-06,
      "loss": 0.3573,
      "step": 690
    },
    {
      "epoch": 8.01,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.6547,
      "step": 700
    },
    {
      "epoch": 8.02,
      "learning_rate": 9.13888888888889e-06,
      "loss": 0.5763,
      "step": 710
    },
    {
      "epoch": 8.02,
      "learning_rate": 9.111111111111112e-06,
      "loss": 0.5645,
      "step": 720
    },
    {
      "epoch": 8.02,
      "eval_accuracy": 0.7073170731707317,
      "eval_loss": 0.46078598499298096,
      "eval_runtime": 8.1101,
      "eval_samples_per_second": 5.055,
      "eval_steps_per_second": 1.356,
      "step": 729
    },
    {
      "epoch": 9.0,
      "learning_rate": 9.083333333333333e-06,
      "loss": 0.5181,
      "step": 730
    },
    {
      "epoch": 9.0,
      "learning_rate": 9.055555555555556e-06,
      "loss": 0.4622,
      "step": 740
    },
    {
      "epoch": 9.01,
      "learning_rate": 9.027777777777779e-06,
      "loss": 0.4728,
      "step": 750
    },
    {
      "epoch": 9.01,
      "learning_rate": 9e-06,
      "loss": 0.3326,
      "step": 760
    },
    {
      "epoch": 9.01,
      "learning_rate": 8.972222222222223e-06,
      "loss": 0.4734,
      "step": 770
    },
    {
      "epoch": 9.01,
      "learning_rate": 8.944444444444446e-06,
      "loss": 0.5118,
      "step": 780
    },
    {
      "epoch": 9.02,
      "learning_rate": 8.916666666666667e-06,
      "loss": 0.4177,
      "step": 790
    },
    {
      "epoch": 9.02,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.5532,
      "step": 800
    },
    {
      "epoch": 9.02,
      "learning_rate": 8.861111111111111e-06,
      "loss": 0.4484,
      "step": 810
    },
    {
      "epoch": 9.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 0.4720686972141266,
      "eval_runtime": 7.9599,
      "eval_samples_per_second": 5.151,
      "eval_steps_per_second": 1.382,
      "step": 810
    },
    {
      "epoch": 10.0,
      "learning_rate": 8.833333333333334e-06,
      "loss": 0.4518,
      "step": 820
    },
    {
      "epoch": 10.01,
      "learning_rate": 8.805555555555557e-06,
      "loss": 0.5042,
      "step": 830
    },
    {
      "epoch": 10.01,
      "learning_rate": 8.777777777777778e-06,
      "loss": 0.6148,
      "step": 840
    },
    {
      "epoch": 10.01,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.2663,
      "step": 850
    },
    {
      "epoch": 10.01,
      "learning_rate": 8.722222222222224e-06,
      "loss": 0.4605,
      "step": 860
    },
    {
      "epoch": 10.02,
      "learning_rate": 8.694444444444445e-06,
      "loss": 0.5055,
      "step": 870
    },
    {
      "epoch": 10.02,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.3305,
      "step": 880
    },
    {
      "epoch": 10.02,
      "learning_rate": 8.63888888888889e-06,
      "loss": 0.3354,
      "step": 890
    },
    {
      "epoch": 10.02,
      "eval_accuracy": 0.7317073170731707,
      "eval_loss": 0.7011194229125977,
      "eval_runtime": 7.8817,
      "eval_samples_per_second": 5.202,
      "eval_steps_per_second": 1.396,
      "step": 891
    },
    {
      "epoch": 11.0,
      "learning_rate": 8.611111111111112e-06,
      "loss": 0.2105,
      "step": 900
    },
    {
      "epoch": 11.0,
      "learning_rate": 8.583333333333333e-06,
      "loss": 0.2104,
      "step": 910
    },
    {
      "epoch": 11.01,
      "learning_rate": 8.555555555555556e-06,
      "loss": 0.7811,
      "step": 920
    },
    {
      "epoch": 11.01,
      "learning_rate": 8.527777777777779e-06,
      "loss": 0.8752,
      "step": 930
    },
    {
      "epoch": 11.01,
      "learning_rate": 8.5e-06,
      "loss": 0.4221,
      "step": 940
    },
    {
      "epoch": 11.01,
      "learning_rate": 8.472222222222223e-06,
      "loss": 0.5703,
      "step": 950
    },
    {
      "epoch": 11.02,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.5222,
      "step": 960
    },
    {
      "epoch": 11.02,
      "learning_rate": 8.416666666666667e-06,
      "loss": 0.2875,
      "step": 970
    },
    {
      "epoch": 11.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 0.44984355568885803,
      "eval_runtime": 7.8985,
      "eval_samples_per_second": 5.191,
      "eval_steps_per_second": 1.393,
      "step": 972
    },
    {
      "epoch": 12.0,
      "learning_rate": 8.38888888888889e-06,
      "loss": 0.3145,
      "step": 980
    },
    {
      "epoch": 12.0,
      "learning_rate": 8.361111111111113e-06,
      "loss": 0.6478,
      "step": 990
    },
    {
      "epoch": 12.01,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.2942,
      "step": 1000
    },
    {
      "epoch": 12.01,
      "learning_rate": 8.305555555555557e-06,
      "loss": 0.5306,
      "step": 1010
    },
    {
      "epoch": 12.01,
      "learning_rate": 8.277777777777778e-06,
      "loss": 0.4272,
      "step": 1020
    },
    {
      "epoch": 12.01,
      "learning_rate": 8.25e-06,
      "loss": 0.502,
      "step": 1030
    },
    {
      "epoch": 12.02,
      "learning_rate": 8.222222222222222e-06,
      "loss": 0.3006,
      "step": 1040
    },
    {
      "epoch": 12.02,
      "learning_rate": 8.194444444444445e-06,
      "loss": 0.3535,
      "step": 1050
    },
    {
      "epoch": 12.02,
      "eval_accuracy": 0.7560975609756098,
      "eval_loss": 0.8161238431930542,
      "eval_runtime": 7.971,
      "eval_samples_per_second": 5.144,
      "eval_steps_per_second": 1.38,
      "step": 1053
    },
    {
      "epoch": 13.0,
      "learning_rate": 8.166666666666668e-06,
      "loss": 0.7329,
      "step": 1060
    },
    {
      "epoch": 13.0,
      "learning_rate": 8.138888888888889e-06,
      "loss": 0.3882,
      "step": 1070
    },
    {
      "epoch": 13.01,
      "learning_rate": 8.111111111111112e-06,
      "loss": 0.5811,
      "step": 1080
    },
    {
      "epoch": 13.01,
      "learning_rate": 8.083333333333334e-06,
      "loss": 0.1979,
      "step": 1090
    },
    {
      "epoch": 13.01,
      "learning_rate": 8.055555555555557e-06,
      "loss": 0.5972,
      "step": 1100
    },
    {
      "epoch": 13.01,
      "learning_rate": 8.027777777777778e-06,
      "loss": 0.4034,
      "step": 1110
