all_results.json

{
    "epoch": 3.0,
    "eval_logits/chosen": 5.65625,
    "eval_logits/rejected": 6.0625,
    "eval_logps/chosen": -63232.0,
    "eval_logps/rejected": -54528.0,
    "eval_loss": 185.16600036621094,
    "eval_rewards/accuracies": 0.435546875,
    "eval_rewards/chosen": -628.0,
    "eval_rewards/margins": -85.5,
    "eval_rewards/rejected": -544.0,
    "eval_runtime": 46.5467,
    "eval_samples": 2000,
    "eval_samples_per_second": 42.968,
    "eval_steps_per_second": 0.687,
    "total_flos": 0.0,
    "train_loss": 0.024607474328222707,
    "train_runtime": 12418.8835,
    "train_samples": 61119,
    "train_samples_per_second": 14.764,
    "train_steps_per_second": 0.231
}