-
-
Notifications
You must be signed in to change notification settings - Fork 1
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Feature Request] New Model Type: torch-kolors #11
Comments
Hmm, I will take a look I think diffusers only has this pipeline packaged up with txt2img and img2img support I would probably need to vendor (copy) their repo into my source tree with their license since there is no pypi packaging Noticing a trend with lack of packaging in these AI projects :) There are a lot of them like this |
Agreed. It looks like the license is compatible with vendored inclusion. EDIT: There is a third-party packaged version, kolors-sdxl (repo), that might be better than a poke with a sharp stick. |
heavily modify original kolors code to add pipelines currently not supported by diffusers, inpaint, and (img2img + controlnet) synthesize currently non-existant (txt2img + controlnet) and (inpaint + controlnet) pipelines for completeness add support for cpu offloading (vram optimizations)
Kwai Kolors is a modified version of SDXL that adds a ChatGLM-based encoder and is customized for Asian users.
There is a basic pipeline in diffusers (KolorsPipeline), and additional diffusers-based code for Kolors-specific inpainting, ControlNets, and IP Adaptors lives here.
The text was updated successfully, but these errors were encountered: