{
    "epoch": 0.9230769230769231,
    "eval_kl/chosen": 9.097276968197954e-13,
    "eval_kl/rejected": 5.868692127248653e-16,
    "eval_logits/chosen": -1.0625,
    "eval_logits/rejected": -1.015625,
    "eval_logps/chosen": -382.0,
    "eval_logps/rejected": -332.0,
    "eval_loss": 0.6977695226669312,
    "eval_nll_loss": 1.6171875,
    "eval_ref_probs/chosen": 1.2864488120600814e-11,
    "eval_ref_probs/rejected": 1.9181423197158582e-15,
    "eval_rejected_term/max": 5.868692127248653e-16,
    "eval_rejected_term/min": 5.868692127248653e-16,
    "eval_rewards/accuracies": 0.3240000009536743,
    "eval_rewards/chosen": 0.03271484375,
    "eval_rewards/margins": 0.01190185546875,
    "eval_rewards/rejected": 0.0208740234375,
    "eval_runtime": 25.5565,
    "eval_samples_per_second": 39.129,
    "eval_steps_per_second": 9.782
}