Feature Request: Qwen2.5-Omni

[Kreijstal]

Prerequisites

• I am running the latest code. Mention the version if possible as well.
• I carefully followed the README.md.
• I searched using keywords relevant to my issue to make sure that I am creating a new issue that is not already open (or closed).
• I reviewed the Discussions, and have a new and useful enhancement to share.

Feature Description

Pleas add support https://github.com/QwenLM/Qwen2.5-Omni

Motivation

This is a multimodal which supports audio/video and text

Possible Implementation

I dont think it is similar to other Qwen 2.5 models, right?

MNN now has support for Qwen2.5-Omni-7B/3B, maybe you can try MNN: https://github.com/alibaba/MNN/tree/master

it works fast with aarch64 cpu, with a acceptable latency using audio/image input.

[BVEsun]

As the server now supports Qwen2-audio and Qwen2.5-VL,
will Qwen2.5-Omni be supported soon?

Maybe just support text, audio, and image at first,
and implement video support later.

[jacekpoplawski]

isn't this already implemented?
#13784
https://huggingface.co/models?search=Qwen2.5-Omni-7B-GGUF

[github-actions]

This issue was closed because it has been inactive for 14 days since being marked as stale.