all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -1.099204421043396,
    "eval_logits/rejected": -0.6287172436714172,
    "eval_logps/chosen": -231.0450897216797,
    "eval_logps/rejected": -222.87437438964844,
    "eval_loss": 0.34230488538742065,
    "eval_rewards/accuracies": 0.7313649654388428,
    "eval_rewards/chosen": -1.0060617923736572,
    "eval_rewards/margins": 0.29796916246414185,
    "eval_rewards/rejected": -1.3040310144424438,
    "eval_rewards/safe_rewards": -0.9952626824378967,
    "eval_rewards/unsafe_rewards": -1.0112754106521606,
    "eval_runtime": 1122.5675,
    "eval_samples": 33044,
    "eval_samples_per_second": 29.436,
    "eval_steps_per_second": 0.92,
    "train_loss": 1.018996798697021,
    "train_runtime": 22449.6551,
    "train_samples": 59478,
    "train_samples_per_second": 2.649,
    "train_steps_per_second": 0.083
}