{
  "best_metric": 3.643535614013672,
  "best_model_checkpoint": "./drive/MyDrive/repositories/torch_example_image-classification/outputs_letter3/checkpoint-700",
  "epoch": 100.0,
  "eval_steps": 500,
  "global_step": 700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.030612244897959183,
      "eval_loss": 3.945223808288574,
      "eval_runtime": 35.9416,
      "eval_samples_per_second": 2.727,
      "eval_steps_per_second": 0.056,
      "step": 7
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.9714285714285718e-05,
      "loss": 3.9498,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.05102040816326531,
      "eval_loss": 3.943814277648926,
      "eval_runtime": 1.6003,
      "eval_samples_per_second": 61.237,
      "eval_steps_per_second": 1.25,
      "step": 14
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.942857142857143e-05,
      "loss": 3.9413,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.04081632653061224,
      "eval_loss": 3.9436657428741455,
      "eval_runtime": 1.5584,
      "eval_samples_per_second": 62.886,
      "eval_steps_per_second": 1.283,
      "step": 21
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.04081632653061224,
      "eval_loss": 3.9431352615356445,
      "eval_runtime": 1.6475,
      "eval_samples_per_second": 59.483,
      "eval_steps_per_second": 1.214,
      "step": 28
    },
    {
      "epoch": 4.29,
      "learning_rate": 1.9142857142857146e-05,
      "loss": 3.9255,
      "step": 30
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.04081632653061224,
      "eval_loss": 3.9424338340759277,
      "eval_runtime": 1.5527,
      "eval_samples_per_second": 63.114,
      "eval_steps_per_second": 1.288,
      "step": 35
    },
    {
      "epoch": 5.71,
      "learning_rate": 1.885714285714286e-05,
      "loss": 3.9132,
      "step": 40
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.030612244897959183,
      "eval_loss": 3.9400839805603027,
      "eval_runtime": 1.6728,
      "eval_samples_per_second": 58.585,
      "eval_steps_per_second": 1.196,
      "step": 42
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.030612244897959183,
      "eval_loss": 3.937299966812134,
      "eval_runtime": 1.6864,
      "eval_samples_per_second": 58.111,
      "eval_steps_per_second": 1.186,
      "step": 49
    },
    {
      "epoch": 7.14,
      "learning_rate": 1.8571428571428575e-05,
      "loss": 3.8913,
      "step": 50
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.02040816326530612,
      "eval_loss": 3.9351377487182617,
      "eval_runtime": 1.7475,
      "eval_samples_per_second": 56.08,
      "eval_steps_per_second": 1.144,
      "step": 56
    },
    {
      "epoch": 8.57,
      "learning_rate": 1.8285714285714288e-05,
      "loss": 3.8685,
      "step": 60
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.02040816326530612,
      "eval_loss": 3.931184768676758,
      "eval_runtime": 1.7165,
      "eval_samples_per_second": 57.092,
      "eval_steps_per_second": 1.165,
      "step": 63
    },
    {
      "epoch": 10.0,
      "learning_rate": 1.8e-05,
      "loss": 3.8413,
      "step": 70
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.030612244897959183,
      "eval_loss": 3.9259071350097656,
      "eval_runtime": 1.7086,
      "eval_samples_per_second": 57.357,
      "eval_steps_per_second": 1.171,
      "step": 70
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.030612244897959183,
      "eval_loss": 3.9218833446502686,
      "eval_runtime": 1.6006,
      "eval_samples_per_second": 61.228,
      "eval_steps_per_second": 1.25,
      "step": 77
    },
    {
      "epoch": 11.43,
      "learning_rate": 1.7714285714285717e-05,
      "loss": 3.8163,
      "step": 80
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.02040816326530612,
      "eval_loss": 3.9182682037353516,
      "eval_runtime": 1.8155,
      "eval_samples_per_second": 53.98,
      "eval_steps_per_second": 1.102,
      "step": 84
    },
    {
      "epoch": 12.86,
