{
    "epoch": 3.0,
    "eval_logits/chosen": 0.3730663061141968,
    "eval_logits/rejected": 0.4475269019603729,
    "eval_logps/chosen": -338.3392028808594,
    "eval_logps/rejected": -370.232666015625,
    "eval_loss": 0.6289177536964417,
    "eval_rewards/accuracies": 0.7405063509941101,
    "eval_rewards/chosen": 0.7478683590888977,
    "eval_rewards/margins": 4.585729122161865,
    "eval_rewards/rejected": -3.8378612995147705,
    "eval_runtime": 70.1775,
    "eval_samples": 2500,
    "eval_samples_per_second": 35.624,
    "eval_steps_per_second": 1.126,
    "total_flos": 0.0,
    "train_loss": 0.5009220597491634,
    "train_runtime": 6227.6413,
    "train_samples": 26990,
    "train_samples_per_second": 13.002,
    "train_steps_per_second": 0.102
}