{
    "epoch": 2.999297541394882,
    "eval_logits/chosen": 0.016507575288414955,
    "eval_logits/rejected": 0.08326081186532974,
    "eval_logps/chosen": -1.3030214309692383,
    "eval_logps/rejected": -1.4992129802703857,
    "eval_loss": 1.6474518775939941,
    "eval_rewards/accuracies": 0.5712166428565979,
    "eval_rewards/chosen": -1.3030214309692383,
    "eval_rewards/margins": 0.19619162380695343,
    "eval_rewards/rejected": -1.4992129802703857,
    "eval_runtime": 39.9919,
    "eval_samples": 1345,
    "eval_samples_per_second": 33.632,
    "eval_steps_per_second": 8.427,
    "total_flos": 0.0,
    "train_loss": 1.6264250374623148,
    "train_runtime": 30148.9193,
    "train_samples": 59790,
    "train_samples_per_second": 5.949,
    "train_steps_per_second": 0.186
}