    },
    {
      "epoch": 13.02,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.4967,
      "step": 1120
    },
    {
      "epoch": 13.02,
      "learning_rate": 7.972222222222224e-06,
      "loss": 0.294,
      "step": 1130
    },
    {
      "epoch": 13.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 0.5175291299819946,
      "eval_runtime": 8.0681,
      "eval_samples_per_second": 5.082,
      "eval_steps_per_second": 1.363,
      "step": 1134
    },
    {
      "epoch": 14.0,
      "learning_rate": 7.944444444444445e-06,
      "loss": 0.3438,
      "step": 1140
    },
    {
      "epoch": 14.0,
      "learning_rate": 7.916666666666667e-06,
      "loss": 0.3211,
      "step": 1150
    },
    {
      "epoch": 14.01,
      "learning_rate": 7.88888888888889e-06,
      "loss": 0.0984,
      "step": 1160
    },
    {
      "epoch": 14.01,
      "learning_rate": 7.861111111111112e-06,
      "loss": 0.1332,
      "step": 1170
    },
    {
      "epoch": 14.01,
      "learning_rate": 7.833333333333333e-06,
      "loss": 0.3631,
      "step": 1180
    },
    {
      "epoch": 14.01,
      "learning_rate": 7.805555555555556e-06,
      "loss": 0.5752,
      "step": 1190
    },
    {
      "epoch": 14.02,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.6094,
      "step": 1200
    },
    {
      "epoch": 14.02,
      "learning_rate": 7.75e-06,
      "loss": 0.1907,
      "step": 1210
    },
    {
      "epoch": 14.02,
      "eval_accuracy": 0.8536585365853658,
      "eval_loss": 0.37496745586395264,
      "eval_runtime": 7.964,
      "eval_samples_per_second": 5.148,
      "eval_steps_per_second": 1.381,
      "step": 1215
    },
    {
      "epoch": 15.0,
      "learning_rate": 7.722222222222223e-06,
      "loss": 0.2175,
      "step": 1220
    },
    {
      "epoch": 15.0,
      "learning_rate": 7.694444444444446e-06,
      "loss": 0.4022,
      "step": 1230
    },
    {
      "epoch": 15.01,
      "learning_rate": 7.666666666666667e-06,
      "loss": 0.2657,
      "step": 1240
    },
    {
      "epoch": 15.01,
      "learning_rate": 7.638888888888888e-06,
      "loss": 0.1851,
      "step": 1250
    },
    {
      "epoch": 15.01,
      "learning_rate": 7.611111111111111e-06,
      "loss": 0.4776,
      "step": 1260
    },
    {
      "epoch": 15.01,
      "learning_rate": 7.583333333333333e-06,
      "loss": 0.2882,
      "step": 1270
    },
    {
      "epoch": 15.02,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.1469,
      "step": 1280
    },
    {
      "epoch": 15.02,
      "learning_rate": 7.527777777777778e-06,
      "loss": 0.2533,
      "step": 1290
    },
    {
      "epoch": 15.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 0.639987587928772,
      "eval_runtime": 8.1114,
      "eval_samples_per_second": 5.055,
      "eval_steps_per_second": 1.356,
      "step": 1296
    },
    {
      "epoch": 16.0,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.4324,
      "step": 1300
    },
    {
      "epoch": 16.0,
      "learning_rate": 7.472222222222223e-06,
      "loss": 0.196,
      "step": 1310
    },
    {
      "epoch": 16.01,
      "learning_rate": 7.444444444444445e-06,
      "loss": 0.2845,
      "step": 1320
    },
    {
      "epoch": 16.01,
      "learning_rate": 7.416666666666668e-06,
      "loss": 0.4057,
      "step": 1330
    },
    {
      "epoch": 16.01,
      "learning_rate": 7.38888888888889e-06,
      "loss": 0.4032,
      "step": 1340
    },
    {
      "epoch": 16.01,
      "learning_rate": 7.361111111111112e-06,
      "loss": 0.3042,
      "step": 1350
    },
    {
      "epoch": 16.02,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.0641,
      "step": 1360
    },
    {
      "epoch": 16.02,
      "learning_rate": 7.305555555555556e-06,
      "loss": 0.2499,
      "step": 1370
    },
    {
      "epoch": 16.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 0.9233259558677673,
      "eval_runtime": 8.1114,
      "eval_samples_per_second": 5.055,
      "eval_steps_per_second": 1.356,
      "step": 1377
    },
    {
      "epoch": 17.0,
      "learning_rate": 7.277777777777778e-06,
      "loss": 0.079,
      "step": 1380
    },
    {
      "epoch": 17.0,
      "learning_rate": 7.25e-06,
      "loss": 0.4094,
      "step": 1390
    },
    {
      "epoch": 17.01,
      "learning_rate": 7.222222222222223e-06,
      "loss": 0.5926,
      "step": 1400
    },
    {
      "epoch": 17.01,
      "learning_rate": 7.194444444444445e-06,
      "loss": 0.1281,
      "step": 1410
    },
    {
      "epoch": 17.01,
      "learning_rate": 7.166666666666667e-06,
      "loss": 0.0788,
      "step": 1420
    },
    {
      "epoch": 17.01,
      "learning_rate": 7.13888888888889e-06,
      "loss": 0.4139,
      "step": 1430
    },
    {
      "epoch": 17.02,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.3324,
      "step": 1440
    },
    {
      "epoch": 17.02,
      "learning_rate": 7.083333333333335e-06,
      "loss": 0.2726,
      "step": 1450
    },
    {
      "epoch": 17.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.0686460733413696,
      "eval_runtime": 7.933,
      "eval_samples_per_second": 5.168,
      "eval_steps_per_second": 1.387,
      "step": 1458
    },
    {
      "epoch": 18.0,
      "learning_rate": 7.055555555555557e-06,
      "loss": 0.1204,
      "step": 1460
    },
    {
      "epoch": 18.0,
      "learning_rate": 7.027777777777778e-06,
      "loss": 0.2403,
      "step": 1470
    },
    {
      "epoch": 18.01,
      "learning_rate": 7e-06,
      "loss": 0.204,
      "step": 1480
    },
    {
      "epoch": 18.01,
      "learning_rate": 6.972222222222223e-06,
      "loss": 0.2375,
      "step": 1490
    },
    {
      "epoch": 18.01,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.5472,
      "step": 1500
    },
    {
      "epoch": 18.01,
      "learning_rate": 6.916666666666667e-06,
      "loss": 0.2903,
      "step": 1510
    },
    {
      "epoch": 18.02,
      "learning_rate": 6.88888888888889e-06,
      "loss": 0.2659,
      "step": 1520
    },
    {
      "epoch": 18.02,
      "learning_rate": 6.861111111111112e-06,
      "loss": 0.4447,
      "step": 1530
    },
    {
      "epoch": 18.02,
