https://livekit.io logo
Join Slack
Powered by
i recently upgraded my livekit-agents version to v...
# ask-ai
r
i recently upgraded my livekit-agents version to v1.2, now I m unable to start my application. Getting this error. What could be the reason and how to fix this?
Copy code
2025-07-26 12:47:06,558 - DEBUG asyncio - Using selector: EpollSelector 
2025-07-26 12:47:06,571 - DEV  livekit.agents - Watching /workspace/LocalVoice 
Created recordings directory at: /workspace/LocalVoice/recordings
2025-07-26 12:47:12,641 - DEBUG asyncio - Using selector: EpollSelector 
2025-07-26 12:47:12,662 - INFO livekit.agents - starting worker {"version": "1.2.2", "rtc-version": "1.0.12"}
2025-07-26 12:47:12,662 - INFO livekit.agents - preloading plugins {"packages": ["livekit.plugins.openai", "livekit.plugins.playai", "livekit.plugins.cartesia", "livekit.plugins.deepgram", "livekit.plugins.elevenlabs", "livekit.plugins.silero", "livekit.plugins.assemblyai", "livekit.plugins.google", "livekit.plugins.turn_detector", "av"]}
2025-07-26 12:47:12,669 - INFO livekit.agents - starting inference executor 
2025-07-26 12:47:19,543 - INFO livekit.agents - initializing process {"pid": 22507, "inference": true}
Created recordings directory at: /workspace/LocalVoice/recordings
2025-07-26 12:47:19,861 - DEBUG livekit.agents - initializing inference runner {"runner": "lk_end_of_utterance_en", "pid": 22507, "inference": true}
2025-07-26 12:47:29,553 - INFO livekit.agents - killing process {"pid": 22507, "inference": true}
2025-07-26 12:47:29,554 - ERROR livekit.agents - worker failed 
Traceback (most recent call last):
  File "/workspace/miniconda3/envs/loc/lib/python3.11/asyncio/tasks.py", line 500, in wait_for
    return fut.result()
           ^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/agents/ipc/channel.py", line 47, in arecv_message
    return _read_message(await dplx.recv_bytes(), messages)
                         ^^^^^^^^^^^^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/agents/utils/aio/duplex_unix.py", line 35, in recv_bytes
    len_bytes = await self._reader.readexactly(4)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/asyncio/streams.py", line 750, in readexactly
    await self._wait_for_data('readexactly')
  File "/workspace/miniconda3/envs/loc/lib/python3.11/asyncio/streams.py", line 543, in _wait_for_data
    await self._waiter
asyncio.exceptions.CancelledError

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/agents/cli/_run.py", line 79, in _worker_run
    await worker.run()
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/agents/worker.py", line 387, in run
    await self._inference_executor.initialize()
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/agents/ipc/supervised_proc.py", line 169, in initialize
    init_res = await asyncio.wait_for(
               ^^^^^^^^^^^^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/asyncio/tasks.py", line 502, in wait_for
    raise exceptions.TimeoutError() from exc
TimeoutError
t
Thanks for reaching out, Taresh. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.
The error you're encountering during the startup of your application with livekit-agents v1.2 is likely due to a timeout during the initialization process. Here are some steps to resolve this issue: • Increase Initialization Timeout: The default timeout for process initialization is 10 seconds. You can increase this value in your configuration. For example:
Copy code
from livekit.agents import WorkerOptions

opts = WorkerOptions(
    initialize_process_timeout=30.0  # Increase to 30 seconds
)
• Check Network Connectivity: Ensure that your application can connect to the LiveKit server and any external APIs it depends on. Network issues can cause timeouts during startup. • Monitor System Resources: Verify that your system has sufficient memory and CPU resources. High resource usage can lead to initialization delays. • Verify LiveKit Server Configuration: Ensure that the required environment variables, such as 
LIVEKIT_URL
, are correctly set and accessible. • Test External API Connections: If your application uses external APIs, test these connections separately to ensure they are responsive and not causing delays. If the issue persists after trying these steps, let me know, and I can assist further. Sources: livekit-agents/livekit/agents/llm/mcp.py
r
okay increasing the timeout fixed it. Now I a facing another issue with EOU model. here is how Im initialising it 
turn_detection=MultilingualModel(),
Copy code
2025-07-26 12:59:41,211 - DEBUG livekit.agents - received user transcript {"user_transcript": "Hey, what's up?", "language": "english", "pid": 32663, "job_id": "AJ_idGDMG4ihsxM"}
2025-07-26 12:59:44,213 - ERROR livekit.agents - Error in _bounce_eou_task
Traceback (most recent call last):
  File "/workspace/miniconda3/envs/loc/lib/python3.11/asyncio/tasks.py", line 500, in wait_for
    return fut.result()
           ^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/agents/ipc/job_proc_lazy_main.py", line 101, in do_inference
    inf_resp = await fut
               ^^^^^^^^^
asyncio.exceptions.CancelledError

