{
  "best_metric": NaN,
  "best_model_checkpoint": "/scratch/czm5kz/finetuned_pythia70M_deduped_cp_14300032_1_0.0003_alternate_no_output/checkpoint-20",
  "epoch": 0.9975062344139651,
  "eval_steps": 20,
  "global_step": 1400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 15
    },
    {
      "epoch": 0.01,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 20
    },
    {
      "epoch": 0.01,
      "eval_loss": NaN,
      "eval_runtime": 16.0642,
      "eval_samples_per_second": 699.008,
      "eval_steps_per_second": 87.399,
      "step": 20
    },
    {
      "epoch": 0.02,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 25
    },
    {
      "epoch": 0.02,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 30
    },
    {
      "epoch": 0.02,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 35
    },
    {
      "epoch": 0.03,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 40
    },
    {
      "epoch": 0.03,
      "eval_loss": NaN,
      "eval_runtime": 15.4028,
      "eval_samples_per_second": 729.023,
      "eval_steps_per_second": 91.152,
      "step": 40
    },
    {
      "epoch": 0.03,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 45
    },
    {
      "epoch": 0.04,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 0.04,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 55
    },
    {
      "epoch": 0.04,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 60
    },
    {
      "epoch": 0.04,
      "eval_loss": NaN,
      "eval_runtime": 15.4237,
      "eval_samples_per_second": 728.035,
      "eval_steps_per_second": 91.029,
      "step": 60
    },
    {
      "epoch": 0.05,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 65
    },
    {
      "epoch": 0.05,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 70
    },
    {
      "epoch": 0.05,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 75
    },
    {
      "epoch": 0.06,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 80
    },
    {
      "epoch": 0.06,
      "eval_loss": NaN,
      "eval_runtime": 15.415,
      "eval_samples_per_second": 728.445,
      "eval_steps_per_second": 91.08,
      "step": 80
    },
    {
      "epoch": 0.06,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 85
    },
    {
      "epoch": 0.06,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 90
    },
    {
      "epoch": 0.07,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 95
    },
    {
      "epoch": 0.07,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 100
    },
    {
      "epoch": 0.07,
      "eval_loss": NaN,
      "eval_runtime": 15.3538,
      "eval_samples_per_second": 731.35,
      "eval_steps_per_second": 91.443,
      "step": 100
    },
    {
      "epoch": 0.07,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 105
    },
    {
      "epoch": 0.08,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 41116.4594,
      "step": 110
    },
    {
      "epoch": 0.08,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 115
    },
    {
      "epoch": 0.09,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 120
    },
    {
      "epoch": 0.09,
      "eval_loss": NaN,
      "eval_runtime": 15.3931,
      "eval_samples_per_second": 729.482,
      "eval_steps_per_second": 91.21,
      "step": 120
    },
    {
      "epoch": 0.09,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 125
    },
    {
      "epoch": 0.09,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 130
    },
    {
      "epoch": 0.1,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 135
    },
    {
      "epoch": 0.1,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 140
    },
    {
      "epoch": 0.1,
      "eval_loss": NaN,
      "eval_runtime": 15.3655,
      "eval_samples_per_second": 730.795,
      "eval_steps_per_second": 91.374,
      "step": 140
    },
    {
      "epoch": 0.1,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 145
    },
    {
      "epoch": 0.11,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 150
    },
    {
      "epoch": 0.11,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 155
    },
    {
      "epoch": 0.11,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 160
    },
    {
      "epoch": 0.11,
      "eval_loss": NaN,
      "eval_runtime": 15.3416,
      "eval_samples_per_second": 731.933,
      "eval_steps_per_second": 91.516,
      "step": 160
    },
    {
      "epoch": 0.12,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 165
    },
    {
      "epoch": 0.12,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 170
    },
    {
      "epoch": 0.12,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 175
    },
    {
      "epoch": 0.13,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.571,
      "step": 180
    },
    {
      "epoch": 0.13,
      "eval_loss": NaN,
      "eval_runtime": 15.3715,