      "eval_accuracy": 0.7073170731707317,
      "eval_loss": 1.0432931184768677,
      "eval_runtime": 7.9292,
      "eval_samples_per_second": 5.171,
      "eval_steps_per_second": 1.387,
      "step": 1539
    },
    {
      "epoch": 19.0,
      "learning_rate": 6.833333333333334e-06,
      "loss": 0.114,
      "step": 1540
    },
    {
      "epoch": 19.0,
      "learning_rate": 6.8055555555555566e-06,
      "loss": 0.0215,
      "step": 1550
    },
    {
      "epoch": 19.01,
      "learning_rate": 6.777777777777779e-06,
      "loss": 0.1346,
      "step": 1560
    },
    {
      "epoch": 19.01,
      "learning_rate": 6.750000000000001e-06,
      "loss": 0.0808,
      "step": 1570
    },
    {
      "epoch": 19.01,
      "learning_rate": 6.7222222222222235e-06,
      "loss": 0.3402,
      "step": 1580
    },
    {
      "epoch": 19.01,
      "learning_rate": 6.694444444444445e-06,
      "loss": 0.2226,
      "step": 1590
    },
    {
      "epoch": 19.02,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.1922,
      "step": 1600
    },
    {
      "epoch": 19.02,
      "learning_rate": 6.6388888888888895e-06,
      "loss": 0.2581,
      "step": 1610
    },
    {
      "epoch": 19.02,
      "learning_rate": 6.6111111111111115e-06,
      "loss": 0.1554,
      "step": 1620
    },
    {
      "epoch": 19.02,
      "eval_accuracy": 0.8292682926829268,
      "eval_loss": 1.0790584087371826,
      "eval_runtime": 8.3557,
      "eval_samples_per_second": 4.907,
      "eval_steps_per_second": 1.316,
      "step": 1620
    },
    {
      "epoch": 20.0,
      "learning_rate": 6.5833333333333335e-06,
      "loss": 0.5207,
      "step": 1630
    },
    {
      "epoch": 20.0,
      "learning_rate": 6.555555555555556e-06,
      "loss": 0.2781,
      "step": 1640
    },
    {
      "epoch": 20.01,
      "learning_rate": 6.5277777777777784e-06,
      "loss": 0.2027,
      "step": 1650
    },
    {
      "epoch": 20.01,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.124,
      "step": 1660
    },
    {
      "epoch": 20.01,
      "learning_rate": 6.472222222222223e-06,
      "loss": 0.0777,
      "step": 1670
    },
    {
      "epoch": 20.02,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.1841,
      "step": 1680
    },
    {
      "epoch": 20.02,
      "learning_rate": 6.416666666666667e-06,
      "loss": 0.5539,
      "step": 1690
    },
    {
      "epoch": 20.02,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 0.0163,
      "step": 1700
    },
    {
      "epoch": 20.02,
      "eval_accuracy": 0.7317073170731707,
      "eval_loss": 1.0938783884048462,
      "eval_runtime": 8.1544,
      "eval_samples_per_second": 5.028,
      "eval_steps_per_second": 1.349,
      "step": 1701
    },
    {
      "epoch": 21.0,
      "learning_rate": 6.361111111111111e-06,
      "loss": 0.0009,
      "step": 1710
    },
    {
      "epoch": 21.0,
      "learning_rate": 6.333333333333333e-06,
      "loss": 0.0007,
      "step": 1720
    },
    {
      "epoch": 21.01,
      "learning_rate": 6.305555555555556e-06,
      "loss": 0.2324,
      "step": 1730
    },
    {
      "epoch": 21.01,
      "learning_rate": 6.277777777777778e-06,
      "loss": 0.0022,
      "step": 1740
    },
    {
      "epoch": 21.01,
      "learning_rate": 6.25e-06,
      "loss": 0.072,
      "step": 1750
    },
    {
      "epoch": 21.01,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.1857,
      "step": 1760
    },
    {
      "epoch": 21.02,
      "learning_rate": 6.194444444444445e-06,
      "loss": 0.2117,
      "step": 1770
    },
    {
      "epoch": 21.02,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.0507,
      "step": 1780
    },
    {
      "epoch": 21.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.16009521484375,
      "eval_runtime": 7.9084,
      "eval_samples_per_second": 5.184,
      "eval_steps_per_second": 1.391,
      "step": 1782
    },
    {
      "epoch": 22.0,
      "learning_rate": 6.13888888888889e-06,
      "loss": 0.0162,
      "step": 1790
    },
    {
      "epoch": 22.0,
      "learning_rate": 6.111111111111112e-06,
      "loss": 0.2656,
      "step": 1800
    },
    {
      "epoch": 22.01,
      "learning_rate": 6.083333333333333e-06,
      "loss": 0.0123,
      "step": 1810
    },
    {
      "epoch": 22.01,
      "learning_rate": 6.055555555555555e-06,
      "loss": 0.0139,
      "step": 1820
    },
    {
      "epoch": 22.01,
      "learning_rate": 6.027777777777778e-06,
      "loss": 0.1534,
      "step": 1830
    },
    {
      "epoch": 22.01,
      "learning_rate": 6e-06,
      "loss": 0.2343,
      "step": 1840
    },
    {
      "epoch": 22.02,
      "learning_rate": 5.972222222222222e-06,
      "loss": 0.0018,
      "step": 1850
    },
    {
      "epoch": 22.02,
      "learning_rate": 5.944444444444445e-06,
      "loss": 0.0116,
      "step": 1860
    },
    {
      "epoch": 22.02,
      "eval_accuracy": 0.7073170731707317,
      "eval_loss": 1.3357629776000977,
      "eval_runtime": 8.2637,
      "eval_samples_per_second": 4.961,
      "eval_steps_per_second": 1.331,
      "step": 1863
    },
    {
      "epoch": 23.0,
      "learning_rate": 5.916666666666667e-06,
      "loss": 0.0023,
      "step": 1870
    },
    {
      "epoch": 23.0,
      "learning_rate": 5.88888888888889e-06,
      "loss": 0.0814,
      "step": 1880
    },
    {
      "epoch": 23.01,
      "learning_rate": 5.861111111111112e-06,
      "loss": 0.4414,
      "step": 1890
    },
    {
      "epoch": 23.01,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.0482,
      "step": 1900
    },
    {
      "epoch": 23.01,
      "learning_rate": 5.805555555555557e-06,
      "loss": 0.0205,
      "step": 1910
    },
    {
      "epoch": 23.01,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.0087,
      "step": 1920
    },
    {
      "epoch": 23.02,
      "learning_rate": 5.75e-06,
      "loss": 0.4235,
      "step": 1930
    },
    {
      "epoch": 23.02,
      "learning_rate": 5.722222222222222e-06,
      "loss": 0.015,
      "step": 1940
    },
    {
      "epoch": 23.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.0973219871520996,
      "eval_runtime": 8.026,
      "eval_samples_per_second": 5.108,
