{
    "epoch": 0.9965156794425087,
    "eval_logits/chosen": -0.16245996952056885,
    "eval_logits/rejected": 0.3127332031726837,
    "eval_logps/chosen": -328.00042724609375,
    "eval_logps/rejected": -417.0811462402344,
    "eval_loss": 0.5565306544303894,
    "eval_rewards/accuracies": 0.7578125,
    "eval_rewards/chosen": -0.653814435005188,
    "eval_rewards/margins": 0.8903896808624268,
    "eval_rewards/rejected": -1.5442042350769043,
    "eval_runtime": 102.1382,
    "eval_samples": 2000,
    "eval_samples_per_second": 19.581,
    "eval_steps_per_second": 0.313,
    "total_flos": 0.0,
    "train_loss": 0.4296882511018873,
    "train_runtime": 3529.992,
    "train_samples": 18339,
    "train_samples_per_second": 5.195,
    "train_steps_per_second": 0.041
}