      "eval_samples_per_second": 730.509,
      "eval_steps_per_second": 91.338,
      "step": 180
    },
    {
      "epoch": 0.13,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 185
    },
    {
      "epoch": 0.14,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 190
    },
    {
      "epoch": 0.14,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 195
    },
    {
      "epoch": 0.14,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 200
    },
    {
      "epoch": 0.14,
      "eval_loss": NaN,
      "eval_runtime": 15.3657,
      "eval_samples_per_second": 730.785,
      "eval_steps_per_second": 91.373,
      "step": 200
    },
    {
      "epoch": 0.15,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 205
    },
    {
      "epoch": 0.15,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 210
    },
    {
      "epoch": 0.15,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 215
    },
    {
      "epoch": 0.16,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 220
    },
    {
      "epoch": 0.16,
      "eval_loss": NaN,
      "eval_runtime": 15.3507,
      "eval_samples_per_second": 731.495,
      "eval_steps_per_second": 91.461,
      "step": 220
    },
    {
      "epoch": 0.16,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 225
    },
    {
      "epoch": 0.16,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 230
    },
    {
      "epoch": 0.17,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 235
    },
    {
      "epoch": 0.17,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 240
    },
    {
      "epoch": 0.17,
      "eval_loss": NaN,
      "eval_runtime": 15.3832,
      "eval_samples_per_second": 729.953,
      "eval_steps_per_second": 91.269,
      "step": 240
    },
    {
      "epoch": 0.17,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 245
    },
    {
      "epoch": 0.18,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 250
    },
    {
      "epoch": 0.18,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 255
    },
    {
      "epoch": 0.19,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 260
    },
    {
      "epoch": 0.19,
      "eval_loss": NaN,
      "eval_runtime": 15.3807,
      "eval_samples_per_second": 730.069,
      "eval_steps_per_second": 91.283,
      "step": 260
    },
    {
      "epoch": 0.19,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 265
    },
    {
      "epoch": 0.19,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 270
    },
    {
      "epoch": 0.2,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 275
    },
    {
      "epoch": 0.2,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 280
    },
    {
      "epoch": 0.2,
      "eval_loss": NaN,
      "eval_runtime": 15.4087,
      "eval_samples_per_second": 728.746,
      "eval_steps_per_second": 91.118,
      "step": 280
    },
    {
      "epoch": 0.2,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 285
    },
    {
      "epoch": 0.21,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 290
    },
    {
      "epoch": 0.21,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 295
    },
    {
      "epoch": 0.21,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 300
    },
    {
      "epoch": 0.21,
      "eval_loss": NaN,
      "eval_runtime": 15.4125,
      "eval_samples_per_second": 728.566,
      "eval_steps_per_second": 91.095,
      "step": 300
    },
    {
      "epoch": 0.22,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 305
    },
    {
      "epoch": 0.22,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 310
    },
    {
      "epoch": 0.22,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 315
    },
    {
      "epoch": 0.23,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 320
    },
    {
      "epoch": 0.23,
      "eval_loss": NaN,
      "eval_runtime": 15.3697,
      "eval_samples_per_second": 730.595,
      "eval_steps_per_second": 91.349,
      "step": 320
    },
    {
      "epoch": 0.23,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 325
    },
    {
      "epoch": 0.24,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 330
    },
    {
      "epoch": 0.24,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 335
    },
    {
      "epoch": 0.24,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 340
    },
    {
      "epoch": 0.24,
      "eval_loss": NaN,
      "eval_runtime": 15.3658,
      "eval_samples_per_second": 730.78,
      "eval_steps_per_second": 91.372,
      "step": 340
    },
    {
      "epoch": 0.25,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 345
    },
    {
      "epoch": 0.25,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 350
    },
    {
      "epoch": 0.25,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 355
    },
    {
      "epoch": 0.26,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 360