      "eval_steps_per_second": 1.371,
      "step": 1944
    },
    {
      "epoch": 24.0,
      "learning_rate": 5.694444444444445e-06,
      "loss": 0.0031,
      "step": 1950
    },
    {
      "epoch": 24.0,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.0475,
      "step": 1960
    },
    {
      "epoch": 24.01,
      "learning_rate": 5.638888888888889e-06,
      "loss": 0.1224,
      "step": 1970
    },
    {
      "epoch": 24.01,
      "learning_rate": 5.611111111111112e-06,
      "loss": 0.0024,
      "step": 1980
    },
    {
      "epoch": 24.01,
      "learning_rate": 5.583333333333334e-06,
      "loss": 0.0033,
      "step": 1990
    },
    {
      "epoch": 24.01,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.2697,
      "step": 2000
    },
    {
      "epoch": 24.02,
      "learning_rate": 5.527777777777779e-06,
      "loss": 0.239,
      "step": 2010
    },
    {
      "epoch": 24.02,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.1993,
      "step": 2020
    },
    {
      "epoch": 24.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.2984837293624878,
      "eval_runtime": 8.0935,
      "eval_samples_per_second": 5.066,
      "eval_steps_per_second": 1.359,
      "step": 2025
    },
    {
      "epoch": 25.0,
      "learning_rate": 5.4722222222222236e-06,
      "loss": 0.1593,
      "step": 2030
    },
    {
      "epoch": 25.0,
      "learning_rate": 5.444444444444445e-06,
      "loss": 0.0006,
      "step": 2040
    },
    {
      "epoch": 25.01,
      "learning_rate": 5.416666666666667e-06,
      "loss": 0.1402,
      "step": 2050
    },
    {
      "epoch": 25.01,
      "learning_rate": 5.388888888888889e-06,
      "loss": 0.0163,
      "step": 2060
    },
    {
      "epoch": 25.01,
      "learning_rate": 5.361111111111112e-06,
      "loss": 0.0498,
      "step": 2070
    },
    {
      "epoch": 25.01,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.0003,
      "step": 2080
    },
    {
      "epoch": 25.02,
      "learning_rate": 5.305555555555556e-06,
      "loss": 0.0002,
      "step": 2090
    },
    {
      "epoch": 25.02,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 0.1496,
      "step": 2100
    },
    {
      "epoch": 25.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.1559369564056396,
      "eval_runtime": 8.2025,
      "eval_samples_per_second": 4.998,
      "eval_steps_per_second": 1.341,
      "step": 2106
    },
    {
      "epoch": 26.0,
      "learning_rate": 5.2500000000000006e-06,
      "loss": 0.011,
      "step": 2110
    },
    {
      "epoch": 26.0,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 0.0007,
      "step": 2120
    },
    {
      "epoch": 26.01,
      "learning_rate": 5.1944444444444454e-06,
      "loss": 0.0007,
      "step": 2130
    },
    {
      "epoch": 26.01,
      "learning_rate": 5.1666666666666675e-06,
      "loss": 0.3386,
      "step": 2140
    },
    {
      "epoch": 26.01,
      "learning_rate": 5.138888888888889e-06,
      "loss": 0.0003,
      "step": 2150
    },
    {
      "epoch": 26.01,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.0032,
      "step": 2160
    },
    {
      "epoch": 26.02,
      "learning_rate": 5.0833333333333335e-06,
      "loss": 0.0002,
      "step": 2170
    },
    {
      "epoch": 26.02,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 0.2059,
      "step": 2180
    },
    {
      "epoch": 26.02,
      "eval_accuracy": 0.7560975609756098,
      "eval_loss": 1.5124356746673584,
      "eval_runtime": 8.1365,
      "eval_samples_per_second": 5.039,
      "eval_steps_per_second": 1.352,
      "step": 2187
    },
    {
      "epoch": 27.0,
      "learning_rate": 5.027777777777778e-06,
      "loss": 0.0062,
      "step": 2190
    },
    {
      "epoch": 27.0,
      "learning_rate": 5e-06,
      "loss": 0.0048,
      "step": 2200
    },
    {
      "epoch": 27.01,
      "learning_rate": 4.9722222222222224e-06,
      "loss": 0.0003,
      "step": 2210
    },
    {
      "epoch": 27.01,
      "learning_rate": 4.944444444444445e-06,
      "loss": 0.2356,
      "step": 2220
    },
    {
      "epoch": 27.01,
      "learning_rate": 4.9166666666666665e-06,
      "loss": 0.2408,
      "step": 2230
    },
    {
      "epoch": 27.01,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.0293,
      "step": 2240
    },
    {
      "epoch": 27.02,
      "learning_rate": 4.861111111111111e-06,
      "loss": 0.0004,
      "step": 2250
    },
    {
      "epoch": 27.02,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.0176,
      "step": 2260
    },
    {
      "epoch": 27.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.4690953493118286,
      "eval_runtime": 8.1532,
      "eval_samples_per_second": 5.029,
      "eval_steps_per_second": 1.349,
      "step": 2268
    },
    {
      "epoch": 28.0,
      "learning_rate": 4.805555555555556e-06,
      "loss": 0.0029,
      "step": 2270
    },
    {
      "epoch": 28.0,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.002,
      "step": 2280
    },
    {
      "epoch": 28.01,
      "learning_rate": 4.75e-06,
      "loss": 0.1602,
      "step": 2290
    },
    {
      "epoch": 28.01,
      "learning_rate": 4.722222222222222e-06,
      "loss": 0.1316,
      "step": 2300
    },
    {
      "epoch": 28.01,
      "learning_rate": 4.694444444444445e-06,
      "loss": 0.0004,
      "step": 2310
    },
    {
      "epoch": 28.01,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.0027,
      "step": 2320
    },
    {
      "epoch": 28.02,
      "learning_rate": 4.638888888888889e-06,
      "loss": 0.2028,
      "step": 2330
    },
    {
      "epoch": 28.02,
      "learning_rate": 4.611111111111112e-06,
      "loss": 0.2477,
      "step": 2340
    },
    {
      "epoch": 28.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5908167362213135,
      "eval_runtime": 8.2176,
      "eval_samples_per_second": 4.989,
      "eval_steps_per_second": 1.339,
      "step": 2349
    },
    {
      "epoch": 29.0,
      "learning_rate": 4.583333333333333e-06,
      "loss": 0.0002,
      "step": 2350
    },
    {
      "epoch": 29.0,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.002,
      "step": 2360
