fix elevenlabs auto mode

[simllll]

Description

fixes #819 (without blingfire implementation)

Changes Made

• added auto_mode flag (kept it to false by default, in python it is true by default though)
• added checks for sentencetokenizer
• added types for sentencetokenizer in the elevenlabs implementation

Pre-Review Checklist

• Build passes: All builds (lint, typecheck, tests) pass locally
• AI-generated code reviewed: Removed unnecessary comments and ensured code quality
• Changes explained: All changes are properly documented and justified above
• Scope appropriate: All changes relate to the PR title, or explanations provided for why they're included

Testing

• manual test locally

[changeset-bot]

🦋 Changeset detected

Latest commit: 19a21a4

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 1 package

Name	Type
@livekit/agents-plugin-elevenlabs	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[simllll]

WARN (285528): skipping user input, current speech generation cannot be interrupted when using the "flush" parameter.

I tried mirroring the python implementation and therefore also added the flush paramater when using sentence tokenizer and automode. but adding "flush" breaks it somehow....I guess I'm missing something that python has and we don't have in typescript.
I'm removing the flush parameter for now again, without it it seems to work quite fine.

[simllll]

Followup on "flush: true".

The reason why this actually broke thigns is that the livekit implemenation currently is not using the multi stream input api.. it starts a new websocket on each turn.

the python version uses this: https://elevenlabs.io/docs/api-reference/text-to-speech/v-1-text-to-speech-voice-id-multi-stream-input, there we also need the flush parameter. I made a quick implementation, it seems to be portable from python but needs more work.

[toubatbrian]

LG! Really nice about the inline comments!

[toubatbrian]

the python version uses this: https://elevenlabs.io/docs/api-reference/text-to-speech/v-1-text-to-speech-voice-id-multi-stream-input, there we also need the flush parameter. I made a quick implementation, it seems to be portable from python but needs more work.

Thanks for digging into this! Let me know if you need help!