      "learning_rate": 1.742857142857143e-05,
      "loss": 3.7912,
      "step": 90
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.04081632653061224,
      "eval_loss": 3.9151253700256348,
      "eval_runtime": 1.704,
      "eval_samples_per_second": 57.511,
      "eval_steps_per_second": 1.174,
      "step": 91
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.030612244897959183,
      "eval_loss": 3.911611795425415,
      "eval_runtime": 1.7037,
      "eval_samples_per_second": 57.521,
      "eval_steps_per_second": 1.174,
      "step": 98
    },
    {
      "epoch": 14.29,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 3.7616,
      "step": 100
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.04081632653061224,
      "eval_loss": 3.9074392318725586,
      "eval_runtime": 1.6801,
      "eval_samples_per_second": 58.331,
      "eval_steps_per_second": 1.19,
      "step": 105
    },
    {
      "epoch": 15.71,
      "learning_rate": 1.6857142857142858e-05,
      "loss": 3.734,
      "step": 110
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.04081632653061224,
      "eval_loss": 3.9029488563537598,
      "eval_runtime": 1.7659,
      "eval_samples_per_second": 55.496,
      "eval_steps_per_second": 1.133,
      "step": 112
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.061224489795918366,
      "eval_loss": 3.8968818187713623,
      "eval_runtime": 1.6451,
      "eval_samples_per_second": 59.57,
      "eval_steps_per_second": 1.216,
      "step": 119
    },
    {
      "epoch": 17.14,
      "learning_rate": 1.6571428571428574e-05,
      "loss": 3.7014,
      "step": 120
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.07142857142857142,
      "eval_loss": 3.890695095062256,
      "eval_runtime": 1.6901,
      "eval_samples_per_second": 57.983,
      "eval_steps_per_second": 1.183,
      "step": 126
    },
    {
      "epoch": 18.57,
      "learning_rate": 1.6285714285714287e-05,
      "loss": 3.6707,
      "step": 130
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.07142857142857142,
      "eval_loss": 3.8844714164733887,
      "eval_runtime": 1.6505,
      "eval_samples_per_second": 59.374,
      "eval_steps_per_second": 1.212,
      "step": 133
    },
    {
      "epoch": 20.0,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 3.6307,
      "step": 140
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.08163265306122448,
      "eval_loss": 3.877913475036621,
      "eval_runtime": 1.8041,
      "eval_samples_per_second": 54.322,
      "eval_steps_per_second": 1.109,
      "step": 140
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.08163265306122448,
      "eval_loss": 3.8703930377960205,
      "eval_runtime": 1.7575,
      "eval_samples_per_second": 55.762,
      "eval_steps_per_second": 1.138,
      "step": 147
    },
    {
      "epoch": 21.43,
      "learning_rate": 1.5714285714285715e-05,
      "loss": 3.596,
      "step": 150
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.09183673469387756,
      "eval_loss": 3.8646252155303955,
      "eval_runtime": 1.6598,
      "eval_samples_per_second": 59.042,
      "eval_steps_per_second": 1.205,
      "step": 154
    },
    {
      "epoch": 22.86,
      "learning_rate": 1.542857142857143e-05,
      "loss": 3.5875,
      "step": 160
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.09183673469387756,
      "eval_loss": 3.8603618144989014,
      "eval_runtime": 1.6878,
      "eval_samples_per_second": 58.065,
      "eval_steps_per_second": 1.185,
      "step": 161
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.09183673469387756,
      "eval_loss": 3.8561482429504395,
      "eval_runtime": 1.6677,
      "eval_samples_per_second": 58.764,
      "eval_steps_per_second": 1.199,
      "step": 168
    },
    {
      "epoch": 24.29,
      "learning_rate": 1.5142857142857144e-05,
      "loss": 3.5532,
      "step": 170
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.09183673469387756,
