https://livekit.io logo
Join Slack
Powered by
im trying to run the new adaptive interruption fea...
# ask-ai
b
im trying to run the new adaptive interruption feature from the 1.5.0 version and im getting error
Copy code
adaptive interruption disabled due to unrecoverable error, falling back to VAD-based    
                                           interruption
here is the full error log
Copy code
00:32:47.031 INFO   livekit.agents     Interruption metrics  
                                         {"model_name": "adaptive interruption", "model_provider": "livekit", "total_duration": 
0.59, "prediction_duration": 0.02, "detection_delay": 0.75, "num_interruptions": 0, "num_backchannels": 1, "num_requests": 8, 
"room": "test_room_xkP5TtDDZA5gNhCPVVNt", "pid": 3492757, "job_id": "AJ_mVU4J52w79xZ", "room_id": "RM_Q3bVhjL7HLbU"}
    00:32:47.432 INFO   livekit.agents     STT metrics  
                                         {"model_name": "deepgram/nova-3", "model_provider": "livekit", "audio_duration": 3.9, 
"room": "test_room_xkP5TtDDZA5gNhCPVVNt", "pid": 3492757, "job_id": "AJ_mVU4J52w79xZ", "room_id": "RM_Q3bVhjL7HLbU"}
    00:32:48.488 WARNI… livekit.agents     adaptive interruption disabled due to unrecoverable error, falling back to VAD-based    
                                           interruption                                                                            
                                         {"room": "test_room_xkP5TtDDZA5gNhCPVVNt", "pid": 3492757, "job_id": "AJ_mVU4J52w79xZ", 
"room_id": "RM_Q3bVhjL7HLbU"}
                 ERROR  livekit.agents     Error in _main_task                                                                     
                                           Traceback (most recent call last):                                                      
                                             File                                                                                  
                                           "/home/mohammedusamaahmed/usama1/Voice-Agents/newLivekitVersion/1_0_21_newLivekitVers…  
                                           line 17, in async_fn_logs                                                               
                                               return await fn(*args, **kwargs)                                                    
                                             File                                                                                  
                                           "/home/mohammedusamaahmed/usama1/Voice-Agents/newLivekitVersion/1_0_21_newLivekitVers…  
                                           line 478, in _main_task                                                                 
                                               return await self._run()                                                            
                                             File                                                                                  
                                           "/home/mohammedusamaahmed/usama1/Voice-Agents/newLivekitVersion/1_0_21_newLivekitVers…  
                                           line 1093, in _run                                                                      
                                               task.result()                                                                       
                                             File                                                                                  
                                           "/home/mohammedusamaahmed/usama1/Voice-Agents/newLivekitVersion/1_0_21_newLivekitVers…  
                                           line 946, in send_task                                                                  
                                               raise APIStatusError(                                                               
                                           livekit.agents._exceptions.APIStatusError: message='interruption inference timed out    
                                           after 1.0s (ws)', status_code=408, retryable=False                                      
                                         {"room": "test_room_xkP5TtDDZA5gNhCPVVNt", "pid": 3492757, "job_id": "AJ_mVU4J52w79xZ", 
"room_id": "RM_Q3bVhjL7HLbU"}
Traceback (most recent call last):
  File 
"/home/mohammedusamaahmed/usama1/Voice-Agents/newLivekitVersion/1_0_21_newLivekitVersionVenv/lib/python3.10/site-packages/livekit/a
gents/utils/log.py", line 17, in async_fn_logs
    return await fn(*args, **kwargs)
  File 
"/home/mohammedusamaahmed/usama1/Voice-Agents/newLivekitVersion/1_0_21_newLivekitVersionVenv/lib/python3.10/site-packages/livekit/a
gents/inference/interruption.py", line 478, in _main_task
    return await self._run()
