agents

agents copied to clipboard

job is unresponsive, killing job

1

hello, I don’t know what this ping method is checking. Everything on the livekit server and the agent worker seems to be normal and can respond normally. However, due to...

leejiliang

Add token_usage_callback to openai llm plugin

1

Now consumers can receive token usage metrics after each LLM completion event with the openai LLM plugin: ```py def token_usage_callback(token_usage: dict[str, int]) -> None: cost_per_prompt_token = 2.5 / 1000000 #...

willsmanley

Error report: https://portola.sentry.io/share/issue/f5d6e682fa3f4233ac6da7d3d3472e90/ Is it possible this is just benign logspam from the Eleven Labs plugin closing its connection (`livekit.plugins.elevenlabs.tts.SynthesizeStream is closed`)?

egoldschmidt

hi, iam facing an issue with livekit agent version 0.8.4 and 0.8.5, it works well in 0.8.2, in this method _will_synthesize_assistant_answer() as per this example(we have a modified version of...

Test-isom

### What's the issue? In the basic VoiceAssistant example, punctuation isn't streamed to the client until tts is complete. This means "Reflection, I hear you would like to talk about...

brentbaum

message_received event cannot be triggered.

7

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))

Emotibot5

Feature: ability to post-process responses from LLM in voice assistant

2

Sometimes it's desirable to control or change the output from LLMs before TTS.

davidzhao

google gemini plugin

6

Hi, I'm trying to build an assistant with google gemini as LLM. Are there any guidelines, docs or examples to follow on how to build a livekit-gemini plugin ? Thanks

makerGeek

support clova speech

4

Plugin for Clova Speech-to-text from Naver cloud platform, Documentation: [ai-application-service-clovaspeech](https://api.ncloud-docs.com/docs/en/ai-application-service-clovaspeech)

josephkieu

Stream Manipulation using Custom Plugins

4

Due to lack of Documentation in plugins section, I'm not sure how to build one let alone how to customise the already existing one. I want manipulate the input and...

parshvadaftari