trainer_state.json

{
  "best_metric": 0.06354796886444092,
  "best_model_checkpoint": "BioM-ELECTRA-Large-SQuAD2-XBD/checkpoint-303",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 303,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 0.266059547662735,
      "eval_runtime": 34.2632,
      "eval_samples_per_second": 23.582,
      "eval_steps_per_second": 2.948,
      "step": 101
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.09467792510986328,
      "eval_runtime": 34.1709,
      "eval_samples_per_second": 23.646,
      "eval_steps_per_second": 2.956,
      "step": 202
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.06354796886444092,
      "eval_runtime": 34.2043,
      "eval_samples_per_second": 23.623,
      "eval_steps_per_second": 2.953,
      "step": 303
    }
  ],
  "logging_steps": 500,
  "max_steps": 303,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1688389311688704.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}