eval_results.json

{
    "epoch": 0.9880609304240429,
    "eval_logits/chosen": 9.16471004486084,
    "eval_logits/rejected": 10.012032508850098,
    "eval_logps/chosen": -84.95478057861328,
    "eval_logps/ref_chosen": -106.16099548339844,
    "eval_logps/ref_rejected": -119.78473663330078,
    "eval_logps/rejected": -98.79496765136719,
    "eval_loss": 0.5348472595214844,
    "eval_rewards/accuracies": 0.5166666507720947,
    "eval_rewards/chosen": 0.2120620608329773,
    "eval_rewards/margins": 0.002164395758882165,
    "eval_rewards/rejected": 0.20989766716957092,
    "eval_runtime": 166.1439,
    "eval_samples": 1080,
    "eval_samples_per_second": 6.5,
    "eval_sft_loss": 0.2975790202617645,
    "eval_steps_per_second": 1.625
}