i have this setup , inspite of having correct setup not able to get the experience same as it shown in this video,

session = AgentSession(
            stt=deepgram.STT(model='nova-3-general',
            language=en-US,
            endpointing_ms=1200,
            smart_format=True
            interim_results=True
            api_key=apikey
            no_delay=False,sample_rate=8000),
            llm=customLLM
            tts=tts,
            vad=vad_instance,
            turn_detection=MultilingualModel(),
            min_interruption_words=1,
            user_away_timeout=user_away_timeout,
        )

Thanks for reaching out, Navin. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

in our case the we are getting early turn detection

To help with those early turn detection hiccups, you can tweak the settings of the turn detection model and Voice Activity Detection (VAD). Here are some friendly steps to guide you: • Try the Turn Detector Plugin: LiveKit offers a custom language model specifically for turn detection. You can set it up like this:

from livekit.plugins.turn_detector.english import EnglishModel

session = AgentSession(
    ...
    turn_detection=EnglishModel(),
)

• Tweak VAD Parameters: Play around with settings like

min_endpointing_delay

and

max_endpointing_delay

to find the right balance between responsiveness and avoiding interruptions. For example:

session = AgentSession(
    ...
    min_endpointing_delay=0.8,
    max_endpointing_delay=4.0,
)

• Go Multilingual if Needed: If you’re working with multiple languages, swap out the English model for the multilingual one like this:

from livekit.plugins.turn_detector.multilingual import MultilingualModel

session = AgentSession(
    ...
    turn_detection=MultilingualModel(),
)

These little adjustments should help cut down on those early interruptions. If you want to chat more about this or need extra help, just let me know!

We tried the above suggestion, none of it seems working ..

I don't have the answer you're looking for. You could also try asking your question: • in one of the other Slack channels or • to https://deepwiki.com/livekit/livekit_composite which is trained on all LiveKit source code If you find the answer, please post it here to help others!