{
    "epoch": 1.0,
    "eval_logits/chosen": -2.484065532684326,
    "eval_logits/rejected": -2.4274542331695557,
    "eval_logps/chosen": -305.43212890625,
    "eval_logps/rejected": -294.2740783691406,
    "eval_loss": 0.5714402198791504,
    "eval_rewards/accuracies": 0.761904776096344,
    "eval_rewards/chosen": -0.4765053391456604,
    "eval_rewards/margins": 1.3102855682373047,
    "eval_rewards/rejected": -1.7867907285690308,
    "eval_runtime": 613.7286,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.259,
    "eval_steps_per_second": 0.103
}