eval_results.json

{
    "epoch": 2.9969690846635686,
    "eval_logits/chosen": -14.27784252166748,
    "eval_logits/rejected": -14.317824363708496,
    "eval_logps/chosen": -1.372594952583313,
    "eval_logps/rejected": -1.4290432929992676,
    "eval_loss": 1.4484930038452148,
    "eval_odds_ratio_loss": 0.7589808702468872,
    "eval_rewards/accuracies": 0.4809090793132782,
    "eval_rewards/chosen": -0.13725949823856354,
    "eval_rewards/margins": 0.005644842050969601,
    "eval_rewards/rejected": -0.1429043412208557,
    "eval_runtime": 410.309,
    "eval_samples_per_second": 2.681,
    "eval_sft_loss": 1.372594952583313,
    "eval_steps_per_second": 1.34
}