{
    "epoch": 3.0,
    "eval_logits/chosen": -0.17340299487113953,
    "eval_logits/rejected": -0.31969568133354187,
    "eval_logps/chosen": -359.973876953125,
    "eval_logps/rejected": -288.69586181640625,
    "eval_loss": 0.6869306564331055,
    "eval_rewards/accuracies": 0.5833333134651184,
    "eval_rewards/chosen": 0.028966180980205536,
    "eval_rewards/margins": 0.01931903511285782,
    "eval_rewards/rejected": 0.009647144004702568,
    "eval_runtime": 155.4392,
    "eval_samples": 2000,
    "eval_samples_per_second": 12.867,
    "eval_steps_per_second": 0.405
}