    },
    {
      "epoch": 29.01,
      "learning_rate": 4.527777777777778e-06,
      "loss": 0.0877,
      "step": 2370
    },
    {
      "epoch": 29.01,
      "learning_rate": 4.5e-06,
      "loss": 0.006,
      "step": 2380
    },
    {
      "epoch": 29.01,
      "learning_rate": 4.472222222222223e-06,
      "loss": 0.0209,
      "step": 2390
    },
    {
      "epoch": 29.01,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.0024,
      "step": 2400
    },
    {
      "epoch": 29.02,
      "learning_rate": 4.416666666666667e-06,
      "loss": 0.3257,
      "step": 2410
    },
    {
      "epoch": 29.02,
      "learning_rate": 4.388888888888889e-06,
      "loss": 0.1004,
      "step": 2420
    },
    {
      "epoch": 29.02,
      "learning_rate": 4.361111111111112e-06,
      "loss": 0.0007,
      "step": 2430
    },
    {
      "epoch": 29.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.3928593397140503,
      "eval_runtime": 8.0733,
      "eval_samples_per_second": 5.078,
      "eval_steps_per_second": 1.363,
      "step": 2430
    },
    {
      "epoch": 30.0,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.0003,
      "step": 2440
    },
    {
      "epoch": 30.0,
      "learning_rate": 4.305555555555556e-06,
      "loss": 0.0473,
      "step": 2450
    },
    {
      "epoch": 30.01,
      "learning_rate": 4.277777777777778e-06,
      "loss": 0.0106,
      "step": 2460
    },
    {
      "epoch": 30.01,
      "learning_rate": 4.25e-06,
      "loss": 0.0002,
      "step": 2470
    },
    {
      "epoch": 30.01,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.0032,
      "step": 2480
    },
    {
      "epoch": 30.02,
      "learning_rate": 4.194444444444445e-06,
      "loss": 0.0023,
      "step": 2490
    },
    {
      "epoch": 30.02,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.0005,
      "step": 2500
    },
    {
      "epoch": 30.02,
      "learning_rate": 4.138888888888889e-06,
      "loss": 0.0003,
      "step": 2510
    },
    {
      "epoch": 30.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.3020046949386597,
      "eval_runtime": 8.1411,
      "eval_samples_per_second": 5.036,
      "eval_steps_per_second": 1.351,
      "step": 2511
    },
    {
      "epoch": 31.0,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.0215,
      "step": 2520
    },
    {
      "epoch": 31.0,
      "learning_rate": 4.083333333333334e-06,
      "loss": 0.0021,
      "step": 2530
    },
    {
      "epoch": 31.01,
      "learning_rate": 4.055555555555556e-06,
      "loss": 0.102,
      "step": 2540
    },
    {
      "epoch": 31.01,
      "learning_rate": 4.027777777777779e-06,
      "loss": 0.0548,
      "step": 2550
    },
    {
      "epoch": 31.01,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0004,
      "step": 2560
    },
    {
      "epoch": 31.01,
      "learning_rate": 3.972222222222223e-06,
      "loss": 0.0002,
      "step": 2570
    },
    {
      "epoch": 31.02,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.0002,
      "step": 2580
    },
    {
      "epoch": 31.02,
      "learning_rate": 3.916666666666667e-06,
      "loss": 0.0002,
      "step": 2590
    },
    {
      "epoch": 31.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5539048910140991,
      "eval_runtime": 7.8769,
      "eval_samples_per_second": 5.205,
      "eval_steps_per_second": 1.396,
      "step": 2592
    },
    {
      "epoch": 32.0,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.0003,
      "step": 2600
    },
    {
      "epoch": 32.0,
      "learning_rate": 3.861111111111112e-06,
      "loss": 0.0002,
      "step": 2610
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.0019,
      "step": 2620
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.8055555555555556e-06,
      "loss": 0.0002,
      "step": 2630
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.0001,
      "step": 2640
    },
    {
      "epoch": 32.01,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0001,
      "step": 2650
    },
    {
      "epoch": 32.02,
      "learning_rate": 3.7222222222222225e-06,
      "loss": 0.0002,
      "step": 2660
    },
    {
      "epoch": 32.02,
      "learning_rate": 3.694444444444445e-06,
      "loss": 0.0001,
      "step": 2670
    },
    {
      "epoch": 32.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.5173004865646362,
      "eval_runtime": 7.8525,
      "eval_samples_per_second": 5.221,
      "eval_steps_per_second": 1.401,
      "step": 2673
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.0011,
      "step": 2680
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.638888888888889e-06,
      "loss": 0.0001,
      "step": 2690
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 0.0022,
      "step": 2700
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.5833333333333335e-06,
      "loss": 0.0799,
      "step": 2710
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.0871,
      "step": 2720
    },
    {
      "epoch": 33.01,
      "learning_rate": 3.5277777777777784e-06,
      "loss": 0.0001,
      "step": 2730
    },
    {
      "epoch": 33.02,
      "learning_rate": 3.5e-06,
      "loss": 0.2127,
      "step": 2740
    },
    {
      "epoch": 33.02,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 0.0001,
      "step": 2750
    },
    {
      "epoch": 33.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.6113324165344238,
      "eval_runtime": 8.0153,
      "eval_samples_per_second": 5.115,
      "eval_steps_per_second": 1.372,
      "step": 2754
    },
    {
      "epoch": 34.0,
      "learning_rate": 3.444444444444445e-06,
      "loss": 0.0003,
      "step": 2760
    },
    {
      "epoch": 34.0,
      "learning_rate": 3.416666666666667e-06,
      "loss": 0.0014,
      "step": 2770
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.0004,
      "step": 2780
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.3611111111111117e-06,
      "loss": 0.0003,
      "step": 2790
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.0007,
      "step": 2800
    },
    {
      "epoch": 34.01,
      "learning_rate": 3.3055555555555558e-06,
      "loss": 0.0003,
      "step": 2810
    },
    {
      "epoch": 34.02,