      "eval_loss": 3.8509910106658936,
      "eval_runtime": 1.7363,
      "eval_samples_per_second": 56.441,
      "eval_steps_per_second": 1.152,
      "step": 175
    },
    {
      "epoch": 25.71,
      "learning_rate": 1.4857142857142858e-05,
      "loss": 3.5374,
      "step": 180
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.09183673469387756,
      "eval_loss": 3.844221353530884,
      "eval_runtime": 1.7776,
      "eval_samples_per_second": 55.131,
      "eval_steps_per_second": 1.125,
      "step": 182
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.10204081632653061,
      "eval_loss": 3.83986496925354,
      "eval_runtime": 1.6514,
      "eval_samples_per_second": 59.342,
      "eval_steps_per_second": 1.211,
      "step": 189
    },
    {
      "epoch": 27.14,
      "learning_rate": 1.4571428571428573e-05,
      "loss": 3.51,
      "step": 190
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.11224489795918367,
      "eval_loss": 3.8350086212158203,
      "eval_runtime": 1.6793,
      "eval_samples_per_second": 58.357,
      "eval_steps_per_second": 1.191,
      "step": 196
    },
    {
      "epoch": 28.57,
      "learning_rate": 1.4285714285714287e-05,
      "loss": 3.4842,
      "step": 200
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.8296377658843994,
      "eval_runtime": 1.7517,
      "eval_samples_per_second": 55.946,
      "eval_steps_per_second": 1.142,
      "step": 203
    },
    {
      "epoch": 30.0,
      "learning_rate": 1.4e-05,
      "loss": 3.4495,
      "step": 210
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.8243255615234375,
      "eval_runtime": 1.7837,
      "eval_samples_per_second": 54.942,
      "eval_steps_per_second": 1.121,
      "step": 210
    },
    {
      "epoch": 31.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.8213043212890625,
      "eval_runtime": 1.6545,
      "eval_samples_per_second": 59.233,
      "eval_steps_per_second": 1.209,
      "step": 217
    },
    {
      "epoch": 31.43,
      "learning_rate": 1.3714285714285716e-05,
      "loss": 3.4155,
      "step": 220
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.815812110900879,
      "eval_runtime": 1.6721,
      "eval_samples_per_second": 58.61,
      "eval_steps_per_second": 1.196,
      "step": 224
    },
    {
      "epoch": 32.86,
      "learning_rate": 1.3428571428571429e-05,
      "loss": 3.4257,
      "step": 230
    },
    {
      "epoch": 33.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.8117594718933105,
      "eval_runtime": 1.6581,
      "eval_samples_per_second": 59.102,
      "eval_steps_per_second": 1.206,
      "step": 231
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.1326530612244898,
      "eval_loss": 3.8061439990997314,
      "eval_runtime": 1.6272,
      "eval_samples_per_second": 60.227,
      "eval_steps_per_second": 1.229,
      "step": 238
    },
    {
      "epoch": 34.29,
      "learning_rate": 1.3142857142857145e-05,
      "loss": 3.395,
      "step": 240
    },
    {
      "epoch": 35.0,
      "eval_accuracy": 0.1326530612244898,
      "eval_loss": 3.8029837608337402,
      "eval_runtime": 1.6449,
      "eval_samples_per_second": 59.577,
      "eval_steps_per_second": 1.216,
      "step": 245
    },
    {
      "epoch": 35.71,
      "learning_rate": 1.2857142857142859e-05,
      "loss": 3.3693,
      "step": 250
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.795672655105591,
      "eval_runtime": 1.6624,
      "eval_samples_per_second": 58.953,
      "eval_steps_per_second": 1.203,
      "step": 252
    },
    {
      "epoch": 37.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.790419578552246,
      "eval_runtime": 1.6455,
      "eval_samples_per_second": 59.556,
      "eval_steps_per_second": 1.215,
      "step": 259
    },
    {
      "epoch": 37.14,
      "learning_rate": 1.2571428571428572e-05,
      "loss": 3.35,
      "step": 260
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.783411741256714,
      "eval_runtime": 1.6562,
