all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -1.9525097608566284,
    "eval_logits/rejected": -1.7713532447814941,
    "eval_logps/chosen": -224.6856231689453,
    "eval_logps/rejected": -193.18142700195312,
    "eval_loss": 4322.0576171875,
    "eval_rewards/accuracies": 0.6033397912979126,
    "eval_rewards/chosen": -0.9425504207611084,
    "eval_rewards/margins": 0.06460745632648468,
    "eval_rewards/rejected": -1.0071579217910767,
    "eval_rewards/safe_rewards": -0.9376580119132996,
    "eval_rewards/unsafe_rewards": -0.9381570816040039,
    "eval_runtime": 993.6614,
    "eval_samples": 33044,
    "eval_samples_per_second": 33.255,
    "eval_steps_per_second": 1.04,
    "train_loss": 11692.422362755651,
    "train_runtime": 23508.9776,
    "train_samples": 59478,
    "train_samples_per_second": 2.53,
    "train_steps_per_second": 0.079
}