eval_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.5884382724761963,
    "eval_logits/rejected": -2.5948920249938965,
    "eval_logps/chosen": -361.0521240234375,
    "eval_logps/rejected": -378.9569091796875,
    "eval_loss": 0.6668171882629395,
    "eval_rewards/accuracies": 0.6136666536331177,
    "eval_rewards/chosen": -0.26719996333122253,
    "eval_rewards/margins": 0.08193866163492203,
    "eval_rewards/rejected": -0.3491385877132416,
    "eval_runtime": 1605.8768,
    "eval_samples": 2994,
    "eval_samples_per_second": 1.864,
    "eval_steps_per_second": 0.234
}