all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.230694055557251,
    "eval_logits/rejected": -2.225895643234253,
    "eval_logps/chosen": -33.75852966308594,
    "eval_logps/rejected": -37.27712631225586,
    "eval_loss": 0.4946732223033905,
    "eval_rewards/accuracies": 0.5365448594093323,
    "eval_rewards/chosen": 0.16561271250247955,
    "eval_rewards/margins": 0.02191758342087269,
    "eval_rewards/rejected": 0.14369513094425201,
    "eval_runtime": 145.7007,
    "eval_samples": 343,
    "eval_samples_per_second": 2.354,
    "eval_steps_per_second": 0.295,
    "train_loss": 0.46439293204963983,
    "train_runtime": 3253.2697,
    "train_samples": 3079,
    "train_samples_per_second": 0.946,
    "train_steps_per_second": 0.118
}