run-2/checkpoint-1070/trainer_state.json

{
  "best_metric": 0.4190778231127276,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-2/checkpoint-1070",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1070,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.93,
      "learning_rate": 4.237866544899179e-06,
      "loss": 0.5637,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5175270438194275,
      "eval_matthews_correlation": 0.40674246350783416,
      "eval_runtime": 0.8013,
      "eval_samples_per_second": 1301.609,
      "eval_steps_per_second": 82.365,
      "step": 535
    },
    {
      "epoch": 1.87,
      "learning_rate": 2.9458340616982097e-06,
      "loss": 0.4571,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5031628012657166,
      "eval_matthews_correlation": 0.4190778231127276,
      "eval_runtime": 1.2207,
      "eval_samples_per_second": 854.397,
      "eval_steps_per_second": 54.065,
      "step": 1070
    }
  ],
  "logging_steps": 500,
  "max_steps": 2140,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 85713651241296.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 5.529899028100149e-06,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 16,
    "seed": 8
  }
}