{
    "epoch": 0.9990049751243781,
    "eval_logits/chosen": 1.5755633115768433,
    "eval_logits/rejected": 1.1696137189865112,
    "eval_logps/chosen": -600.6971435546875,
    "eval_logps/rejected": -556.9004516601562,
    "eval_loss": 0.3912544250488281,
    "eval_rewards/accuracies": 0.9417910575866699,
    "eval_rewards/chosen": 0.17757247388362885,
    "eval_rewards/margins": 0.851599395275116,
    "eval_rewards/rejected": -0.6740269660949707,
    "eval_runtime": 879.0322,
    "eval_samples": 5359,
    "eval_samples_per_second": 6.096,
    "eval_steps_per_second": 0.381,
    "total_flos": 0.0,
    "train_loss": 0.45811287039621257,
    "train_runtime": 34472.405,
    "train_samples": 48227,
    "train_samples_per_second": 1.399,
    "train_steps_per_second": 0.022
}