all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -1.583479404449463,
    "eval_logits/rejected": -1.729608416557312,
    "eval_logps/chosen": -202.55133056640625,
    "eval_logps/rejected": -174.8009033203125,
    "eval_loss": 4659.3857421875,
    "eval_rewards/accuracies": 0.6341965198516846,
    "eval_rewards/chosen": -0.22735264897346497,
    "eval_rewards/margins": 0.03718877211213112,
    "eval_rewards/rejected": -0.2645414173603058,
    "eval_rewards/safe_rewards": -0.2253904640674591,
    "eval_rewards/unsafe_rewards": -0.22526215016841888,
    "eval_runtime": 1015.3903,
    "eval_samples": 33044,
    "eval_samples_per_second": 32.543,
    "eval_steps_per_second": 1.017,
    "train_loss": 4602.791315297503,
    "train_runtime": 29489.6101,
    "train_samples": 297394,
    "train_samples_per_second": 10.085,
    "train_steps_per_second": 0.158
}