参考了 https://github.com/FunAudioLLM/SenseVoice/issues/30 , 删除了不必要的字段

api.py

@@ -222,7 +222,7 @@ async def process_audio(audio_data: bytes, language: str = "auto") -> str:
             language=language,
             use_itn=True,
             batch_size_s=500,
-            merge_vad=True
+            # merge_vad=True
         )
         
         # Format result