  File 
"/home/mohammedusamaahmed/usama1/Voice-Agents/newLivekitVersion/1_0_21_newLivekitVersionVenv/lib/python3.10/site-packages/livekit/a
gents/inference/interruption.py", line 1093, in _run
    task.result()
  File 
"/home/mohammedusamaahmed/usama1/Voice-Agents/newLivekitVersion/1_0_21_newLivekitVersionVenv/lib/python3.10/site-packages/livekit/a
gents/inference/interruption.py", line 946, in send_task
    raise APIStatusError(
livekit.agents._exceptions.APIStatusError: message='interruption inference timed out after 1.0s (ws)', status_code=408, 
retryable=False
    00:32:50.083 INFO   livekit.agents     STT metrics  
                                         {"model_name": "deepgram/nova-3", "model_provider": "livekit", "audio_duration": 2.65, 
"room": "test_room_xkP5TtDDZA5gNhCPVVNt", "pid": 3492757, "job_id": "AJ_mVU4J52w79xZ", "room_id": "RM_Q3bVhjL7HLbU"}
    00:32:50.939 INFO   livekit.agents     STT metrics  
                                         {"model_name": "deepgram/nova-3", "model_provider": "livekit", "audio_duration": 0.85, 
"room": "test_room_xkP5TtDDZA5gNhCPVVNt", "pid": 3492757, "job_id": "AJ_mVU4J52w79xZ", "room_id": "RM_Q3bVhjL7HLbU"}
    00:32:50.940 DEBUG  livekit.agents     received user transcript  
                                         {"user_transcript": "All right.", "language": "en", "transcript_delay": 
0.5605952739715576, "room": "test_room_xkP5TtDDZA5gNhCPVVNt", "pid": 3492757, "job_id": "AJ_mVU4J52w79xZ", "room_id": 
"RM_Q3bVhjL7HLbU"}
does anyone have any idea what this is, why is the adaptive interruption getting disabled?
✅ 1
t
Thanks for reaching out, Mohammed. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.
Thanks — I looked at the logs you pasted. You're hitting an adaptive-interruption inference timeout: • The agent logs show: "interruption inference timed out after 1.0s (ws)" and then "adaptive interruption disabled due to unrecoverable error, falling back to VAD-based interruption." • That means the agent attempted to call LiveKit’s adaptive interruption inference service over WebSocket, the call did not respond within the 1s window (HTTP 408), and the agent disabled the adaptive model and reverted to VAD. What to check next (quick checklist): • Are you running the agent on LiveKit Cloud or self‑hosted? Adaptive interruption requires connectivity to LiveKit’s inference endpoint. • Is there any firewall, proxy, or network rule that could block or add latency to WebSocket connections from the agent host to LiveKit? • Is the host experiencing high CPU/memory or network congestion that could delay requests? • Are you consistently seeing timeouts or only intermittently? (Frequency helps diagnosis.) If you can share: • Deployment type (LiveKit Cloud vs self‑hosted) • Whether this happens every run or intermittently • Any relevant network environment details (behind corporate proxy, firewall rules, region) I’ll use that to suggest the next steps (network checks, increasing timeouts if possible, or further logs to capture). Sources: Solving unwanted interruptions with Adaptive Interruption Handling | LiveKi... | Solving unwanted interruptions with Adaptive Interruption Handling | LiveKi... | Adaptive interruption handling | LiveKit Documentation, with interactive elements
b
im consistently seeing this error
t
Thanks — I see you’re consistently getting the timeout: "interruption inference timed out after 1.0s (ws)" which causes the agent to disable adaptive interruption and fall back to VAD. Quick checklist and next steps to diagnose and fix this: 1. Confirm deployment ◦ Are you running the agent on LiveKit Cloud or self‑hosted? 1. Network & connectivity checks ◦ Ensure the agent host can open stable WebSocket connections to LiveKit inference endpoints (no firewall/proxy blocking or packet filtering). ◦ Check for high network latency or packet loss from the agent host to the cloud (run ping/traceroute to the LiveKit region or check your monitoring). 1. Host resource checks ◦ Verify the agent host isn’t CPU- or network‑constrained during runs (spikes can delay requests). 1. STT & settings ◦ Confirm VAD is enabled and your STT model supports aligned transcripts (
stt.capabilities.aligned_transcript
). ◦ If you want to temporarily avoid timeouts, set interruption mode to VAD explicitly: set 
interruption.mode
to 
"vad"
in 
TurnHandlingOptions
. Please reply with: • Deployment type (LiveKit Cloud vs self‑hosted) • Whether you’re behind a proxy/firewall • Region of your agent host and approximate network latency to LiveKit (if available) With that I’ll suggest targeted network tests and next configuration steps. Sources: Adaptive interruption handling | LiveKit Documentation | Solving unwanted interruptions with Adaptive Interruption Handling | LiveKi..., with interactive elements
3 Views
Open in Slack
PreviousNext
Powered by