all_results.json

{
    "epoch": 3.0,
    "eval_logits/chosen": -2.4314932823181152,
    "eval_logits/rejected": -2.403698682785034,
    "eval_logps/chosen": -326.85711669921875,
    "eval_logps/rejected": -393.3514709472656,
    "eval_loss": 0.6730212569236755,
    "eval_rewards/accuracies": 0.7074999809265137,
    "eval_rewards/chosen": -13.26190185546875,
    "eval_rewards/margins": 8.8816556930542,
    "eval_rewards/rejected": -22.143556594848633,
    "eval_runtime": 139.8238,
    "eval_samples": 3156,
    "eval_samples_per_second": 22.571,
    "eval_steps_per_second": 0.358,
    "train_loss": 0.2455477410652717,
    "train_runtime": 50581.1093,
    "train_samples": 82404,
    "train_samples_per_second": 4.887,
    "train_steps_per_second": 0.306
}