      "eval_samples_per_second": 59.172,
      "eval_steps_per_second": 1.208,
      "step": 266
    },
    {
      "epoch": 38.57,
      "learning_rate": 1.2285714285714288e-05,
      "loss": 3.3453,
      "step": 270
    },
    {
      "epoch": 39.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.778721332550049,
      "eval_runtime": 1.7932,
      "eval_samples_per_second": 54.651,
      "eval_steps_per_second": 1.115,
      "step": 273
    },
    {
      "epoch": 40.0,
      "learning_rate": 1.2e-05,
      "loss": 3.2977,
      "step": 280
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.772735357284546,
      "eval_runtime": 1.6336,
      "eval_samples_per_second": 59.991,
      "eval_steps_per_second": 1.224,
      "step": 280
    },
    {
      "epoch": 41.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.7680680751800537,
      "eval_runtime": 1.7171,
      "eval_samples_per_second": 57.072,
      "eval_steps_per_second": 1.165,
      "step": 287
    },
    {
      "epoch": 41.43,
      "learning_rate": 1.1714285714285716e-05,
      "loss": 3.2875,
      "step": 290
    },
    {
      "epoch": 42.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.762765407562256,
      "eval_runtime": 1.6284,
      "eval_samples_per_second": 60.181,
      "eval_steps_per_second": 1.228,
      "step": 294
    },
    {
      "epoch": 42.86,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 3.2504,
      "step": 300
    },
    {
      "epoch": 43.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.758150100708008,
      "eval_runtime": 1.6731,
      "eval_samples_per_second": 58.575,
      "eval_steps_per_second": 1.195,
      "step": 301
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.7527263164520264,
      "eval_runtime": 1.7657,
      "eval_samples_per_second": 55.502,
      "eval_steps_per_second": 1.133,
      "step": 308
    },
    {
      "epoch": 44.29,
      "learning_rate": 1.1142857142857143e-05,
      "loss": 3.2772,
      "step": 310
    },
    {
      "epoch": 45.0,
      "eval_accuracy": 0.12244897959183673,
      "eval_loss": 3.749258041381836,
      "eval_runtime": 1.6842,
      "eval_samples_per_second": 58.188,
      "eval_steps_per_second": 1.188,
      "step": 315
    },
    {
      "epoch": 45.71,
      "learning_rate": 1.0857142857142858e-05,
      "loss": 3.2353,
      "step": 320
    },
    {
      "epoch": 46.0,
      "eval_accuracy": 0.11224489795918367,
      "eval_loss": 3.7462167739868164,
      "eval_runtime": 1.6536,
      "eval_samples_per_second": 59.264,
      "eval_steps_per_second": 1.209,
      "step": 322
    },
    {
      "epoch": 47.0,
      "eval_accuracy": 0.1326530612244898,
      "eval_loss": 3.743089437484741,
      "eval_runtime": 1.639,
      "eval_samples_per_second": 59.793,
      "eval_steps_per_second": 1.22,
      "step": 329
    },
    {
      "epoch": 47.14,
      "learning_rate": 1.0571428571428572e-05,
      "loss": 3.2198,
      "step": 330
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.1326530612244898,
      "eval_loss": 3.7392406463623047,
      "eval_runtime": 1.6659,
      "eval_samples_per_second": 58.827,
      "eval_steps_per_second": 1.201,
      "step": 336
    },
    {
      "epoch": 48.57,
      "learning_rate": 1.0285714285714285e-05,
      "loss": 3.204,
      "step": 340
    },
    {
      "epoch": 49.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.73702073097229,
      "eval_runtime": 1.7576,
      "eval_samples_per_second": 55.758,
      "eval_steps_per_second": 1.138,
      "step": 343
    },
    {
      "epoch": 50.0,
      "learning_rate": 1e-05,
      "loss": 3.1762,
      "step": 350
    },
    {
      "epoch": 50.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.733855962753296,
      "eval_runtime": 1.6472,
      "eval_samples_per_second": 59.494,
      "eval_steps_per_second": 1.214,
      "step": 350
    },
    {
      "epoch": 51.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.730581283569336,
