Skip to content

ENH: support JoyAI image editing models - #5458

Merged
qinxuye merged 4 commits into
xorbitsai:mainfrom
Minamiyama:models/image/joyai-image
Sep 1, 2026
Merged

ENH: support JoyAI image editing models#5458
qinxuye merged 4 commits into
xorbitsai:mainfrom
Minamiyama:models/image/joyai-image

Conversation

@Minamiyama

Copy link
Copy Markdown
Collaborator

Summary

Add built-in support for the following JoyAI image editing models:

  • joyai-image-edit
  • joyai-image-edit-plus

Both Hugging Face and ModelScope sources are supported.

Implementation

  • Register model metadata, default generation settings, and virtualenv dependencies.
  • Use DiffusionPipeline because JoyAI edit pipelines are not registered in AutoPipelineForText2Image.
  • Support JoyAI Image Edit Plus multi-image input:
    • Merge the primary image and reference_images.
    • Pass images through the pipeline's images argument.
    • Always provide a list, including single-image requests.
    • Avoid passing unsupported width and height arguments.
  • Add built-in model documentation.
  • Add focused metadata, loading, and multi-image tests.

Model Sources

Model Hugging Face Revision ModelScope Revision
joyai-image-edit jdopensource/JoyAI-Image-Edit-Diffusers main jd-opensource/JoyAI-Image-Edit-Diffusers master
joyai-image-edit-plus jdopensource/JoyAI-Image-Edit-Plus-Diffusers main jd-opensource/JoyAI-Image-Edit-Plus-Diffusers master

Official project: https://github.com/jd-opensource/JoyAI-Image

@XprobeBot XprobeBot added the enhancement New feature or request label Aug 30, 2026
@gemini-code-assist

Copy link
Copy Markdown
Contributor

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

@XprobeBot XprobeBot added this to the v3.x milestone Aug 30, 2026
Comment thread xinference/model/image/stable_diffusion/core.py
JoyAI Image Edit Plus does not expose num_images_per_prompt, so invoke
the pipeline once per requested output and preserve per-image seeds.
Comment thread xinference/model/image/stable_diffusion/core.py
Map each JoyAI Plus invocation into one request-wide progress range so
100% is reported only after the final output.

@qinxuye qinxuye left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@qinxuye
qinxuye merged commit 9f6119a into xorbitsai:main Sep 1, 2026
14 of 15 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

enhancement New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants