{
    "epoch": 0.9963369963369964,
    "eval_logits/chosen": -2.650862693786621,
    "eval_logits/rejected": -2.649400234222412,
    "eval_logps/chosen": -448.55865478515625,
    "eval_logps/rejected": -538.1439819335938,
    "eval_loss": 0.41933417320251465,
    "eval_rewards/accuracies": 0.71875,
    "eval_rewards/chosen": -0.6056715250015259,
    "eval_rewards/margins": 1.091530680656433,
    "eval_rewards/rejected": -1.697202205657959,
    "eval_runtime": 2.9829,
    "eval_samples": 200,
    "eval_samples_per_second": 67.048,
    "eval_steps_per_second": 0.67
}