run-3/checkpoint-536/trainer_state.json

{
  "best_metric": 0.4440192352078645,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-3/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": 0.5107208490371704,
      "eval_matthews_correlation": 0.40032442840601,
      "eval_runtime": 0.7767,
      "eval_samples_per_second": 1342.914,
      "eval_steps_per_second": 84.978,
      "step": 268
    },
    {
      "epoch": 1.87,
      "grad_norm": 6.027254104614258,
      "learning_rate": 6.64979179347967e-07,
      "loss": 0.5057,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5090423226356506,
      "eval_matthews_correlation": 0.4440192352078645,
      "eval_runtime": 0.9,
      "eval_samples_per_second": 1158.903,
      "eval_steps_per_second": 73.334,
      "step": 536
    }
  ],
  "logging_steps": 500,
  "max_steps": 536,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 95617141643160.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 9.900801114736397e-06,
    "num_train_epochs": 2,
    "per_device_train_batch_size": 32,
    "seed": 26
  }
}