{
  "best_metric": 0.89004,
  "best_model_checkpoint": "../../checkpoint/imdb/bert-large/checkpoint-13294",
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 15640,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.88516,
      "eval_loss": 0.2848346531391144,
      "eval_runtime": 108.2274,
      "eval_samples_per_second": 230.995,
      "eval_steps_per_second": 0.906,
      "step": 782
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.680306905370844e-05,
      "loss": 0.3133,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.88884,
      "eval_loss": 0.30378255248069763,
      "eval_runtime": 107.9566,
      "eval_samples_per_second": 231.574,
      "eval_steps_per_second": 0.908,
      "step": 1564
    },
    {
      "epoch": 2.56,
      "learning_rate": 4.360613810741688e-05,
      "loss": 0.1751,
      "step": 2000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.87908,
      "eval_loss": 0.5034909248352051,
      "eval_runtime": 108.1984,
      "eval_samples_per_second": 231.057,
      "eval_steps_per_second": 0.906,
      "step": 2346
    },
    {
      "epoch": 3.84,
      "learning_rate": 4.040920716112532e-05,
      "loss": 0.1057,
      "step": 3000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.885,
      "eval_loss": 0.5942110419273376,
      "eval_runtime": 108.005,
      "eval_samples_per_second": 231.471,
      "eval_steps_per_second": 0.907,
      "step": 3128
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.87644,
      "eval_loss": 0.5219671726226807,
      "eval_runtime": 107.7329,
      "eval_samples_per_second": 232.055,
      "eval_steps_per_second": 0.91,
      "step": 3910
    },
    {
      "epoch": 5.12,
      "learning_rate": 3.721227621483376e-05,
      "loss": 0.0733,
      "step": 4000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.88228,
      "eval_loss": 0.6980640888214111,
      "eval_runtime": 107.7427,
      "eval_samples_per_second": 232.034,
      "eval_steps_per_second": 0.91,
      "step": 4692
    },
    {
      "epoch": 6.39,
      "learning_rate": 3.40153452685422e-05,
      "loss": 0.0439,
      "step": 5000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.88328,
      "eval_loss": 0.6775364279747009,
      "eval_runtime": 107.8186,
      "eval_samples_per_second": 231.871,
      "eval_steps_per_second": 0.909,
      "step": 5474
    },
    {
      "epoch": 7.67,
      "learning_rate": 3.081841432225064e-05,
      "loss": 0.0371,
      "step": 6000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.88908,
      "eval_loss": 0.6118285059928894,
      "eval_runtime": 108.1846,
      "eval_samples_per_second": 231.086,
      "eval_steps_per_second": 0.906,
      "step": 6256
    },
    {
      "epoch": 8.95,
      "learning_rate": 2.7621483375959077e-05,
      "loss": 0.0277,
      "step": 7000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.88644,
      "eval_loss": 0.7127733826637268,
      "eval_runtime": 108.2656,
      "eval_samples_per_second": 230.914,
      "eval_steps_per_second": 0.905,
      "step": 7038
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8868,
      "eval_loss": 0.7554897665977478,
      "eval_runtime": 108.2355,
      "eval_samples_per_second": 230.978,
      "eval_steps_per_second": 0.905,
      "step": 7820
    },
    {
      "epoch": 10.23,
      "learning_rate": 2.442455242966752e-05,
      "loss": 0.0202,
      "step": 8000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.88876,
      "eval_loss": 0.7617989182472229,
      "eval_runtime": 107.7019,
      "eval_samples_per_second": 232.122,
      "eval_steps_per_second": 0.91,
      "step": 8602
    },
    {
      "epoch": 11.51,
      "learning_rate": 2.122762148337596e-05,
      "loss": 0.0141,
      "step": 9000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.88424,
      "eval_loss": 0.765361487865448,
      "eval_runtime": 108.098,
      "eval_samples_per_second": 231.272,
      "eval_steps_per_second": 0.907,
      "step": 9384
    },
    {
      "epoch": 12.79,
      "learning_rate": 1.80306905370844e-05,
      "loss": 0.0125,
      "step": 10000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.88668,
      "eval_loss": 0.834495484828949,
      "eval_runtime": 108.2866,
      "eval_samples_per_second": 230.869,
      "eval_steps_per_second": 0.905,
      "step": 10166
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8844,
      "eval_loss": 0.8072969317436218,
      "eval_runtime": 107.9547,
      "eval_samples_per_second": 231.579,
      "eval_steps_per_second": 0.908,
      "step": 10948
    },
    {
      "epoch": 14.07,
      "learning_rate": 1.483375959079284e-05,
      "loss": 0.0077,
      "step": 11000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.88872,
      "eval_loss": 0.7046570181846619,
      "eval_runtime": 108.2002,
      "eval_samples_per_second": 231.053,
      "eval_steps_per_second": 0.906,
      "step": 11730
    },
    {
      "epoch": 15.35,
      "learning_rate": 1.163682864450128e-05,
      "loss": 0.0071,
      "step": 12000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.88908,
      "eval_loss": 0.8622098565101624,
      "eval_runtime": 108.34,
      "eval_samples_per_second": 230.755,
      "eval_steps_per_second": 0.905,
      "step": 12512
    },
    {
      "epoch": 16.62,
      "learning_rate": 8.439897698209718e-06,
      "loss": 0.004,
      "step": 13000
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.89004,
      "eval_loss": 0.8655158877372742,
      "eval_runtime": 108.4486,
      "eval_samples_per_second": 230.524,
      "eval_steps_per_second": 0.904,
      "step": 13294
    },
    {
      "epoch": 17.9,
      "learning_rate": 5.242966751918159e-06,
      "loss": 0.0031,
      "step": 14000
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.88984,
      "eval_loss": 0.9096148610115051,
      "eval_runtime": 108.8181,
      "eval_samples_per_second": 229.741,
      "eval_steps_per_second": 0.901,
      "step": 14076
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.88916,
      "eval_loss": 0.945423424243927,
      "eval_runtime": 108.3869,
      "eval_samples_per_second": 230.655,
      "eval_steps_per_second": 0.904,
      "step": 14858
    },
    {
      "epoch": 19.18,
      "learning_rate": 2.0460358056265987e-06,
      "loss": 0.0016,
      "step": 15000
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.88868,
      "eval_loss": 0.9621471762657166,
      "eval_runtime": 108.2519,
      "eval_samples_per_second": 230.943,
      "eval_steps_per_second": 0.905,
      "step": 15640
    },
    {
      "epoch": 20.0,
      "step": 15640,
      "total_flos": 1.16491420416e+17,
      "train_loss": 0.054148883717444245,
      "train_runtime": 8090.5095,
      "train_samples_per_second": 61.801,
      "train_steps_per_second": 1.933
    }
  ],
  "logging_steps": 1000,
  "max_steps": 15640,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 1.16491420416e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}