{
    "epoch": 0.996510067114094,
    "eval_logits/chosen": -3.1117970943450928,
    "eval_logits/rejected": -3.3963212966918945,
    "eval_logps/chosen": -2103.136962890625,
    "eval_logps/rejected": -10148.9853515625,
    "eval_loss": 0.17164821922779083,
    "eval_rewards/accuracies": 0.9180327653884888,
    "eval_rewards/chosen": -2.3896186351776123,
    "eval_rewards/margins": 8.561161994934082,
    "eval_rewards/rejected": -10.950779914855957,
    "eval_runtime": 428.9453,
    "eval_samples": 1951,
    "eval_samples_per_second": 4.548,
    "eval_steps_per_second": 0.569,
    "total_flos": 0.0,
    "train_loss": 0.2600768296497649,
    "train_runtime": 7822.1359,
    "train_samples": 14897,
    "train_samples_per_second": 1.904,
    "train_steps_per_second": 0.015
}