-
Notifications
You must be signed in to change notification settings - Fork 6k
APG: Eliminating Oversaturation and Artifacts of High Guidance Scales in Diffusion Models #9585
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
Comments
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. |
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. |
Hi, I'm looking into this now. Is there an open source implementation available by the authors? I couldn't find any from a quick look. I'll first be adding newer CFG methods based on them being easily available openly, and later look into implementing other ones myself |
Oh, I couldn't find a linked github but there was a X post by author: https://x.com/Msadat97/status/1842246625147920722 This should be sufficient for me to quickly support the method! |
Hey @a-r-r-o-w - I believe the answer is yes. Please take a look at Page 22 of the original paper. |
New paper just released, a new way to do classifier-free guidance (CFG) without affecting image saturation, introducing artifacts, etc. Could be huge.
https://huggingface.co/papers/2410.02416
The text was updated successfully, but these errors were encountered: