delay gemini multimodal live websocket connect

examples/foundational/07d-interruptible-elevenlabs.py

@@ -48,6 +48,7 @@ async def main():
         tts = ElevenLabsTTSService(
             api_key=os.getenv("ELEVENLABS_API_KEY", ""),
             voice_id=os.getenv("ELEVENLABS_VOICE_ID", ""),
+            model="eleven_flash_v2_5",
         )
 
         llm = OpenAILLMService(api_key=os.getenv("OPENAI_API_KEY"), model="gpt-4o")
@@ -79,7 +80,7 @@ async def main():
                 allow_interruptions=True,
                 enable_metrics=True,
                 enable_usage_metrics=True,
-                report_only_initial_ttfb=True,
+                # report_only_initial_ttfb=True,
             ),
         )
 

src/pipecat/services/gemini_multimodal_live/gemini.py

@@ -177,6 +177,7 @@ class GeminiMultimodalLiveLLMService(LLMService):
         self._receive_task = None
         self._context = None
 
+        self._connected = False
         self._disconnecting = False
         self._api_session_ready = False
         self._run_llm_when_api_session_ready = False
@@ -230,7 +231,6 @@ class GeminiMultimodalLiveLLMService(LLMService):
 
     async def start(self, frame: StartFrame):
         await super().start(frame)
-        await self._connect()
 
     async def stop(self, frame: EndFrame):
         await super().stop(frame)
@@ -434,8 +434,9 @@ class GeminiMultimodalLiveLLMService(LLMService):
     async def _ws_send(self, message):
         # logger.debug(f"Sending message to websocket: {message}")
         try:
-            if self._websocket:
-                await self._websocket.send(json.dumps(message))
+            if not self._websocket:
+                await self._connect()
+            await self._websocket.send(json.dumps(message))
         except Exception as e:
             if self._disconnecting:
                 return