{
    "epoch": 1.0,
    "eval_logits/chosen": -3.164646625518799,
    "eval_logits/rejected": -3.153470039367676,
    "eval_logps/chosen": -543.001708984375,
    "eval_logps/rejected": -567.34228515625,
    "eval_loss": 0.7142251133918762,
    "eval_rewards/accuracies": 0.546875,
    "eval_rewards/chosen": -1.4159148931503296,
    "eval_rewards/margins": 0.08076205104589462,
    "eval_rewards/rejected": -1.496677041053772,
    "eval_runtime": 125.7685,
    "eval_samples": 1000,
    "eval_samples_per_second": 7.951,
    "eval_steps_per_second": 0.254,
    "train_loss": 0.6047722502396657,
    "train_runtime": 5332.3567,
    "train_samples": 19958,
    "train_samples_per_second": 3.75,
    "train_steps_per_second": 0.059
}