all_results.json

{
    "epoch": 0.9978947368421053,
    "eval_logits/chosen": -2.123624324798584,
    "eval_logits/rejected": -1.5094114542007446,
    "eval_logps/chosen": -344.99029541015625,
    "eval_logps/rejected": -369.51580810546875,
    "eval_loss": 0.09113045036792755,
    "eval_rewards/accuracies": 1.0,
    "eval_rewards/chosen": 1.1668342351913452,
    "eval_rewards/margins": 2.6867592334747314,
    "eval_rewards/rejected": -1.5199248790740967,
    "eval_runtime": 33.1453,
    "eval_samples_per_second": 1.509,
    "eval_steps_per_second": 0.211
}