message_received event cannot be triggered.

[Emotibot5]

message_received event cannot be triggered.

chat = rtc.ChatManager(ctx.room)

async def answer_from_text(txt: str):
    chat_ctx = assistant.chat_ctx.copy()
    chat_ctx.append(role="user", text=txt)
    stream = llm_plugin.chat(chat_ctx=chat_ctx)
    await assistant.say(stream)

@chat.on("message_received")
def on_chat_received(msg: rtc.ChatMessage):
    if msg.message:
        asyncio.create_task(answer_from_text(msg.message))

[davidzhao]

how are you sending the chat message? the client-side application would need to send it

[Emotibot5]

I am using local stt, llm, and tts, I can talk with the agent.

assistant = VoiceAssistant(
    vad=vad,
    stt=whisper_stt,
    llm=_llm,
    tts=_tts,
    chat_ctx=initial_ctx,
)

assistant.start(ctx.room)

# listen to incoming chat messages, only required if you'd like the agent to
# answer incoming messages from Chat

chat_mng = rtc.ChatManager(ctx.room)

async def answer_from_text(txt: str):
    chat_ctx = assistant.chat_ctx.copy()
    chat_ctx.append(role="user", text=txt)
    stream = _llm.chat(chat_ctx=chat_ctx)
    await assistant.say(stream)

@chat_mng.on("message_received")
def on_chat_received(msg: rtc.ChatMessage):
    if msg.message:
        print("=====",msg.message, " received")
        asyncio.create_task(answer_from_text(msg.message))

[Emotibot5]

I am using local stt, llm, and tts, I can talk with the agent.

assistant = VoiceAssistant(
    vad=vad,
    stt=whisper_stt,
    llm=_llm,
    tts=_tts,
    chat_ctx=initial_ctx,
)

assistant.start(ctx.room)

# listen to incoming chat messages, only required if you'd like the agent to
# answer incoming messages from Chat

chat_mng = rtc.ChatManager(ctx.room)

async def answer_from_text(txt: str):
    chat_ctx = assistant.chat_ctx.copy()
    chat_ctx.append(role="user", text=txt)
    stream = _llm.chat(chat_ctx=chat_ctx)
    await assistant.say(stream)

@chat_mng.on("message_received")
def on_chat_received(msg: rtc.ChatMessage):
    if msg.message:
        print("=====",msg.message, " received")
        asyncio.create_task(answer_from_text(msg.message))

it is wired that I can talk with the agent, but I set breakpoint in the debug mode, the program cannot suspended.

[davidzhao]

message_received would not be fired unless you are sending Chat messages from the client. this is expected

[Emotibot5]

I am using local stt, llm, and tts, I can talk with the agent.

assistant = VoiceAssistant(
    vad=vad,
    stt=whisper_stt,
    llm=_llm,
    tts=_tts,
    chat_ctx=initial_ctx,
)

assistant.start(ctx.room)

# listen to incoming chat messages, only required if you'd like the agent to
# answer incoming messages from Chat

chat_mng = rtc.ChatManager(ctx.room)

async def answer_from_text(txt: str):
    chat_ctx = assistant.chat_ctx.copy()
    chat_ctx.append(role="user", text=txt)
    stream = _llm.chat(chat_ctx=chat_ctx)
    await assistant.say(stream)

@chat_mng.on("message_received")
def on_chat_received(msg: rtc.ChatMessage):
    if msg.message:
        print("=====",msg.message, " received")
        asyncio.create_task(answer_from_text(msg.message))

it is wired that I can talk with the agent, but I set breakpoint in the debug mode, the program cannot suspended.

My goal is to using stream mode to input llm trunk data to tts and then accelerate the voice response.

[Emotibot5]

I am running this examaple:

https://github.com/livekit/agents/blob/main/examples/voice-assistant/minimal_assistant.py