diff --git a/src/voxcpm/model/voxcpm.py b/src/voxcpm/model/voxcpm.py
index 3af0af9..0d8f1c2 100644
--- a/src/voxcpm/model/voxcpm.py
+++ b/src/voxcpm/model/voxcpm.py
@@ -283,8 +283,10 @@ class VoxCPMModel(nn.Module):
                 else:
                     break
             else:
-                break
-        return self.audio_vae.decode(latent_pred.to(torch.float32)).squeeze(1).cpu()
+                
+        decode_audio = self.audio_vae.decode(latent_pred.to(torch.float32)).squeeze(1).cpu()  
+        decode_audio = decode_audio[..., 640:-640] # trick: trim the start and end of the audio
+        return decode_audio
     
     @torch.inference_mode()
     def build_prompt_cache(
@@ -468,7 +470,8 @@ class VoxCPMModel(nn.Module):
             else:
                 break
         decode_audio = self.audio_vae.decode(latent_pred.to(torch.float32)).squeeze(1).cpu()
-            
+        decode_audio = decode_audio[..., 640:-640] # trick: trim the start and end of the audio
+
         return (
             decode_audio,
             target_text_token,
@@ -580,7 +583,6 @@ class VoxCPMModel(nn.Module):
         pred_feat_seq = torch.cat(pred_feat_seq, dim=1)  # b, t, p, d
 
         feat_pred = rearrange(pred_feat_seq, "b t p d -> b d (t p)", b=B, p=self.patch_size)
-        feat_pred = feat_pred[..., 1:-1] # trick: remove the first and last token
         return feat_pred, pred_feat_seq.squeeze(0).cpu()
 
     @classmethod