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/agents/utils/log.py", line 16, in async_fn_logs
    return await fn(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/agents/voice/audio_recognition.py", line 358, in _bounce_eou_task
    end_of_turn_probability = await turn_detector.predict_end_of_turn(chat_ctx)
                              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/plugins/turn_detector/multilingual.py", line 62, in predict_end_of_turn
    return await super().predict_end_of_turn(chat_ctx, timeout=timeout)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/site-packages/livekit/plugins/turn_detector/base.py", line 230, in predict_end_of_turn
    result = await asyncio.wait_for(
             ^^^^^^^^^^^^^^^^^^^^^^^
  File "/workspace/miniconda3/envs/loc/lib/python3.11/asyncio/tasks.py", line 502, in wait_for
    raise exceptions.TimeoutError() from exc
TimeoutError {"pid": 32663, "job_id": "AJ_idGDMG4ihsxM"}
t
Hey there! I totally understand how frustrating TimeoutErrors can be. Let me help you troubleshoot the MultilingualModel turn detection issue with some friendly, easy-to-follow steps: • Increase Timeout Parameter: No worries! The default timeout is just 3 seconds, which can sometimes be too quick. You can easily give your system a bit more breathing room by extending the timeout: 
probability = await turn_detector.predict_end_of_turn(chat_ctx, timeout=10)
• Check Remote Inference Configuration: Quick check - make sure your 
LIVEKIT_REMOTE_EOT_URL
is set up correctly and that your remote server is responding. Sometimes it's just a small connection hiccup! • Ensure Model Files Are Downloaded: Let's make sure all your model files are ready to go. Just run this simple command: 
python agent.py download-files
• Monitor System Resources: Don't stress - the model is pretty lightweight, needing less than 500MB of RAM and running smoothly on your CPU. Just do a quick check to ensure you've got enough free resources. • Fallback Behavior: The good news? This model is super smart and can handle timeouts gracefully. If something doesn't work perfectly, it'll automatically fall back to local inference or provide a default value. Give these steps a try, and if you're still seeing any issues, I'm right here to help you out! We'll get this sorted together. 😊 Sources: livekit-agents/livekit/agents/voice/agent_activity.py | LiveKit turn detector plugin | LiveKit Docs
r
the EnglishModel() is working fine, Im only facing this issue in the Multilingual() model. does it have any dependency on an external inference server? How to check if its configured.
t
The MultilingualModel for turn detection is actually pretty cool and flexible! By default, it runs right on your computer using your CPU, and it's super lightweight – needing less than 500MB of RAM. But here's the neat part: if you want, you can also set it up to use a remote inference server. Want to check how your MultilingualModel is configured? Here are some friendly ways to do that: • Check the Environment Variable: Take a peek at the 
LIVEKIT_REMOTE_EOT_URL
environment variable in your system setup. If it's there, the model will chat with a remote server for inference. • Observe the Behavior: ◦ If 
LIVEKIT_REMOTE_EOT_URL
is not set or is empty, the model will happily run locally and load its language configs during startup. ◦ If 
LIVEKIT_REMOTE_EOT_URL
is set, the model will reach out to the remote server for inference and threshold checks. • Inspect the Code: The 
_remote_inference_url()
function is your behind-the-scenes detective. If it returns 
None
, that means local inference mode is in action. If you decide you'd rather keep things local and want to disable remote inference, just unset or remove the LIVEKIT_REMOTE_EOT_URL variable, and your model will run locally without missing a beat! Sources: livekit-agents/livekit/agents/voice/agent_activity.py | LiveKit turn detector plugin | LiveKit Docs | LiveKit turn detector plugin | LiveKit Docs
2 Views
Open in Slack
PreviousNext
Powered by