    },
    {
      "epoch": 0.26,
      "eval_loss": NaN,
      "eval_runtime": 15.367,
      "eval_samples_per_second": 730.722,
      "eval_steps_per_second": 91.365,
      "step": 360
    },
    {
      "epoch": 0.26,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 365
    },
    {
      "epoch": 0.26,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 370
    },
    {
      "epoch": 0.27,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 375
    },
    {
      "epoch": 0.27,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 380
    },
    {
      "epoch": 0.27,
      "eval_loss": NaN,
      "eval_runtime": 15.4049,
      "eval_samples_per_second": 728.925,
      "eval_steps_per_second": 91.14,
      "step": 380
    },
    {
      "epoch": 0.27,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 385
    },
    {
      "epoch": 0.28,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 390
    },
    {
      "epoch": 0.28,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 5.1255,
      "step": 395
    },
    {
      "epoch": 0.29,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 400
    },
    {
      "epoch": 0.29,
      "eval_loss": NaN,
      "eval_runtime": 15.3905,
      "eval_samples_per_second": 729.605,
      "eval_steps_per_second": 91.225,
      "step": 400
    },
    {
      "epoch": 0.29,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 405
    },
    {
      "epoch": 0.29,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 410
    },
    {
      "epoch": 0.3,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 415
    },
    {
      "epoch": 0.3,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 420
    },
    {
      "epoch": 0.3,
      "eval_loss": NaN,
      "eval_runtime": 15.3964,
      "eval_samples_per_second": 729.327,
      "eval_steps_per_second": 91.19,
      "step": 420
    },
    {
      "epoch": 0.3,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 425
    },
    {
      "epoch": 0.31,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 430
    },
    {
      "epoch": 0.31,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 435
    },
    {
      "epoch": 0.31,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 440
    },
    {
      "epoch": 0.31,
      "eval_loss": NaN,
      "eval_runtime": 15.3551,
      "eval_samples_per_second": 731.287,
      "eval_steps_per_second": 91.435,
      "step": 440
    },
    {
      "epoch": 0.32,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 445
    },
    {
      "epoch": 0.32,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 450
    },
    {
      "epoch": 0.32,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 455
    },
    {
      "epoch": 0.33,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 460
    },
    {
      "epoch": 0.33,
      "eval_loss": NaN,
      "eval_runtime": 15.3729,
      "eval_samples_per_second": 730.443,
      "eval_steps_per_second": 91.33,
      "step": 460
    },
    {
      "epoch": 0.33,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 465
    },
    {
      "epoch": 0.33,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 470
    },
    {
      "epoch": 0.34,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 475
    },
    {
      "epoch": 0.34,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 480
    },
    {
      "epoch": 0.34,
      "eval_loss": NaN,
      "eval_runtime": 15.3536,
      "eval_samples_per_second": 731.36,
      "eval_steps_per_second": 91.444,
      "step": 480
    },
    {
      "epoch": 0.35,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 485
    },
    {
      "epoch": 0.35,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 490
    },
    {
      "epoch": 0.35,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 495
    },
    {
      "epoch": 0.36,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 500
    },
    {
      "epoch": 0.36,
      "eval_loss": NaN,
      "eval_runtime": 15.351,
      "eval_samples_per_second": 731.484,
      "eval_steps_per_second": 91.46,
      "step": 500
    },
    {
      "epoch": 0.36,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 505
    },
    {
      "epoch": 0.36,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 510
    },
    {
      "epoch": 0.37,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 515
    },
    {
      "epoch": 0.37,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 520
    },
    {
      "epoch": 0.37,
      "eval_loss": NaN,
      "eval_runtime": 15.3883,
      "eval_samples_per_second": 729.709,
      "eval_steps_per_second": 91.238,
      "step": 520
    },
    {
      "epoch": 0.37,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 525
    },
    {
      "epoch": 0.38,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 530
    },
    {
