[New Model]: aya 32b vision support

[devops724]

🚀 The feature, motivation and pitch

Hi
Cohere just released aya vision 32b
https://huggingface.co/CohereForAI/aya-vision-32b
it reach to qwen vl 2.5 72B with only 32b parameters
can we support this model

Alternatives

No response

Additional context

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.

[JenZhao]

Could I take this one

[devops724]

Could I take one this one

ensure it work in bitsandbytes too,please, most low vram developers use 32b models in bitsandbytes

[saurabhdash]

@JenZhao if you implement it, could you please tag me and I'll be happy to review it.

[JenZhao]

@JenZhao if you implement it, could you please tag me and I'll be happy to review it.

will do thank you!

[devops724]

aya vision is in top 10 in huggingface trending models

[bikevit2008]

Too much waits aya vision support