      "eval_runtime": 1.6737,
      "eval_samples_per_second": 58.554,
      "eval_steps_per_second": 1.195,
      "step": 357
    },
    {
      "epoch": 51.43,
      "learning_rate": 9.714285714285715e-06,
      "loss": 3.1741,
      "step": 360
    },
    {
      "epoch": 52.0,
      "eval_accuracy": 0.16326530612244897,
      "eval_loss": 3.7267162799835205,
      "eval_runtime": 1.669,
      "eval_samples_per_second": 58.718,
      "eval_steps_per_second": 1.198,
      "step": 364
    },
    {
      "epoch": 52.86,
      "learning_rate": 9.42857142857143e-06,
      "loss": 3.1757,
      "step": 370
    },
    {
      "epoch": 53.0,
      "eval_accuracy": 0.16326530612244897,
      "eval_loss": 3.7221927642822266,
      "eval_runtime": 1.6595,
      "eval_samples_per_second": 59.054,
      "eval_steps_per_second": 1.205,
      "step": 371
    },
    {
      "epoch": 54.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.717994451522827,
      "eval_runtime": 1.6603,
      "eval_samples_per_second": 59.027,
      "eval_steps_per_second": 1.205,
      "step": 378
    },
    {
      "epoch": 54.29,
      "learning_rate": 9.142857142857144e-06,
      "loss": 3.1492,
      "step": 380
    },
    {
      "epoch": 55.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.7148733139038086,
      "eval_runtime": 1.6452,
      "eval_samples_per_second": 59.566,
      "eval_steps_per_second": 1.216,
      "step": 385
    },
    {
      "epoch": 55.71,
      "learning_rate": 8.857142857142858e-06,
      "loss": 3.1442,
      "step": 390
    },
    {
      "epoch": 56.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.7107248306274414,
      "eval_runtime": 1.6947,
      "eval_samples_per_second": 57.827,
      "eval_steps_per_second": 1.18,
      "step": 392
    },
    {
      "epoch": 57.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.7084951400756836,
      "eval_runtime": 1.6087,
      "eval_samples_per_second": 60.919,
      "eval_steps_per_second": 1.243,
      "step": 399
    },
    {
      "epoch": 57.14,
      "learning_rate": 8.571428571428571e-06,
      "loss": 3.1174,
      "step": 400
    },
    {
      "epoch": 58.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.705909013748169,
      "eval_runtime": 1.7081,
      "eval_samples_per_second": 57.372,
      "eval_steps_per_second": 1.171,
      "step": 406
    },
    {
      "epoch": 58.57,
      "learning_rate": 8.285714285714287e-06,
      "loss": 3.0962,
      "step": 410
    },
    {
      "epoch": 59.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.7031126022338867,
      "eval_runtime": 1.5815,
      "eval_samples_per_second": 61.967,
      "eval_steps_per_second": 1.265,
      "step": 413
    },
    {
      "epoch": 60.0,
      "learning_rate": 8.000000000000001e-06,
      "loss": 3.1237,
      "step": 420
    },
    {
      "epoch": 60.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.701946258544922,
      "eval_runtime": 1.7026,
      "eval_samples_per_second": 57.559,
      "eval_steps_per_second": 1.175,
      "step": 420
    },
    {
      "epoch": 61.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6996471881866455,
      "eval_runtime": 1.6975,
      "eval_samples_per_second": 57.732,
      "eval_steps_per_second": 1.178,
      "step": 427
    },
    {
      "epoch": 61.43,
      "learning_rate": 7.714285714285716e-06,
      "loss": 3.1229,
      "step": 430
    },
    {
      "epoch": 62.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6955974102020264,
      "eval_runtime": 1.6826,
      "eval_samples_per_second": 58.242,
      "eval_steps_per_second": 1.189,
      "step": 434
    },
    {
      "epoch": 62.86,
      "learning_rate": 7.428571428571429e-06,
      "loss": 3.0946,
      "step": 440
    },
    {
      "epoch": 63.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.692981719970703,
      "eval_runtime": 1.7683,
      "eval_samples_per_second": 55.42,
      "eval_steps_per_second": 1.131,
      "step": 441
    },
    {