      "epoch": 0.38,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 535
    },
    {
      "epoch": 0.38,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 540
    },
    {
      "epoch": 0.38,
      "eval_loss": NaN,
      "eval_runtime": 15.3597,
      "eval_samples_per_second": 731.068,
      "eval_steps_per_second": 91.408,
      "step": 540
    },
    {
      "epoch": 0.39,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 545
    },
    {
      "epoch": 0.39,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 550
    },
    {
      "epoch": 0.4,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 555
    },
    {
      "epoch": 0.4,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 560
    },
    {
      "epoch": 0.4,
      "eval_loss": NaN,
      "eval_runtime": 15.3532,
      "eval_samples_per_second": 731.379,
      "eval_steps_per_second": 91.447,
      "step": 560
    },
    {
      "epoch": 0.4,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 565
    },
    {
      "epoch": 0.41,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 570
    },
    {
      "epoch": 0.41,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 575
    },
    {
      "epoch": 0.41,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 580
    },
    {
      "epoch": 0.41,
      "eval_loss": NaN,
      "eval_runtime": 15.3602,
      "eval_samples_per_second": 731.044,
      "eval_steps_per_second": 91.405,
      "step": 580
    },
    {
      "epoch": 0.42,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 585
    },
    {
      "epoch": 0.42,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 590
    },
    {
      "epoch": 0.42,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 595
    },
    {
      "epoch": 0.43,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 600
    },
    {
      "epoch": 0.43,
      "eval_loss": NaN,
      "eval_runtime": 15.3417,
      "eval_samples_per_second": 731.929,
      "eval_steps_per_second": 91.516,
      "step": 600
    },
    {
      "epoch": 0.43,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 605
    },
    {
      "epoch": 0.43,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 610
    },
    {
      "epoch": 0.44,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 615
    },
    {
      "epoch": 0.44,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 620
    },
    {
      "epoch": 0.44,
      "eval_loss": NaN,
      "eval_runtime": 15.3943,
      "eval_samples_per_second": 729.425,
      "eval_steps_per_second": 91.203,
      "step": 620
    },
    {
      "epoch": 0.45,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 625
    },
    {
      "epoch": 0.45,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 630
    },
    {
      "epoch": 0.45,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 635
    },
    {
      "epoch": 0.46,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 640
    },
    {
      "epoch": 0.46,
      "eval_loss": NaN,
      "eval_runtime": 15.3821,
      "eval_samples_per_second": 730.003,
      "eval_steps_per_second": 91.275,
      "step": 640
    },
    {
      "epoch": 0.46,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 645
    },
    {
      "epoch": 0.46,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 650
    },
    {
      "epoch": 0.47,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 655
    },
    {
      "epoch": 0.47,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 660
    },
    {
      "epoch": 0.47,
      "eval_loss": NaN,
      "eval_runtime": 15.3877,
      "eval_samples_per_second": 729.738,
      "eval_steps_per_second": 91.242,
      "step": 660
    },
    {
      "epoch": 0.47,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 665
    },
    {
      "epoch": 0.48,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 670
    },
    {
      "epoch": 0.48,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 675
    },
    {
      "epoch": 0.48,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 680
    },
    {
      "epoch": 0.48,
      "eval_loss": NaN,
      "eval_runtime": 15.369,
      "eval_samples_per_second": 730.629,
      "eval_steps_per_second": 91.353,
      "step": 680
    },
    {
      "epoch": 0.49,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 685
    },
    {
      "epoch": 0.49,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 690
    },
    {
      "epoch": 0.5,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 695
    },
    {
      "epoch": 0.5,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 700
    },
    {
      "epoch": 0.5,
      "eval_loss": NaN,
      "eval_runtime": 15.3711,
      "eval_samples_per_second": 730.526,
      "eval_steps_per_second": 91.34,
      "step": 700
    },
    {
      "epoch": 0.5,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 705
    },
    {
