Nvidia PersonaPlex 7b Integration as a Realtime Audio-to-Audio model

[arpanpreneur]

Feature Type

Nice to have

Feature Description

Excited about Nvidia's open source Audio-to-Audio model Personaplex 7b. Would love to see the livekit nvidia plugin updated to support the same, and instructions in the doc on how to run the model locally.

This would open up a lot of experimentation at the very least.

https://huggingface.co/nvidia/personaplex-7b-v1
https://research.nvidia.com/labs/adlr/personaplex/

Workarounds / Alternatives

No response

Additional Context

No response

[tinalenguyen]

Hi, thanks for creating this report! We will keep this model request in mind as we work on actively improving Agents 😄

[Punchy25]

THATS A MUST! not nice-to-have

[aholtheuerl]

Where is going to be the plugin added?

[domino14]

please to be adding plugin thanks

[milanperovic]

+1! This would be a game-changer for local audio-to-audio experimentation. Eagerly waiting for this plugin support!

[milanperovic]

Hi everyone! 👋
I've prepared a PR that addresses this feature request. You can check it out here: #4660
Any feedback, suggestions, or comments are more than welcome!

[neo-lyzr]

Thanks @milanperovic This is amazing. My company needs this plugin in Livekit

Who should we tag to get the PR along?🤔

[ChrisFeldmeier]

+1

[umairrkhn]

+1

[aheadrox]

+1

[tallshandy]

WE WANT LIVEKIT. WE WANT LIVEKIT