      "learning_rate": 3.277777777777778e-06,
      "loss": 0.0017,
      "step": 2820
    },
    {
      "epoch": 34.02,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0918,
      "step": 2830
    },
    {
      "epoch": 34.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5837070941925049,
      "eval_runtime": 7.8617,
      "eval_samples_per_second": 5.215,
      "eval_steps_per_second": 1.399,
      "step": 2835
    },
    {
      "epoch": 35.0,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 0.0716,
      "step": 2840
    },
    {
      "epoch": 35.0,
      "learning_rate": 3.1944444444444443e-06,
      "loss": 0.1896,
      "step": 2850
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.1666666666666667e-06,
      "loss": 0.0375,
      "step": 2860
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.138888888888889e-06,
      "loss": 0.0005,
      "step": 2870
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.0412,
      "step": 2880
    },
    {
      "epoch": 35.01,
      "learning_rate": 3.0833333333333336e-06,
      "loss": 0.0022,
      "step": 2890
    },
    {
      "epoch": 35.02,
      "learning_rate": 3.055555555555556e-06,
      "loss": 0.0002,
      "step": 2900
    },
    {
      "epoch": 35.02,
      "learning_rate": 3.0277777777777776e-06,
      "loss": 0.0001,
      "step": 2910
    },
    {
      "epoch": 35.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.492816686630249,
      "eval_runtime": 7.885,
      "eval_samples_per_second": 5.2,
      "eval_steps_per_second": 1.395,
      "step": 2916
    },
    {
      "epoch": 36.0,
      "learning_rate": 3e-06,
      "loss": 0.0001,
      "step": 2920
    },
    {
      "epoch": 36.0,
      "learning_rate": 2.9722222222222225e-06,
      "loss": 0.0001,
      "step": 2930
    },
    {
      "epoch": 36.01,
      "learning_rate": 2.944444444444445e-06,
      "loss": 0.0001,
      "step": 2940
    },
    {
      "epoch": 36.01,
      "learning_rate": 2.916666666666667e-06,
      "loss": 0.0002,
      "step": 2950
    },
    {
      "epoch": 36.01,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.0001,
      "step": 2960
    },
    {
      "epoch": 36.01,
      "learning_rate": 2.861111111111111e-06,
      "loss": 0.0002,
      "step": 2970
    },
    {
      "epoch": 36.02,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.008,
      "step": 2980
    },
    {
      "epoch": 36.02,
      "learning_rate": 2.805555555555556e-06,
      "loss": 0.0001,
      "step": 2990
    },
    {
      "epoch": 36.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.413764238357544,
      "eval_runtime": 7.9092,
      "eval_samples_per_second": 5.184,
      "eval_steps_per_second": 1.391,
      "step": 2997
    },
    {
      "epoch": 37.0,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 0.0001,
      "step": 3000
    },
    {
      "epoch": 37.0,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.0001,
      "step": 3010
    },
    {
      "epoch": 37.01,
      "learning_rate": 2.7222222222222224e-06,
      "loss": 0.0001,
      "step": 3020
    },
    {
      "epoch": 37.01,
      "learning_rate": 2.6944444444444444e-06,
      "loss": 0.2049,
      "step": 3030
    },
    {
      "epoch": 37.01,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.0158,
      "step": 3040
    },
    {
      "epoch": 37.01,
      "learning_rate": 2.6388888888888893e-06,
      "loss": 0.0001,
      "step": 3050
    },
    {
      "epoch": 37.02,
      "learning_rate": 2.6111111111111113e-06,
      "loss": 0.0058,
      "step": 3060
    },
    {
      "epoch": 37.02,
      "learning_rate": 2.5833333333333337e-06,
      "loss": 0.2197,
      "step": 3070
    },
    {
      "epoch": 37.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.3980056047439575,
      "eval_runtime": 7.7795,
      "eval_samples_per_second": 5.27,
      "eval_steps_per_second": 1.414,
      "step": 3078
    },
    {
      "epoch": 38.0,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 0.019,
      "step": 3080
    },
    {
      "epoch": 38.0,
      "learning_rate": 2.5277777777777778e-06,
      "loss": 0.0001,
      "step": 3090
    },
    {
      "epoch": 38.01,
      "learning_rate": 2.5e-06,
      "loss": 0.0001,
      "step": 3100
    },
    {
      "epoch": 38.01,
      "learning_rate": 2.4722222222222226e-06,
      "loss": 0.0001,
      "step": 3110
    },
    {
      "epoch": 38.01,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.1238,
      "step": 3120
    },
    {
      "epoch": 38.01,
      "learning_rate": 2.4166666666666667e-06,
      "loss": 0.0008,
      "step": 3130
    },
    {
      "epoch": 38.02,
      "learning_rate": 2.388888888888889e-06,
      "loss": 0.1425,
      "step": 3140
    },
    {
      "epoch": 38.02,
      "learning_rate": 2.361111111111111e-06,
      "loss": 0.0001,
      "step": 3150
    },
    {
      "epoch": 38.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.6004482507705688,
      "eval_runtime": 8.0544,
      "eval_samples_per_second": 5.09,
      "eval_steps_per_second": 1.366,
      "step": 3159
    },
    {
      "epoch": 39.0,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.0057,
      "step": 3160
    },
    {
      "epoch": 39.0,
      "learning_rate": 2.305555555555556e-06,
      "loss": 0.0001,
      "step": 3170
    },
    {
      "epoch": 39.01,
      "learning_rate": 2.277777777777778e-06,
      "loss": 0.0146,
      "step": 3180
    },
    {
      "epoch": 39.01,
      "learning_rate": 2.25e-06,
      "loss": 0.1196,
      "step": 3190
    },
    {
      "epoch": 39.01,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.0003,
      "step": 3200
    },
    {
      "epoch": 39.01,
      "learning_rate": 2.1944444444444445e-06,
      "loss": 0.0001,
      "step": 3210
    },
    {
      "epoch": 39.02,
      "learning_rate": 2.166666666666667e-06,
      "loss": 0.0001,
      "step": 3220
    },
    {
      "epoch": 39.02,
      "learning_rate": 2.138888888888889e-06,
      "loss": 0.0001,
      "step": 3230
    },
    {
      "epoch": 39.02,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 0.0002,
      "step": 3240
    },
    {
      "epoch": 39.02,
