K-Diffusion + Stable Diffusion Pipeline

[patrickvonplaten]

Model/Pipeline/Scheduler description

It would be really nice to have a pipeline that integrates well with k-diffusion so that all schedulers be used out of the box will all checkpoints compatible with diffusers.

diffusers is starting to have many variations of stable diffusion but does not yet have all the samplers/schedulers of Katherine's amazing k-diffusion library. Would be cool to combine the two in a pipeline that we could put under:

src/pipelines/k_diffusion/pipeline_k_diffusion.py

If anybody is happy to give it a try - this: https://github.com/huggingface/diffusers/pull/857 seems stale - it'd be very nice!

Open source status

• [ ] The model implementation is available
• [ ] The model weights are available (Only relevant if addition is not a scheduler).

Provide useful links for the implementation

More than happy to help, make it work :-)

[keturn]

@Birch-san Do you have an implementation of this you can submit as a PR here?

[patrickvonplaten]

@keturn, actually I had some time to look into it today: https://github.com/huggingface/diffusers/pull/1360

Could you try this?

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.