trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0005162622612287042,
  "eval_steps": 100,
  "global_step": 1,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 0.0,
      "logits/chosen": -2.0846190452575684,
      "logits/rejected": -1.8105568885803223,
      "logps/chosen": -181.74118041992188,
      "logps/rejected": -198.60940551757812,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.0,
      "step": 1,
      "total_flos": 0.0,
      "train_loss": 0.6931471824645996,
      "train_runtime": 5.5573,
      "train_samples_per_second": 5.758,
      "train_steps_per_second": 0.18
    }
  ],
  "logging_steps": 10,
  "max_steps": 1,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}