[New Model]: CosyVoice2-0.5B

[MengAiDev]

The model to consider.

https://huggingface.co/FunAudioLLM/CosyVoice2-0.5B

The closest model vllm already supports.

No response

What's your difficulty of supporting the model you want?

No response

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

[ZJY0516]

@DarkLight1337 Would it be possible to assign this task to me?

[DarkLight1337]

Sure. For the audio output, you can have the model just output audio tokens, and let the user handle the responsibility of decoding them.

[deyituo]

there are two versions for this vllm model...

[github-actions]

This issue has been automatically marked as stale because it has not had any activity within 90 days. It will be automatically closed if no further activity occurs within 30 days. Leave a comment if you feel this issue should remain open. Thank you!

[DarkLight1337]

For TTS models, please direct your requests to vllm-project/vllm-omni which support multimodal outputs. Thanks!