{
    "epoch": 1.0,
    "eval_logits/chosen": -1.2812360525131226,
    "eval_logits/rejected": -1.175386905670166,
    "eval_logps/chosen": -475.6828918457031,
    "eval_logps/rejected": -535.7244873046875,
    "eval_loss": 0.5020151734352112,
    "eval_rewards/accuracies": 0.7053571343421936,
    "eval_rewards/chosen": -2.1711955070495605,
    "eval_rewards/margins": 0.7908250093460083,
    "eval_rewards/rejected": -2.9620203971862793,
    "eval_runtime": 218.5799,
    "eval_samples": 2000,
    "eval_samples_per_second": 9.15,
    "eval_steps_per_second": 0.384,
    "total_flos": 0.0,
    "train_loss": 0.5397420258873861,
    "train_runtime": 14909.9591,
    "train_samples": 61134,
    "train_samples_per_second": 4.1,
    "train_steps_per_second": 0.043
}