{ "epoch": 0.9989528795811519, "eval_KL/chosen_KL_mean": -467.01068115234375, "eval_KL/mean": -554.3729858398438, "eval_KL/rejected_KL_mean": -641.7353515625, "eval_KL/std": 329.4678649902344, "eval_fcm_dpo/beta": 0.0010197119554504752, "eval_logits/chosen": -0.7331064939498901, "eval_logits/rejected": -0.7050662040710449, "eval_logps/chosen": -754.8375244140625, "eval_logps/ref_chosen": -287.8268127441406, "eval_logps/ref_rejected": -266.9300231933594, "eval_logps/rejected": -908.6654663085938, "eval_loss": 0.6327955722808838, "eval_margin_dpo/margin_mean": 174.72463989257812, "eval_margin_dpo/margin_std": 379.1971740722656, "eval_runtime": 78.0079, "eval_samples": 2000, "eval_samples_per_second": 25.638, "eval_steps_per_second": 3.205 }