      "epoch": 64.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6915957927703857,
      "eval_runtime": 1.6752,
      "eval_samples_per_second": 58.501,
      "eval_steps_per_second": 1.194,
      "step": 448
    },
    {
      "epoch": 64.29,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 3.0861,
      "step": 450
    },
    {
      "epoch": 65.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6893138885498047,
      "eval_runtime": 1.6711,
      "eval_samples_per_second": 58.644,
      "eval_steps_per_second": 1.197,
      "step": 455
    },
    {
      "epoch": 65.71,
      "learning_rate": 6.857142857142858e-06,
      "loss": 3.0406,
      "step": 460
    },
    {
      "epoch": 66.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6859352588653564,
      "eval_runtime": 1.7238,
      "eval_samples_per_second": 56.85,
      "eval_steps_per_second": 1.16,
      "step": 462
    },
    {
      "epoch": 67.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6839077472686768,
      "eval_runtime": 1.6549,
      "eval_samples_per_second": 59.218,
      "eval_steps_per_second": 1.209,
      "step": 469
    },
    {
      "epoch": 67.14,
      "learning_rate": 6.571428571428572e-06,
      "loss": 3.077,
      "step": 470
    },
    {
      "epoch": 68.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6815552711486816,
      "eval_runtime": 1.673,
      "eval_samples_per_second": 58.578,
      "eval_steps_per_second": 1.195,
      "step": 476
    },
    {
      "epoch": 68.57,
      "learning_rate": 6.285714285714286e-06,
      "loss": 3.0555,
      "step": 480
    },
    {
      "epoch": 69.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.678163766860962,
      "eval_runtime": 1.6518,
      "eval_samples_per_second": 59.329,
      "eval_steps_per_second": 1.211,
      "step": 483
    },
    {
      "epoch": 70.0,
      "learning_rate": 6e-06,
      "loss": 3.035,
      "step": 490
    },
    {
      "epoch": 70.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6762583255767822,
      "eval_runtime": 1.6591,
      "eval_samples_per_second": 59.069,
      "eval_steps_per_second": 1.205,
      "step": 490
    },
    {
      "epoch": 71.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.672853469848633,
      "eval_runtime": 1.6495,
      "eval_samples_per_second": 59.412,
      "eval_steps_per_second": 1.212,
      "step": 497
    },
    {
      "epoch": 71.43,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 3.0246,
      "step": 500
    },
    {
      "epoch": 72.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6719117164611816,
      "eval_runtime": 1.6901,
      "eval_samples_per_second": 57.986,
      "eval_steps_per_second": 1.183,
      "step": 504
    },
    {
      "epoch": 72.86,
      "learning_rate": 5.428571428571429e-06,
      "loss": 3.0282,
      "step": 510
    },
    {
      "epoch": 73.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.670848846435547,
      "eval_runtime": 1.7293,
      "eval_samples_per_second": 56.669,
      "eval_steps_per_second": 1.157,
      "step": 511
    },
    {
      "epoch": 74.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.6683461666107178,
      "eval_runtime": 1.6959,
      "eval_samples_per_second": 57.785,
      "eval_steps_per_second": 1.179,
      "step": 518
    },
    {
      "epoch": 74.29,
      "learning_rate": 5.142857142857142e-06,
      "loss": 3.0293,
      "step": 520
    },
    {
      "epoch": 75.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.665170669555664,
      "eval_runtime": 1.6438,
      "eval_samples_per_second": 59.617,
      "eval_steps_per_second": 1.217,
      "step": 525
    },
    {
      "epoch": 75.71,
      "learning_rate": 4.857142857142858e-06,
      "loss": 2.9893,
      "step": 530
    },
    {
      "epoch": 76.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.66402268409729,
      "eval_runtime": 1.6852,
      "eval_samples_per_second": 58.155,
      "eval_steps_per_second": 1.187,
