{
    "epoch": 3.0,
    "eval_logits/chosen": -2.924863338470459,
    "eval_logits/rejected": -2.9190196990966797,
    "eval_logps/chosen": -73.79141235351562,
    "eval_logps/rejected": -81.6350326538086,
    "eval_loss": 0.6785153150558472,
    "eval_rewards/accuracies": 0.6085036993026733,
    "eval_rewards/chosen": -0.15079528093338013,
    "eval_rewards/margins": 0.03375381976366043,
    "eval_rewards/rejected": -0.18454909324645996,
    "eval_runtime": 383.6713,
    "eval_samples": 4304,
    "eval_samples_per_second": 11.218,
    "eval_steps_per_second": 1.402
}