{
  "best_metric": 0.0,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-1/checkpoint-2138",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 4276,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "learning_rate": 6.433717444611503e-05,
      "loss": 0.628,
      "step": 500
    },
    {
      "epoch": 0.47,
      "learning_rate": 5.581795113492395e-05,
      "loss": 0.6274,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.729872782373287e-05,
      "loss": 0.6194,
      "step": 1500
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.8779504512541786e-05,
      "loss": 0.6078,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.6168875098228455,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7747,
      "eval_samples_per_second": 1346.33,
      "eval_steps_per_second": 85.194,
      "step": 2138
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.0260281201350713e-05,
      "loss": 0.6329,
      "step": 2500
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.174105789015963e-05,
      "loss": 0.6121,
      "step": 3000
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.3221834578968553e-05,
      "loss": 0.6075,
      "step": 3500
    },
    {
      "epoch": 1.87,
      "learning_rate": 4.7026112677774754e-06,
      "loss": 0.6006,
      "step": 4000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.6148878335952759,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7365,
      "eval_samples_per_second": 1416.138,
      "eval_steps_per_second": 89.612,
      "step": 4276
    }
  ],
  "logging_steps": 500,
  "max_steps": 4276,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 65322209786076.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 7.28563977573061e-05,
    "num_train_epochs": 2,
    "per_device_train_batch_size": 4,
    "seed": 23
  }
}