      "step": 532
    },
    {
      "epoch": 77.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.6634998321533203,
      "eval_runtime": 1.7537,
      "eval_samples_per_second": 55.882,
      "eval_steps_per_second": 1.14,
      "step": 539
    },
    {
      "epoch": 77.14,
      "learning_rate": 4.571428571428572e-06,
      "loss": 2.9888,
      "step": 540
    },
    {
      "epoch": 78.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.6618170738220215,
      "eval_runtime": 1.7774,
      "eval_samples_per_second": 55.135,
      "eval_steps_per_second": 1.125,
      "step": 546
    },
    {
      "epoch": 78.57,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 2.9833,
      "step": 550
    },
    {
      "epoch": 79.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.659451484680176,
      "eval_runtime": 1.6474,
      "eval_samples_per_second": 59.487,
      "eval_steps_per_second": 1.214,
      "step": 553
    },
    {
      "epoch": 80.0,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.9739,
      "step": 560
    },
    {
      "epoch": 80.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.6578376293182373,
      "eval_runtime": 1.6531,
      "eval_samples_per_second": 59.283,
      "eval_steps_per_second": 1.21,
      "step": 560
    },
    {
      "epoch": 81.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.656160593032837,
      "eval_runtime": 1.7409,
      "eval_samples_per_second": 56.294,
      "eval_steps_per_second": 1.149,
      "step": 567
    },
    {
      "epoch": 81.43,
      "learning_rate": 3.7142857142857146e-06,
      "loss": 2.9513,
      "step": 570
    },
    {
      "epoch": 82.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.655242681503296,
      "eval_runtime": 1.6811,
      "eval_samples_per_second": 58.294,
      "eval_steps_per_second": 1.19,
      "step": 574
    },
    {
      "epoch": 82.86,
      "learning_rate": 3.428571428571429e-06,
      "loss": 2.9503,
      "step": 580
    },
    {
      "epoch": 83.0,
      "eval_accuracy": 0.14285714285714285,
      "eval_loss": 3.6539218425750732,
      "eval_runtime": 1.7574,
      "eval_samples_per_second": 55.764,
      "eval_steps_per_second": 1.138,
      "step": 581
    },
    {
      "epoch": 84.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.653193950653076,
      "eval_runtime": 1.6302,
      "eval_samples_per_second": 60.114,
      "eval_steps_per_second": 1.227,
      "step": 588
    },
    {
      "epoch": 84.29,
      "learning_rate": 3.142857142857143e-06,
      "loss": 2.9792,
      "step": 590
    },
    {
      "epoch": 85.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.651691436767578,
      "eval_runtime": 1.7277,
      "eval_samples_per_second": 56.724,
      "eval_steps_per_second": 1.158,
      "step": 595
    },
    {
      "epoch": 85.71,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 2.9561,
      "step": 600
    },
    {
      "epoch": 86.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6496691703796387,
      "eval_runtime": 1.6772,
      "eval_samples_per_second": 58.432,
      "eval_steps_per_second": 1.192,
      "step": 602
    },
    {
      "epoch": 87.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6485908031463623,
      "eval_runtime": 1.664,
      "eval_samples_per_second": 58.893,
      "eval_steps_per_second": 1.202,
      "step": 609
    },
    {
      "epoch": 87.14,
      "learning_rate": 2.571428571428571e-06,
      "loss": 2.964,
      "step": 610
    },
    {
      "epoch": 88.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.647573709487915,
      "eval_runtime": 1.7778,
      "eval_samples_per_second": 55.124,
      "eval_steps_per_second": 1.125,
      "step": 616
    },
    {
      "epoch": 88.57,
      "learning_rate": 2.285714285714286e-06,
      "loss": 2.9665,
      "step": 620
    },
    {
      "epoch": 89.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6470184326171875,
      "eval_runtime": 1.6689,
      "eval_samples_per_second": 58.722,
