{ "epoch": 1.0, "eval_logits/chosen": -923703389.9449542, "eval_logits/rejected": -831308124.5957447, "eval_logps/chosen": -219.38191657110093, "eval_logps/rejected": -231.15425531914894, "eval_loss": 0.37631505727767944, "eval_rewards/chosen": 0.6432788393913059, "eval_rewards/margins": 1.9445905758025428, "eval_rewards/rejected": -1.3013117364112368, "eval_runtime": 1556.3946, "eval_samples_per_second": 0.643, "eval_steps_per_second": 0.161, "kl": 0.0 }