{
    "epoch": 0.99,
    "eval_logits/chosen": -3.918135404586792,
    "eval_logits/rejected": -3.7465851306915283,
    "eval_logps/chosen": -233.3091278076172,
    "eval_logps/rejected": -518.7740478515625,
    "eval_loss": 0.5462985038757324,
    "eval_rewards/accuracies": 0.8125,
    "eval_rewards/chosen": -0.5233834981918335,
    "eval_rewards/margins": 0.8745772242546082,
    "eval_rewards/rejected": -1.3979607820510864,
    "eval_runtime": 37.2528,
    "eval_samples": 290,
    "eval_samples_per_second": 7.785,
    "eval_steps_per_second": 0.268,
    "train_loss": 0.49557680382833375,
    "train_runtime": 1348.5444,
    "train_samples": 5827,
    "train_samples_per_second": 4.321,
    "train_steps_per_second": 0.067
}