      "eval_accuracy": 0.8048780487804879,
      "eval_loss": 1.5589700937271118,
      "eval_runtime": 8.1123,
      "eval_samples_per_second": 5.054,
      "eval_steps_per_second": 1.356,
      "step": 3240
    },
    {
      "epoch": 40.0,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 0.0003,
      "step": 3250
    },
    {
      "epoch": 40.01,
      "learning_rate": 2.0555555555555555e-06,
      "loss": 0.066,
      "step": 3260
    },
    {
      "epoch": 40.01,
      "learning_rate": 2.027777777777778e-06,
      "loss": 0.0001,
      "step": 3270
    },
    {
      "epoch": 40.01,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.023,
      "step": 3280
    },
    {
      "epoch": 40.01,
      "learning_rate": 1.9722222222222224e-06,
      "loss": 0.0003,
      "step": 3290
    },
    {
      "epoch": 40.02,
      "learning_rate": 1.944444444444445e-06,
      "loss": 0.0002,
      "step": 3300
    },
    {
      "epoch": 40.02,
      "learning_rate": 1.916666666666667e-06,
      "loss": 0.0001,
      "step": 3310
    },
    {
      "epoch": 40.02,
      "learning_rate": 1.888888888888889e-06,
      "loss": 0.0246,
      "step": 3320
    },
    {
      "epoch": 40.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5889155864715576,
      "eval_runtime": 8.7954,
      "eval_samples_per_second": 4.662,
      "eval_steps_per_second": 1.251,
      "step": 3321
    },
    {
      "epoch": 41.0,
      "learning_rate": 1.8611111111111113e-06,
      "loss": 0.0001,
      "step": 3330
    },
    {
      "epoch": 41.0,
      "learning_rate": 1.8333333333333333e-06,
      "loss": 0.0001,
      "step": 3340
    },
    {
      "epoch": 41.01,
      "learning_rate": 1.8055555555555557e-06,
      "loss": 0.0001,
      "step": 3350
    },
    {
      "epoch": 41.01,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.0862,
      "step": 3360
    },
    {
      "epoch": 41.01,
      "learning_rate": 1.75e-06,
      "loss": 0.0001,
      "step": 3370
    },
    {
      "epoch": 41.01,
      "learning_rate": 1.7222222222222224e-06,
      "loss": 0.0001,
      "step": 3380
    },
    {
      "epoch": 41.02,
      "learning_rate": 1.6944444444444446e-06,
      "loss": 0.0001,
      "step": 3390
    },
    {
      "epoch": 41.02,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.0002,
      "step": 3400
    },
    {
      "epoch": 41.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.6164361238479614,
      "eval_runtime": 9.5435,
      "eval_samples_per_second": 4.296,
      "eval_steps_per_second": 1.153,
      "step": 3402
    },
    {
      "epoch": 42.0,
      "learning_rate": 1.638888888888889e-06,
      "loss": 0.0001,
      "step": 3410
    },
    {
      "epoch": 42.0,
      "learning_rate": 1.6111111111111113e-06,
      "loss": 0.0002,
      "step": 3420
    },
    {
      "epoch": 42.01,
      "learning_rate": 1.5833333333333333e-06,
      "loss": 0.0007,
      "step": 3430
    },
    {
      "epoch": 42.01,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.0093,
      "step": 3440
    },
    {
      "epoch": 42.01,
      "learning_rate": 1.527777777777778e-06,
      "loss": 0.0001,
      "step": 3450
    },
    {
      "epoch": 42.01,
      "learning_rate": 1.5e-06,
      "loss": 0.0001,
      "step": 3460
    },
    {
      "epoch": 42.02,
      "learning_rate": 1.4722222222222225e-06,
      "loss": 0.0029,
      "step": 3470
    },
    {
      "epoch": 42.02,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 0.0001,
      "step": 3480
    },
    {
      "epoch": 42.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.6071969270706177,
      "eval_runtime": 9.3129,
      "eval_samples_per_second": 4.402,
      "eval_steps_per_second": 1.181,
      "step": 3483
    },
    {
      "epoch": 43.0,
      "learning_rate": 1.4166666666666667e-06,
      "loss": 0.0001,
      "step": 3490
    },
    {
      "epoch": 43.0,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 0.0001,
      "step": 3500
    },
    {
      "epoch": 43.01,
      "learning_rate": 1.3611111111111112e-06,
      "loss": 0.0001,
      "step": 3510
    },
    {
      "epoch": 43.01,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.0001,
      "step": 3520
    },
    {
      "epoch": 43.01,
      "learning_rate": 1.3055555555555556e-06,
      "loss": 0.0183,
      "step": 3530
    },
    {
      "epoch": 43.01,
      "learning_rate": 1.2777777777777779e-06,
      "loss": 0.0001,
      "step": 3540
    },
    {
      "epoch": 43.02,
      "learning_rate": 1.25e-06,
      "loss": 0.0003,
      "step": 3550
    },
    {
      "epoch": 43.02,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 0.0001,
      "step": 3560
    },
    {
      "epoch": 43.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.6258982419967651,
      "eval_runtime": 9.617,
      "eval_samples_per_second": 4.263,
      "eval_steps_per_second": 1.144,
      "step": 3564
    },
    {
      "epoch": 44.0,
      "learning_rate": 1.1944444444444446e-06,
      "loss": 0.0001,
      "step": 3570
    },
    {
      "epoch": 44.0,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 0.0001,
      "step": 3580
    },
    {
      "epoch": 44.01,
      "learning_rate": 1.138888888888889e-06,
      "loss": 0.0001,
      "step": 3590
    },
    {
      "epoch": 44.01,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.0001,
      "step": 3600
    },
    {
      "epoch": 44.01,
      "learning_rate": 1.0833333333333335e-06,
      "loss": 0.0001,
      "step": 3610
    },
    {
      "epoch": 44.01,
      "learning_rate": 1.0555555555555557e-06,
      "loss": 0.0001,
      "step": 3620
    },
    {
      "epoch": 44.02,
      "learning_rate": 1.0277777777777777e-06,
      "loss": 0.0001,
      "step": 3630
    },
    {
      "epoch": 44.02,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.0001,
      "step": 3640
    },
    {
      "epoch": 44.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5989917516708374,
      "eval_runtime": 9.0925,
      "eval_samples_per_second": 4.509,
      "eval_steps_per_second": 1.21,
      "step": 3645
    },
    {
      "epoch": 45.0,
      "learning_rate": 9.722222222222224e-07,
      "loss": 0.0001,
      "step": 3650
