trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 100,
  "global_step": 5,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8,
      "learning_rate": 5e-07,
      "logits/chosen": -2.3972699642181396,
      "logits/rejected": -2.39332914352417,
      "logps/chosen": -153.26783752441406,
      "logps/rejected": -146.77935791015625,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.8,
      "eval_logits/chosen": -2.1716835498809814,
      "eval_logits/rejected": -2.2157046794891357,
      "eval_logps/chosen": -103.76641845703125,
      "eval_logps/rejected": -101.35842895507812,
      "eval_loss": 0.6931473016738892,
      "eval_rewards/accuracies": 0.0,
      "eval_rewards/chosen": 0.0,
      "eval_rewards/margins": 0.0,
      "eval_rewards/rejected": 0.0,
      "eval_runtime": 6.5465,
      "eval_samples_per_second": 4.583,
      "eval_steps_per_second": 0.153,
      "step": 1
    },
    {
      "epoch": 1.6,
      "eval_logits/chosen": -2.1715452671051025,
      "eval_logits/rejected": -2.2146592140197754,
      "eval_logps/chosen": -103.4699935913086,
      "eval_logps/rejected": -101.27898406982422,
      "eval_loss": 0.694814920425415,
      "eval_rewards/accuracies": 0.5,
      "eval_rewards/chosen": 0.02964324876666069,
      "eval_rewards/margins": 0.021698763594031334,
      "eval_rewards/rejected": 0.007944487035274506,
      "eval_runtime": 6.0506,
      "eval_samples_per_second": 4.958,
      "eval_steps_per_second": 0.165,
      "step": 2
    },
    {
      "epoch": 2.4,
      "eval_logits/chosen": -2.170850992202759,
      "eval_logits/rejected": -2.215348482131958,
      "eval_logps/chosen": -103.48912811279297,
      "eval_logps/rejected": -101.26887512207031,
      "eval_loss": 0.6913403868675232,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": 0.027730178087949753,
      "eval_rewards/margins": 0.018774602562189102,
      "eval_rewards/rejected": 0.00895557552576065,
      "eval_runtime": 6.0481,
      "eval_samples_per_second": 4.96,
      "eval_steps_per_second": 0.165,
      "step": 3
    },
    {
      "epoch": 4.0,
      "eval_logits/chosen": -2.1724095344543457,
      "eval_logits/rejected": -2.2160496711730957,
      "eval_logps/chosen": -102.96302795410156,
      "eval_logps/rejected": -101.06044006347656,
      "eval_loss": 0.6873850226402283,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.08033924549818039,
      "eval_rewards/margins": 0.05054035410284996,
      "eval_rewards/rejected": 0.02979888767004013,
      "eval_runtime": 6.0751,
      "eval_samples_per_second": 4.938,
      "eval_steps_per_second": 0.165,
      "step": 5
    },
    {
      "epoch": 4.0,
      "step": 5,
      "total_flos": 0.0,
      "train_loss": 0.6922631859779358,
      "train_runtime": 883.8621,
      "train_samples_per_second": 3.541,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 10,
  "max_steps": 5,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}