all_results.json

{
    "epoch": 0.9997382884061764,
    "eval_logits/chosen": -0.34449103474617004,
    "eval_logits/rejected": -0.3073786199092865,
    "eval_logps/chosen": -293.0410461425781,
    "eval_logps/ref_response": -0.5363935232162476,
    "eval_logps/rejected": -285.7321472167969,
    "eval_loss": 0.5171638131141663,
    "eval_rewards/accuracies": 0.7300000190734863,
    "eval_rewards/chosen": -0.04938787221908569,
    "eval_rewards/margins": 0.8561664819717407,
    "eval_rewards/rejected": -0.9055544137954712,
    "eval_runtime": 348.587,
    "eval_samples": 2000,
    "eval_samples_per_second": 5.737,
    "eval_steps_per_second": 0.359,
    "total_flos": 0.0,
    "train_loss": 0.5383632463934533,
    "train_runtime": 19109.9128,
    "train_samples": 61135,
    "train_samples_per_second": 3.199,
    "train_steps_per_second": 0.05
}