    },
    {
      "epoch": 45.0,
      "learning_rate": 9.444444444444445e-07,
      "loss": 0.0001,
      "step": 3660
    },
    {
      "epoch": 45.01,
      "learning_rate": 9.166666666666666e-07,
      "loss": 0.0001,
      "step": 3670
    },
    {
      "epoch": 45.01,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.0001,
      "step": 3680
    },
    {
      "epoch": 45.01,
      "learning_rate": 8.611111111111112e-07,
      "loss": 0.0001,
      "step": 3690
    },
    {
      "epoch": 45.01,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.0001,
      "step": 3700
    },
    {
      "epoch": 45.02,
      "learning_rate": 8.055555555555557e-07,
      "loss": 0.0011,
      "step": 3710
    },
    {
      "epoch": 45.02,
      "learning_rate": 7.777777777777779e-07,
      "loss": 0.0002,
      "step": 3720
    },
    {
      "epoch": 45.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5516998767852783,
      "eval_runtime": 9.3085,
      "eval_samples_per_second": 4.405,
      "eval_steps_per_second": 1.182,
      "step": 3726
    },
    {
      "epoch": 46.0,
      "learning_rate": 7.5e-07,
      "loss": 0.0002,
      "step": 3730
    },
    {
      "epoch": 46.0,
      "learning_rate": 7.222222222222222e-07,
      "loss": 0.0001,
      "step": 3740
    },
    {
      "epoch": 46.01,
      "learning_rate": 6.944444444444446e-07,
      "loss": 0.0001,
      "step": 3750
    },
    {
      "epoch": 46.01,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.0001,
      "step": 3760
    },
    {
      "epoch": 46.01,
      "learning_rate": 6.388888888888889e-07,
      "loss": 0.0001,
      "step": 3770
    },
    {
      "epoch": 46.01,
      "learning_rate": 6.111111111111112e-07,
      "loss": 0.0004,
      "step": 3780
    },
    {
      "epoch": 46.02,
      "learning_rate": 5.833333333333334e-07,
      "loss": 0.0002,
      "step": 3790
    },
    {
      "epoch": 46.02,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.0,
      "step": 3800
    },
    {
      "epoch": 46.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5259727239608765,
      "eval_runtime": 9.3182,
      "eval_samples_per_second": 4.4,
      "eval_steps_per_second": 1.18,
      "step": 3807
    },
    {
      "epoch": 47.0,
      "learning_rate": 5.277777777777779e-07,
      "loss": 0.0001,
      "step": 3810
    },
    {
      "epoch": 47.0,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.0001,
      "step": 3820
    },
    {
      "epoch": 47.01,
      "learning_rate": 4.7222222222222226e-07,
      "loss": 0.0001,
      "step": 3830
    },
    {
      "epoch": 47.01,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.0001,
      "step": 3840
    },
    {
      "epoch": 47.01,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 0.0001,
      "step": 3850
    },
    {
      "epoch": 47.01,
      "learning_rate": 3.8888888888888895e-07,
      "loss": 0.0001,
      "step": 3860
    },
    {
      "epoch": 47.02,
      "learning_rate": 3.611111111111111e-07,
      "loss": 0.0001,
      "step": 3870
    },
    {
      "epoch": 47.02,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.0146,
      "step": 3880
    },
    {
      "epoch": 47.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.529157280921936,
      "eval_runtime": 9.2844,
      "eval_samples_per_second": 4.416,
      "eval_steps_per_second": 1.185,
      "step": 3888
    },
    {
      "epoch": 48.0,
      "learning_rate": 3.055555555555556e-07,
      "loss": 0.0001,
      "step": 3890
    },
    {
      "epoch": 48.0,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.0001,
      "step": 3900
    },
    {
      "epoch": 48.01,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.0001,
      "step": 3910
    },
    {
      "epoch": 48.01,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.0001,
      "step": 3920
    },
    {
      "epoch": 48.01,
      "learning_rate": 1.9444444444444447e-07,
      "loss": 0.0001,
      "step": 3930
    },
    {
      "epoch": 48.01,
      "learning_rate": 1.6666666666666668e-07,
      "loss": 0.0001,
      "step": 3940
    },
    {
      "epoch": 48.02,
      "learning_rate": 1.3888888888888888e-07,
      "loss": 0.0002,
      "step": 3950
    },
    {
      "epoch": 48.02,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 0.0001,
      "step": 3960
    },
    {
      "epoch": 48.02,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5404993295669556,
      "eval_runtime": 9.0406,
      "eval_samples_per_second": 4.535,
      "eval_steps_per_second": 1.217,
      "step": 3969
    },
    {
      "epoch": 49.0,
      "learning_rate": 8.333333333333334e-08,
      "loss": 0.0001,
      "step": 3970
    },
    {
      "epoch": 49.0,
      "learning_rate": 5.555555555555556e-08,
      "loss": 0.0001,
      "step": 3980
    },
    {
      "epoch": 49.01,
      "learning_rate": 2.777777777777778e-08,
      "loss": 0.0001,
      "step": 3990
    },
    {
      "epoch": 49.01,
      "learning_rate": 0.0,
      "loss": 0.0001,
      "step": 4000
    },
    {
      "epoch": 49.01,
      "eval_accuracy": 0.7804878048780488,
      "eval_loss": 1.5413603782653809,
      "eval_runtime": 9.061,
      "eval_samples_per_second": 4.525,
      "eval_steps_per_second": 1.214,
      "step": 4000
    },
    {
      "epoch": 49.01,
      "step": 4000,
      "total_flos": 6.9826100465849205e+19,
      "train_loss": 0.21371395963054965,
      "train_runtime": 6987.4008,
      "train_samples_per_second": 2.29,
      "train_steps_per_second": 0.572
    },
    {
      "epoch": 49.01,
      "eval_accuracy": 0.7317073170731707,
      "eval_loss": 0.8472180366516113,
      "eval_runtime": 10.3104,
      "eval_samples_per_second": 3.977,
      "eval_steps_per_second": 1.067,
      "step": 4000
    },
    {
      "epoch": 49.01,
      "eval_accuracy": 0.7317073170731707,
      "eval_loss": 0.8472180366516113,
      "eval_runtime": 9.5225,
      "eval_samples_per_second": 4.306,
      "eval_steps_per_second": 1.155,
      "step": 4000
    }
  ],
  "logging_steps": 10,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "total_flos": 6.9826100465849205e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}