all_results.json

{
    "epoch": 0.9989701338825953,
    "eval_logits/chosen": -2.072295904159546,
    "eval_logits/rejected": -1.9723700284957886,
    "eval_logps/chosen": -326.1224060058594,
    "eval_logps/rejected": -410.1681213378906,
    "eval_loss": 0.530189573764801,
    "eval_rewards/accuracies": 0.7421875,
    "eval_rewards/chosen": -1.0375562906265259,
    "eval_rewards/margins": 0.7862817049026489,
    "eval_rewards/rejected": -1.8238379955291748,
    "eval_runtime": 67.7016,
    "eval_samples_per_second": 14.771,
    "eval_steps_per_second": 0.236
}