{
    "epoch": 1.0,
    "eval_logits/chosen": -2.243804693222046,
    "eval_logits/rejected": -2.238922119140625,
    "eval_logps/chosen": -34.02475357055664,
    "eval_logps/rejected": -37.51246643066406,
    "eval_loss": 0.6927995681762695,
    "eval_rewards/accuracies": 0.5099667906761169,
    "eval_rewards/chosen": 0.0019583876710385084,
    "eval_rewards/margins": 0.0011271554976701736,
    "eval_rewards/rejected": 0.0008312325226143003,
    "eval_runtime": 145.6648,
    "eval_samples": 343,
    "eval_samples_per_second": 2.355,
    "eval_steps_per_second": 0.295
}