      "epoch": 0.51,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 710
    },
    {
      "epoch": 0.51,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 715
    },
    {
      "epoch": 0.51,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 720
    },
    {
      "epoch": 0.51,
      "eval_loss": NaN,
      "eval_runtime": 15.3467,
      "eval_samples_per_second": 731.689,
      "eval_steps_per_second": 91.486,
      "step": 720
    },
    {
      "epoch": 0.52,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 725
    },
    {
      "epoch": 0.52,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 730
    },
    {
      "epoch": 0.52,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 735
    },
    {
      "epoch": 0.53,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 740
    },
    {
      "epoch": 0.53,
      "eval_loss": NaN,
      "eval_runtime": 15.3106,
      "eval_samples_per_second": 733.414,
      "eval_steps_per_second": 91.701,
      "step": 740
    },
    {
      "epoch": 0.53,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 745
    },
    {
      "epoch": 0.53,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 750
    },
    {
      "epoch": 0.54,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 755
    },
    {
      "epoch": 0.54,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 760
    },
    {
      "epoch": 0.54,
      "eval_loss": NaN,
      "eval_runtime": 15.3582,
      "eval_samples_per_second": 731.14,
      "eval_steps_per_second": 91.417,
      "step": 760
    },
    {
      "epoch": 0.55,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 765
    },
    {
      "epoch": 0.55,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 770
    },
    {
      "epoch": 0.55,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 775
    },
    {
      "epoch": 0.56,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 780
    },
    {
      "epoch": 0.56,
      "eval_loss": NaN,
      "eval_runtime": 15.3395,
      "eval_samples_per_second": 732.032,
      "eval_steps_per_second": 91.528,
      "step": 780
    },
    {
      "epoch": 0.56,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 785
    },
    {
      "epoch": 0.56,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 47.4835,
      "step": 790
    },
    {
      "epoch": 0.57,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 795
    },
    {
      "epoch": 0.57,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 800
    },
    {
      "epoch": 0.57,
      "eval_loss": NaN,
      "eval_runtime": 15.3823,
      "eval_samples_per_second": 729.995,
      "eval_steps_per_second": 91.274,
      "step": 800
    },
    {
      "epoch": 0.57,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 805
    },
    {
      "epoch": 0.58,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 810
    },
    {
      "epoch": 0.58,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 815
    },
    {
      "epoch": 0.58,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 820
    },
    {
      "epoch": 0.58,
      "eval_loss": NaN,
      "eval_runtime": 15.3882,
      "eval_samples_per_second": 729.716,
      "eval_steps_per_second": 91.239,
      "step": 820
    },
    {
      "epoch": 0.59,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 825
    },
    {
      "epoch": 0.59,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 830
    },
    {
      "epoch": 0.59,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 835
    },
    {
      "epoch": 0.6,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 840
    },
    {
      "epoch": 0.6,
      "eval_loss": NaN,
      "eval_runtime": 15.3598,
      "eval_samples_per_second": 731.066,
      "eval_steps_per_second": 91.408,
      "step": 840
    },
    {
      "epoch": 0.6,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 845
    },
    {
      "epoch": 0.61,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 850
    },
    {
      "epoch": 0.61,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 855
    },
    {
      "epoch": 0.61,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 860
    },
    {
      "epoch": 0.61,
      "eval_loss": NaN,
      "eval_runtime": 15.3292,
      "eval_samples_per_second": 732.523,
      "eval_steps_per_second": 91.59,
      "step": 860
    },
    {
      "epoch": 0.62,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 865
    },
    {
      "epoch": 0.62,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 870
    },
    {
      "epoch": 0.62,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 875
    },
    {
      "epoch": 0.63,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 880
    },
    {
      "epoch": 0.63,
      "eval_loss": NaN,
      "eval_runtime": 15.3262,
      "eval_samples_per_second": 732.668,
      "eval_steps_per_second": 91.608,
      "step": 880
    },
