Add support for LongCat-Image

[junqiangwu]

What does this PR do?

This PR introduces the LongCat-Image Series into the diffusers library. The LongCat-Image Series comprises powerful and highly efficient 6B-parameter models, featuring both a state-of-the-art image generation model and a versatile image editing model. Designed to be friendly for consumer-grade hardware, this series demonstrates strong capabilities in high-fidelity image synthesis and precise, controllable image editing. The technical report and checkpoints for both models have been released.

Thanks for the support of @yiyixuxu @apolinario.

• GitHub: https://github.com/meituan-longcat/LongCat-Image
• LongCat-Image on Hugging Face: https://huggingface.co/meituan-longcat/LongCat-Image
• LongCat-Image-Edit on Hugging Face: https://huggingface.co/meituan-longcat/LongCat-Image-Edit

[yiyixuxu]

thanks a lot for the PR!
looks really good to me, I left some very small feedbacks, I think we can merge this soon:)

we also need doc & tests but can work on that in a follow-up PR if you prefer. Let us know

[yiyixuxu]

thanks! I left a few more comments

[yiyixuxu]

are we only support single prompt? if so maybe can add a check in check_inputs

[junqiangwu]

We’ve updated the code so that the text-to-image pipeline now accepts multiple prompts, while the edit pipeline still takes a single prompt. Both behaviors are enforced via checks in check_inputs.

[yiyixuxu]

I think you need to add the doc page into the index here https://github.com/huggingface/diffusers/blob/main/docs/source/en/_toctree.yml

also can you run make style and make fix-copies?

[junqiangwu]

Thanks for the feedback — we’ve updated the code accordingly and run make style and make fix-copies in this commit. Please let us know if you spot anything else that should be addressed. @yiyixuxu