Skip to content

Add dynamic Caching when batch_size = 1 for flux model #138

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

Merged
merged 4 commits into from
Mar 9, 2025

Conversation

ita9naiwa
Copy link
Collaborator

#105
Hi, this pr introduces ParaAttention's dynamic caching during diffusion process.
I implement only for flux model, but it is not that difficult to extend to other models, so I want to ask maintainers of this awesome library to go with it or not to proceed further.

Previous PR #113 targets to main branch, so I recreated PR to target dev branch.

@ita9naiwa
Copy link
Collaborator Author

@lmxyy

@lmxyy
Copy link
Collaborator

lmxyy commented Mar 4, 2025

Will test your PR tomorrow.

@lmxyy
Copy link
Collaborator

lmxyy commented Mar 9, 2025

start reviewing your PR. finding a way to resolve conflicts

@lmxyy lmxyy merged commit c2d7f1b into mit-han-lab:dev Mar 9, 2025
@lmxyy
Copy link
Collaborator

lmxyy commented Mar 9, 2025

I fixed the conflicts and merged it. I will test it. If this works, I will merge it to the main branch. Thanks for your PR!

This was referenced Mar 11, 2025
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

Successfully merging this pull request may close these issues.

2 participants