eval_results.json

{
    "epoch": 3.0,
    "eval_logits/chosen": -2.355067729949951,
    "eval_logits/rejected": -2.3430099487304688,
    "eval_logps/chosen": -164.92103576660156,
    "eval_logps/rejected": -186.04312133789062,
    "eval_loss": 0.6500672101974487,
    "eval_rewards/accuracies": 0.6031598448753357,
    "eval_rewards/chosen": -1.0590553283691406,
    "eval_rewards/margins": 0.1738797128200531,
    "eval_rewards/rejected": -1.2329350709915161,
    "eval_runtime": 384.1905,
    "eval_samples": 4304,
    "eval_samples_per_second": 11.203,
    "eval_steps_per_second": 1.4
}