{
    "epoch": 1.0,
    "eval_log_ratio_diff/mean": 1.9296875,
    "eval_logits/chosen": -2.015625,
    "eval_logits/rejected": -1.9765625,
    "eval_logps/chosen": -396.0,
    "eval_logps/rejected": -340.0,
    "eval_loss": 0.6354262828826904,
    "eval_nll_loss": 1.7109375,
    "eval_ref_probs/chosen": 8.935757983433244e-15,
    "eval_ref_probs/rejected": 1.2251821244682734e-15,
    "eval_rejected_term/max": 1.0652459139391321e-13,
    "eval_rejected_term/min": 1.0652459139391321e-13,
    "eval_rewards/accuracies": 0.5720000267028809,
    "eval_rewards/chosen": 0.26171875,
    "eval_rewards/margins": 0.193359375,
    "eval_rewards/rejected": 0.06884765625,
    "eval_runtime": 17.1008,
    "eval_samples_per_second": 58.477,
    "eval_steps_per_second": 14.619
}