trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": null,
  "eval_steps": 500,
  "global_step": 0,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "eval_accuracy": 0.04516129032258064,
      "eval_loss": 2.444662094116211,
      "eval_runtime": 31.3777,
      "eval_samples_per_second": 4.94,
      "eval_steps_per_second": 2.486,
      "step": 0
    },
    {
      "eval_accuracy": 0.04516129032258064,
      "eval_loss": 2.444662094116211,
      "eval_runtime": 27.1309,
      "eval_samples_per_second": 5.713,
      "eval_steps_per_second": 2.875,
      "step": 0
    }
  ],
  "logging_steps": 500,
  "max_steps": 0,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 0,
  "save_steps": 500,
  "total_flos": 0,
  "train_batch_size": null,
  "trial_name": null,
  "trial_params": null
}