all_results.json

{
    "epoch": 2.0,
    "eval_logits/chosen": -2.388157844543457,
    "eval_logits/rejected": -2.1907498836517334,
    "eval_logps/chosen": -288.2511291503906,
    "eval_logps/rejected": -264.61138916015625,
    "eval_loss": 0.011413076892495155,
    "eval_rewards/accuracies": 0.39399999380111694,
    "eval_rewards/chosen": -0.08743663877248764,
    "eval_rewards/margins": -0.022926397621631622,
    "eval_rewards/rejected": -0.06451024115085602,
    "eval_runtime": 1422.7975,
    "eval_samples": 2000,
    "eval_samples_per_second": 1.406,
    "eval_steps_per_second": 0.703,
    "train_loss": 0.014390302914878392,
    "train_runtime": 11750.1681,
    "train_samples": 61135,
    "train_samples_per_second": 0.112,
    "train_steps_per_second": 0.056
}