    {
      "epoch": 0.63,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 885
    },
    {
      "epoch": 0.63,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 890
    },
    {
      "epoch": 0.64,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 895
    },
    {
      "epoch": 0.64,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 900
    },
    {
      "epoch": 0.64,
      "eval_loss": NaN,
      "eval_runtime": 15.3603,
      "eval_samples_per_second": 731.039,
      "eval_steps_per_second": 91.404,
      "step": 900
    },
    {
      "epoch": 0.64,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 905
    },
    {
      "epoch": 0.65,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 910
    },
    {
      "epoch": 0.65,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 915
    },
    {
      "epoch": 0.66,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 920
    },
    {
      "epoch": 0.66,
      "eval_loss": NaN,
      "eval_runtime": 15.3565,
      "eval_samples_per_second": 731.221,
      "eval_steps_per_second": 91.427,
      "step": 920
    },
    {
      "epoch": 0.66,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 925
    },
    {
      "epoch": 0.66,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 930
    },
    {
      "epoch": 0.67,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 935
    },
    {
      "epoch": 0.67,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 940
    },
    {
      "epoch": 0.67,
      "eval_loss": NaN,
      "eval_runtime": 15.3771,
      "eval_samples_per_second": 730.243,
      "eval_steps_per_second": 91.305,
      "step": 940
    },
    {
      "epoch": 0.67,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 945
    },
    {
      "epoch": 0.68,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 950
    },
    {
      "epoch": 0.68,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 955
    },
    {
      "epoch": 0.68,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 960
    },
    {
      "epoch": 0.68,
      "eval_loss": NaN,
      "eval_runtime": 15.3362,
      "eval_samples_per_second": 732.188,
      "eval_steps_per_second": 91.548,
      "step": 960
    },
    {
      "epoch": 0.69,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 965
    },
    {
      "epoch": 0.69,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 970
    },
    {
      "epoch": 0.69,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 975
    },
    {
      "epoch": 0.7,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 980
    },
    {
      "epoch": 0.7,
      "eval_loss": NaN,
      "eval_runtime": 15.3547,
      "eval_samples_per_second": 731.308,
      "eval_steps_per_second": 91.438,
      "step": 980
    },
    {
      "epoch": 0.7,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 985
    },
    {
      "epoch": 0.71,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 990
    },
    {
      "epoch": 0.71,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 995
    },
    {
      "epoch": 0.71,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1000
    },
    {
      "epoch": 0.71,
      "eval_loss": NaN,
      "eval_runtime": 15.3384,
      "eval_samples_per_second": 732.084,
      "eval_steps_per_second": 91.535,
      "step": 1000
    },
    {
      "epoch": 0.72,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1005
    },
    {
      "epoch": 0.72,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1010
    },
    {
      "epoch": 0.72,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1015
    },
    {
      "epoch": 0.73,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1020
    },
    {
      "epoch": 0.73,
      "eval_loss": NaN,
      "eval_runtime": 15.403,
      "eval_samples_per_second": 729.014,
      "eval_steps_per_second": 91.151,
      "step": 1020
    },
    {
      "epoch": 0.73,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1025
    },
    {
      "epoch": 0.73,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1030
    },
    {
      "epoch": 0.74,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1035
    },
    {
      "epoch": 0.74,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1040
    },
    {
      "epoch": 0.74,
      "eval_loss": NaN,
      "eval_runtime": 15.3539,
      "eval_samples_per_second": 731.345,
      "eval_steps_per_second": 91.443,
      "step": 1040
    },
    {
      "epoch": 0.74,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1045
    },
    {
      "epoch": 0.75,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1050
    },
    {
      "epoch": 0.75,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1055
    },
    {
      "epoch": 0.76,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1060
    },
    {
      "epoch": 0.76,
      "eval_loss": NaN,
      "eval_runtime": 15.3654,
