run-1/checkpoint-1605/trainer_state.json

{
  "best_metric": 0.3731107016234198,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-1/checkpoint-1605",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1605,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.93,
      "grad_norm": 4.029795169830322,
      "learning_rate": 1.7163903856453514e-06,
      "loss": 0.6033,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5766915082931519,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7097,
      "eval_samples_per_second": 1469.555,
      "eval_steps_per_second": 92.992,
      "step": 535
    },
    {
      "epoch": 1.87,
      "grad_norm": 4.412518501281738,
      "learning_rate": 1.1931006339242075e-06,
      "loss": 0.5433,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5463550686836243,
      "eval_matthews_correlation": 0.2835704013245166,
      "eval_runtime": 1.1159,
      "eval_samples_per_second": 934.633,
      "eval_steps_per_second": 59.143,
      "step": 1070
    },
    {
      "epoch": 2.8,
      "grad_norm": 4.607611656188965,
      "learning_rate": 6.69810882203064e-07,
      "loss": 0.491,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.5242806077003479,
      "eval_matthews_correlation": 0.3731107016234198,
      "eval_runtime": 0.7889,
      "eval_samples_per_second": 1322.174,
      "eval_steps_per_second": 83.666,
      "step": 1605
    }
  ],
  "logging_steps": 500,
  "max_steps": 2140,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 128596608126132.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 2.2396801373664952e-06,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 16,
    "seed": 33
  }
}