save code

services/asr.py

@@ -15,25 +15,36 @@ class ASRService:
         self._ready = False
         self._attempted = False
         self._init_error: Optional[str] = None
+        self._device: str = "uninitialized"
 
     def _ensure_loaded(self) -> None:
         if self._ready or self._attempted:
             return
         self._attempted = True
-        try:
-            from funasr import AutoModel
-            from funasr.utils.postprocess_utils import rich_transcription_postprocess
+        from funasr import AutoModel
+        from funasr.utils.postprocess_utils import rich_transcription_postprocess
 
-            self._model = AutoModel(
-                model="FunAudioLLM/SenseVoiceSmall",
-                device="cuda:0",
-                hub="hf",
-            )
-            self._post = rich_transcription_postprocess
-            self._ready = True
-        except Exception as exc:  # pragma: no cover
-            self._init_error = str(exc)
-            logger.warning("ASR fallback mode: %s", exc)
+        errors: list[str] = []
+        for device in ("cuda:0", "cpu"):
+            try:
+                self._model = AutoModel(
+                    model="FunAudioLLM/SenseVoiceSmall",
+                    device=device,
+                    hub="hf",
+                )
+                self._post = rich_transcription_postprocess
+                self._ready = True
+                self._init_error = None
+                self._device = device
+                if device != "cuda:0":
+                    logger.warning("ASR running in degraded mode on %s", device)
+                return
+            except Exception as exc:  # pragma: no cover
+                errors.append(f"{device}: {exc}")
+
+        self._init_error = " | ".join(errors)
+        self._device = "unavailable"
+        logger.warning("ASR unavailable: %s", self._init_error)
 
     async def transcribe(self, audio_16k: np.ndarray) -> str:
         self._ensure_loaded()
@@ -46,9 +57,9 @@ class ASRService:
             )
             return self._post(res[0]["text"])
 
-        # Fallback for initial bring-up
-        return "（ASR降级）语音已接收"
+        # When ASR is unavailable, avoid emitting fake user text that would trigger a bogus LLM reply.
+        return ""
 
     @property
     def health(self) -> dict:
-        return {"ready": self._ready, "attempted": self._attempted, "error": self._init_error}
+        return {"ready": self._ready, "attempted": self._attempted, "device": self._device, "error": self._init_error}