eval_results.json

{
    "epoch": 2.9865871833084947,
    "eval_log_odds_chosen": 0.28559842705726624,
    "eval_log_odds_ratio": -0.6970076560974121,
    "eval_logits/chosen": 297.1682434082031,
    "eval_logits/rejected": 268.0281982421875,
    "eval_logps/chosen": -1.1085351705551147,
    "eval_logps/rejected": -1.2919707298278809,
    "eval_loss": 1.5517091751098633,
    "eval_nll_loss": 1.4855411052703857,
    "eval_rewards/accuracies": 0.5611510872840881,
    "eval_rewards/chosen": -0.055426761507987976,
    "eval_rewards/margins": 0.009171773679554462,
    "eval_rewards/rejected": -0.06459853798151016,
    "eval_runtime": 112.1993,
    "eval_samples": 553,
    "eval_samples_per_second": 4.929,
    "eval_steps_per_second": 1.239
}