      "eval_samples_per_second": 730.796,
      "eval_steps_per_second": 91.374,
      "step": 1060
    },
    {
      "epoch": 0.76,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1065
    },
    {
      "epoch": 0.76,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1070
    },
    {
      "epoch": 0.77,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1075
    },
    {
      "epoch": 0.77,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1080
    },
    {
      "epoch": 0.77,
      "eval_loss": NaN,
      "eval_runtime": 15.3857,
      "eval_samples_per_second": 729.833,
      "eval_steps_per_second": 91.253,
      "step": 1080
    },
    {
      "epoch": 0.77,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1085
    },
    {
      "epoch": 0.78,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1090
    },
    {
      "epoch": 0.78,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1095
    },
    {
      "epoch": 0.78,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1100
    },
    {
      "epoch": 0.78,
      "eval_loss": NaN,
      "eval_runtime": 15.3963,
      "eval_samples_per_second": 729.329,
      "eval_steps_per_second": 91.19,
      "step": 1100
    },
    {
      "epoch": 0.79,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1105
    },
    {
      "epoch": 0.79,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1110
    },
    {
      "epoch": 0.79,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1115
    },
    {
      "epoch": 0.8,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1120
    },
    {
      "epoch": 0.8,
      "eval_loss": NaN,
      "eval_runtime": 15.4262,
      "eval_samples_per_second": 727.917,
      "eval_steps_per_second": 91.014,
      "step": 1120
    },
    {
      "epoch": 0.8,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1125
    },
    {
      "epoch": 0.81,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1130
    },
    {
      "epoch": 0.81,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1135
    },
    {
      "epoch": 0.81,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1140
    },
    {
      "epoch": 0.81,
      "eval_loss": NaN,
      "eval_runtime": 15.4177,
      "eval_samples_per_second": 728.321,
      "eval_steps_per_second": 91.064,
      "step": 1140
    },
    {
      "epoch": 0.82,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1145
    },
    {
      "epoch": 0.82,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1150
    },
    {
      "epoch": 0.82,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1155
    },
    {
      "epoch": 0.83,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1160
    },
    {
      "epoch": 0.83,
      "eval_loss": NaN,
      "eval_runtime": 15.3795,
      "eval_samples_per_second": 730.13,
      "eval_steps_per_second": 91.291,
      "step": 1160
    },
    {
      "epoch": 0.83,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1165
    },
    {
      "epoch": 0.83,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1170
    },
    {
      "epoch": 0.84,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1175
    },
    {
      "epoch": 0.84,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1180
    },
    {
      "epoch": 0.84,
      "eval_loss": NaN,
      "eval_runtime": 15.3815,
      "eval_samples_per_second": 730.031,
      "eval_steps_per_second": 91.278,
      "step": 1180
    },
    {
      "epoch": 0.84,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 1408.5843,
      "step": 1185
    },
    {
      "epoch": 0.85,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1190
    },
    {
      "epoch": 0.85,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1195
    },
    {
      "epoch": 0.86,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1200
    },
    {
      "epoch": 0.86,
      "eval_loss": NaN,
      "eval_runtime": 15.3703,
      "eval_samples_per_second": 730.564,
      "eval_steps_per_second": 91.345,
      "step": 1200
    },
    {
      "epoch": 0.86,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1205
    },
    {
      "epoch": 0.86,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1210
    },
    {
      "epoch": 0.87,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1215
    },
    {
      "epoch": 0.87,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1220
    },
    {
      "epoch": 0.87,
      "eval_loss": NaN,
      "eval_runtime": 15.3663,
      "eval_samples_per_second": 730.756,
      "eval_steps_per_second": 91.369,
      "step": 1220
    },
    {
      "epoch": 0.87,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1225
    },
    {
      "epoch": 0.88,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1230
    },
    {
      "epoch": 0.88,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1235
    },
    {
