{
    "epoch": 4.0,
    "eval_logits/chosen": -2.9570438861846924,
    "eval_logits/rejected": -2.951200246810913,
    "eval_logps/chosen": -71.36335754394531,
    "eval_logps/rejected": -78.77714538574219,
    "eval_loss": 0.6802749633789062,
    "eval_rewards/accuracies": 0.6036245226860046,
    "eval_rewards/chosen": -0.12651462852954865,
    "eval_rewards/margins": 0.02945556864142418,
    "eval_rewards/rejected": -0.15597018599510193,
    "eval_runtime": 383.8578,
    "eval_samples": 4304,
    "eval_samples_per_second": 11.212,
    "eval_steps_per_second": 1.402,
    "total_flos": 0.0,
    "train_loss": 0.6773310824150385,
    "train_runtime": 111666.1731,
    "train_samples": 92858,
    "train_samples_per_second": 3.326,
    "train_steps_per_second": 0.052
}