diff --git a/src/llamafactory/train/dpo/trainer.py b/src/llamafactory/train/dpo/trainer.py
index fdc41dd7..7e76dee2 100644
--- a/src/llamafactory/train/dpo/trainer.py
+++ b/src/llamafactory/train/dpo/trainer.py
@@ -255,10 +255,10 @@ class CustomDPOTrainer(DPOTrainer):
         metrics[f"{prefix}rewards/rejected"] = rejected_rewards.mean().item()
         metrics[f"{prefix}rewards/accuracies"] = (chosen_rewards > rejected_rewards).float().mean().item()
         metrics[f"{prefix}rewards/margins"] = (chosen_rewards - rejected_rewards).mean().item()
-        metrics[f"{prefix}logps/rejected"] = policy_chosen_logps.mean().item()
-        metrics[f"{prefix}logps/chosen"] = policy_rejected_logps.mean().item()
-        metrics[f"{prefix}logits/rejected"] = policy_chosen_logits.mean().item()
-        metrics[f"{prefix}logits/chosen"] = policy_rejected_logits.mean().item()
+        metrics[f"{prefix}logps/chosen"] = policy_chosen_logps.mean().item()
+        metrics[f"{prefix}logps/rejected"] = policy_rejected_logps.mean().item()
+        metrics[f"{prefix}logits/chosen"] = policy_chosen_logits.mean().item()
+        metrics[f"{prefix}logits/rejected"] = policy_rejected_logits.mean().item()
         if self.loss_type == "orpo":
             metrics[f"{prefix}sft_loss"] = sft_loss.mean().item()
             metrics[f"{prefix}odds_ratio_loss"] = ((losses - sft_loss) / self.beta).mean().item()