run-1/checkpoint-536/trainer_state.json

{
  "best_metric": 0.4959240027089484,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-1/checkpoint-536",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 536,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 0.4628433287143707,
      "eval_matthews_correlation": 0.4621927387299967,
      "eval_runtime": 29.9442,
      "eval_samples_per_second": 34.831,
      "eval_steps_per_second": 2.204,
      "step": 268
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.765374183654785,
      "learning_rate": 2.822226275780235e-06,
      "loss": 0.4202,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4986613690853119,
      "eval_matthews_correlation": 0.4959240027089484,
      "eval_runtime": 31.4713,
      "eval_samples_per_second": 33.141,
      "eval_steps_per_second": 2.097,
      "step": 536
    }
  ],
  "logging_steps": 500,
  "max_steps": 536,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 96028256284692.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 4.201981343939461e-05,
    "num_train_epochs": 2,
    "per_device_train_batch_size": 32,
    "seed": 25
  }
}