{
    "epoch": 1.0,
    "eval_logits/chosen": -2.5093116760253906,
    "eval_logits/rejected": -2.4554219245910645,
    "eval_logps/chosen": -253.41494750976562,
    "eval_logps/rejected": -298.073486328125,
    "eval_loss": 0.5370525121688843,
    "eval_rewards/accuracies": 0.7264150977134705,
    "eval_rewards/chosen": -0.9334509372711182,
    "eval_rewards/margins": 0.7120789289474487,
    "eval_rewards/rejected": -1.6455297470092773,
    "eval_runtime": 287.3881,
    "eval_samples": 418,
    "eval_samples_per_second": 1.454,
    "eval_steps_per_second": 0.184,
    "train_loss": 0.5672297686573089,
    "train_runtime": 12946.9794,
    "train_samples": 8130,
    "train_samples_per_second": 0.628,
    "train_steps_per_second": 0.078
}