search

oobabooga / text-generation-webui

A Gradio web UI for Large Language Models.
GNU Affero General Public License v3.0
39.67k stars 5.21k forks source link

[Feature request]: fast-whisper+streaming #5054

Closed Katehuuh closed 6 months ago

Katehuuh commented 9 months ago

Description

whisper-large-v3 and distil-large-v2 with integration of https://github.com/Vaibhavs10/insanely-fast-whisper could be interesting for real-time dialogue. Also streaming option to not have to click on microphone where for example current extensions could start record in def output_modifier for trigger document.querySelector('.mic-wrap button').click(); and send message when detect silence than loop.

Additional Context

In my junky trial loop as custom_js(): execute once, it works in interval, a single record could be easily managed.

Edit: I modify current whisper_stt with gradio feature streaming=True

Urammar commented 8 months ago

Seconded

github-actions[bot] commented 6 months ago

This issue has been closed due to inactivity for 2 months. If you believe it is still relevant, please leave a comment below. You can tag a developer in your comment.

sheepdestroyer commented 6 months ago

@oobabooga Is that how I am supposed to tag this issue? It seems it would be a really nice feature.