Skip to content
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

"refine_steps" argument name and docs are misleading #54

Open
iuliaturc opened this issue May 9, 2024 · 0 comments
Open

"refine_steps" argument name and docs are misleading #54

iuliaturc opened this issue May 9, 2024 · 0 comments

Comments

@iuliaturc
Copy link

iuliaturc commented May 9, 2024

When setting refiner=base_image_refiner, the refine_steps argument promises to control how many denoising steps will be performed with the refiner:

Screenshot 2024-05-09 at 11 39 35 AM

However, in practice, this is only true if you additional set prompt_strength=1.0. The reason is that:

  • cog-sdxl sets num_inference_steps=refiner_steps here, and
  • diffusers computes the init_step for refinement based both on num_inference_steps and strength here.

This results in surprising behavior, where you set refiner_steps=N only to see that it ran for fewer steps. I think you could either:

  • Update your documentation to reflect this reality, or
  • Automatically set strength=1 here
@iuliaturc iuliaturc changed the title "refiner_steps" argument name and docs are misleading "refine_steps" argument name and docs are misleading May 9, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant