{
    "epoch": 1.0,
    "eval_logits/chosen": -2.8066177368164062,
    "eval_logits/rejected": -2.768425226211548,
    "eval_logps/chosen": -284.29595947265625,
    "eval_logps/rejected": -259.2527160644531,
    "eval_loss": 0.6889536380767822,
    "eval_rewards/accuracies": 0.675000011920929,
    "eval_rewards/chosen": 0.002974556991830468,
    "eval_rewards/margins": 0.009712548926472664,
    "eval_rewards/margins_max": 0.04481280967593193,
    "eval_rewards/margins_min": -0.02172652818262577,
    "eval_rewards/margins_std": 0.021889150142669678,
    "eval_rewards/rejected": -0.006737993564456701,
    "eval_runtime": 427.88,
    "eval_samples": 2000,
    "eval_samples_per_second": 4.674,
    "eval_steps_per_second": 0.292
}