run-17/checkpoint-500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.5723270440251573,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7290322580645161,
      "eval_loss": 3.165098190307617,
      "eval_runtime": 2.394,
      "eval_samples_per_second": 1294.905,
      "eval_steps_per_second": 27.151,
      "step": 318
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.685534591194969e-05,
      "loss": 3.6948,
      "step": 500
    }
  ],
  "logging_steps": 500,
  "max_steps": 3180,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 236585578739496.0,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.3328723302231085,
    "num_train_epochs": 10,
    "temperature": 5
  }
}