      "epoch": 0.88,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1240
    },
    {
      "epoch": 0.88,
      "eval_loss": NaN,
      "eval_runtime": 15.3411,
      "eval_samples_per_second": 731.955,
      "eval_steps_per_second": 91.519,
      "step": 1240
    },
    {
      "epoch": 0.89,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1245
    },
    {
      "epoch": 0.89,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1250
    },
    {
      "epoch": 0.89,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1255
    },
    {
      "epoch": 0.9,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1260
    },
    {
      "epoch": 0.9,
      "eval_loss": NaN,
      "eval_runtime": 15.35,
      "eval_samples_per_second": 731.53,
      "eval_steps_per_second": 91.466,
      "step": 1260
    },
    {
      "epoch": 0.9,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1265
    },
    {
      "epoch": 0.9,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1270
    },
    {
      "epoch": 0.91,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1275
    },
    {
      "epoch": 0.91,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1280
    },
    {
      "epoch": 0.91,
      "eval_loss": NaN,
      "eval_runtime": 15.3759,
      "eval_samples_per_second": 730.297,
      "eval_steps_per_second": 91.311,
      "step": 1280
    },
    {
      "epoch": 0.92,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1285
    },
    {
      "epoch": 0.92,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1290
    },
    {
      "epoch": 0.92,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1295
    },
    {
      "epoch": 0.93,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1300
    },
    {
      "epoch": 0.93,
      "eval_loss": NaN,
      "eval_runtime": 15.3714,
      "eval_samples_per_second": 730.513,
      "eval_steps_per_second": 91.339,
      "step": 1300
    },
    {
      "epoch": 0.93,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1305
    },
    {
      "epoch": 0.93,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1310
    },
    {
      "epoch": 0.94,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1315
    },
    {
      "epoch": 0.94,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1320
    },
    {
      "epoch": 0.94,
      "eval_loss": NaN,
      "eval_runtime": 15.4163,
      "eval_samples_per_second": 728.384,
      "eval_steps_per_second": 91.072,
      "step": 1320
    },
    {
      "epoch": 0.94,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1325
    },
    {
      "epoch": 0.95,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1330
    },
    {
      "epoch": 0.95,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1335
    },
    {
      "epoch": 0.95,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1340
    },
    {
      "epoch": 0.95,
      "eval_loss": NaN,
      "eval_runtime": 15.4318,
      "eval_samples_per_second": 727.651,
      "eval_steps_per_second": 90.981,
      "step": 1340
    },
    {
      "epoch": 0.96,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1345
    },
    {
      "epoch": 0.96,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1350
    },
    {
      "epoch": 0.97,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1355
    },
    {
      "epoch": 0.97,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1360
    },
    {
      "epoch": 0.97,
      "eval_loss": NaN,
      "eval_runtime": 15.4043,
      "eval_samples_per_second": 728.952,
      "eval_steps_per_second": 91.143,
      "step": 1360
    },
    {
      "epoch": 0.97,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1365
    },
    {
      "epoch": 0.98,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1370
    },
    {
      "epoch": 0.98,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1375
    },
    {
      "epoch": 0.98,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1380
    },
    {
      "epoch": 0.98,
      "eval_loss": NaN,
      "eval_runtime": 15.383,
      "eval_samples_per_second": 729.96,
      "eval_steps_per_second": 91.269,
      "step": 1380
    },
    {
      "epoch": 0.99,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1385
    },
    {
      "epoch": 0.99,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1390
    },
    {
      "epoch": 0.99,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1395
    },
    {
      "epoch": 1.0,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.0,
      "step": 1400
    },
    {
      "epoch": 1.0,
      "eval_loss": NaN,
      "eval_runtime": 15.4904,
      "eval_samples_per_second": 724.9,
      "eval_steps_per_second": 90.637,
      "step": 1400
    }
  ],
  "logging_steps": 5,
  "max_steps": 1403,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 20,
  "total_flos": 1536241035116544.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}