trainer_state.json

{
  "best_metric": 0.2785862785862786,
  "best_model_checkpoint": "test-hasy-7/checkpoint-541",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 541,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.92,
      "grad_norm": 3.252389669418335,
      "learning_rate": 1.5157116451016638e-06,
      "loss": 4.3305,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2785862785862786,
      "eval_loss": 4.025966167449951,
      "eval_runtime": 2.0515,
      "eval_samples_per_second": 234.458,
      "eval_steps_per_second": 29.734,
      "step": 541
    },
    {
      "epoch": 1.0,
      "step": 541,
      "total_flos": 3.355193271048192e+17,
      "train_loss": 4.313544084756961,
      "train_runtime": 56.6601,
      "train_samples_per_second": 76.332,
      "train_steps_per_second": 9.548
    }
  ],
  "logging_steps": 500,
  "max_steps": 541,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 3.355193271048192e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}