run-1/checkpoint-321/trainer_state.json

{
  "best_metric": 0.5280629029350907,
  "best_model_checkpoint": "deberta-v3-large-finetuned-cola-midterm/run-1/checkpoint-321",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 321,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 0.5697229504585266,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 7.6758,
      "eval_samples_per_second": 135.882,
      "eval_steps_per_second": 8.598,
      "step": 107
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4736369848251343,
      "eval_matthews_correlation": 0.37472354614099185,
      "eval_runtime": 8.2012,
      "eval_samples_per_second": 127.176,
      "eval_steps_per_second": 8.048,
      "step": 214
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.41821202635765076,
      "eval_matthews_correlation": 0.5280629029350907,
      "eval_runtime": 10.1308,
      "eval_samples_per_second": 102.953,
      "eval_steps_per_second": 6.515,
      "step": 321
    }
  ],
  "logging_steps": 500,
  "max_steps": 535,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.2524709009088454e-06,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 8,
    "seed": 39
  }
}