      "eval_steps_per_second": 1.198,
      "step": 623
    },
    {
      "epoch": 90.0,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.9439,
      "step": 630
    },
    {
      "epoch": 90.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.646164894104004,
      "eval_runtime": 2.2025,
      "eval_samples_per_second": 44.495,
      "eval_steps_per_second": 0.908,
      "step": 630
    },
    {
      "epoch": 91.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6452953815460205,
      "eval_runtime": 1.6826,
      "eval_samples_per_second": 58.242,
      "eval_steps_per_second": 1.189,
      "step": 637
    },
    {
      "epoch": 91.43,
      "learning_rate": 1.7142857142857145e-06,
      "loss": 2.9369,
      "step": 640
    },
    {
      "epoch": 92.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.645128011703491,
      "eval_runtime": 1.646,
      "eval_samples_per_second": 59.539,
      "eval_steps_per_second": 1.215,
      "step": 644
    },
    {
      "epoch": 92.86,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 2.9619,
      "step": 650
    },
    {
      "epoch": 93.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.64505934715271,
      "eval_runtime": 1.7663,
      "eval_samples_per_second": 55.483,
      "eval_steps_per_second": 1.132,
      "step": 651
    },
    {
      "epoch": 94.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6448814868927,
      "eval_runtime": 1.6466,
      "eval_samples_per_second": 59.516,
      "eval_steps_per_second": 1.215,
      "step": 658
    },
    {
      "epoch": 94.29,
      "learning_rate": 1.142857142857143e-06,
      "loss": 2.955,
      "step": 660
    },
    {
      "epoch": 95.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6444060802459717,
      "eval_runtime": 1.6863,
      "eval_samples_per_second": 58.115,
      "eval_steps_per_second": 1.186,
      "step": 665
    },
    {
      "epoch": 95.71,
      "learning_rate": 8.571428571428572e-07,
      "loss": 2.9323,
      "step": 670
    },
    {
      "epoch": 96.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6440815925598145,
      "eval_runtime": 1.626,
      "eval_samples_per_second": 60.271,
      "eval_steps_per_second": 1.23,
      "step": 672
    },
    {
      "epoch": 97.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6438138484954834,
      "eval_runtime": 1.7724,
      "eval_samples_per_second": 55.292,
      "eval_steps_per_second": 1.128,
      "step": 679
    },
    {
      "epoch": 97.14,
      "learning_rate": 5.714285714285715e-07,
      "loss": 2.9466,
      "step": 680
    },
    {
      "epoch": 98.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.643672466278076,
      "eval_runtime": 1.7638,
      "eval_samples_per_second": 55.56,
      "eval_steps_per_second": 1.134,
      "step": 686
    },
    {
      "epoch": 98.57,
      "learning_rate": 2.8571428571428575e-07,
      "loss": 2.945,
      "step": 690
    },
    {
      "epoch": 99.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.6435706615448,
      "eval_runtime": 1.6937,
      "eval_samples_per_second": 57.862,
      "eval_steps_per_second": 1.181,
      "step": 693
    },
    {
      "epoch": 100.0,
      "learning_rate": 0.0,
      "loss": 2.9665,
      "step": 700
    },
    {
      "epoch": 100.0,
      "eval_accuracy": 0.15306122448979592,
      "eval_loss": 3.643535614013672,
      "eval_runtime": 1.66,
      "eval_samples_per_second": 59.037,
      "eval_steps_per_second": 1.205,
      "step": 700
    },
    {
      "epoch": 100.0,
      "step": 700,
      "total_flos": 4.2794747466153984e+18,
      "train_loss": 3.289549721309117,
      "train_runtime": 4059.8356,
      "train_samples_per_second": 13.597,
      "train_steps_per_second": 0.172
    }
  ],
  "logging_steps": 10,
  "max_steps": 700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 500,
  "total_flos": 4.2794747466153984e+18,
  "train_batch_size": 80,
  "trial_name": null,
  "trial_params": null
}