-
Notifications
You must be signed in to change notification settings - Fork 5.6k
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
[Core] feat: enable fused attention projections for other SD and SDXL pipelines #6179
Conversation
Will @sayakpaul add it for Controlnet? Are there benchmark results? |
@@ -23,6 +23,7 @@ | |||
from ...image_processor import PipelineImageInput, VaeImageProcessor |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The pipeline is deprecated let's not change it anymore ideally
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
@sayakpaul , @patrickvonplaten Install:
I tested this optimization code. Only 3 seconds faster for 10 samples. Is this true? |
… pipelines (huggingface#6179) * feat: enable fused attention projections for other SD and SDXL pipelines * add: test for SD fused projections.
… pipelines (huggingface#6179) * feat: enable fused attention projections for other SD and SDXL pipelines * add: test for SD fused projections.
… pipelines (huggingface#6179) * feat: enable fused attention projections for other SD and SDXL pipelines * add: test for SD fused projections.
What does this PR do?
A follow-up of #6030.
Allow the main SD and SDXL pipelines to benefit from horizontal fusion of attention projection matrices.
Related to: https://github.com/huggingface/sdxl-fast.