{
    "epoch": 0.9994353472614342,
    "eval_logits/chosen": -2.932816743850708,
    "eval_logits/rejected": -2.902209758758545,
    "eval_logps/chosen": -250.2433319091797,
    "eval_logps/rejected": -229.81752014160156,
    "eval_loss": 0.6546631455421448,
    "eval_rewards/accuracies": 0.591292142868042,
    "eval_rewards/chosen": -0.2775766849517822,
    "eval_rewards/margins": 0.10209956020116806,
    "eval_rewards/rejected": -0.3796762526035309,
    "eval_runtime": 230.8386,
    "eval_samples": 712,
    "eval_samples_per_second": 3.084,
    "eval_steps_per_second": 0.386,
    "total_flos": 0.0,
    "train_loss": 0.6718437324135991,
    "train_runtime": 9969.8232,
    "train_samples": 14167,
    "train_samples_per_second": 1.421,
    "train_steps_per_second": 0.089
}