{
    "epoch": 1.0,
    "eval_logits/chosen": -23.520809173583984,
    "eval_logits/rejected": -23.798635482788086,
    "eval_logps/chosen": -0.5238760113716125,
    "eval_logps/rejected": -0.7024902105331421,
    "eval_loss": 3.6378085613250732,
    "eval_rewards/accuracies": 1.0,
    "eval_rewards/chosen": -5.238759994506836,
    "eval_rewards/margins": 1.786142349243164,
    "eval_rewards/rejected": -7.02490234375,
    "eval_runtime": 1.2939,
    "eval_samples": 100,
    "eval_samples_per_second": 77.284,
    "eval_steps_per_second": 0.773
}