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.

Sign up for GitHub

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

Jump to bottom

[GPU] MLP : 2fcs + swiglu fusion #27831

Open
wants to merge 10 commits into
base: master
Choose a base branch
from
Open

[GPU] MLP : 2fcs + swiglu fusion #27831

wants to merge 10 commits into from
+436 −48

Conversation

yeonbok
Copy link
Contributor

@yeonbok yeonbok commented Nov 29, 2024

Details:

  • 2 FCs + swiglu in MLP pattern are fused
  • Only applied to cldnn && #EUs > 128 && glu type with swiglu

Tickets:

  • 152163

@yeonbok
swiglu_fuse_infra
3ad5ccb 
swiglu kernel
working with static kernel only

swiglu : working with dynamic shape only

swiglu works with dynamic shape + static shape

no print

refactor

Fix phi3 crash

Support dynamic quantization path

Fix 1st token regression

Apply only for cldnn

clean up
@yeonbok
Fix rebase error
a15090e
@yeonbok
clean up
ec42903
@yeonbok yeonbok requested review from a team as code owners November 29, 2024 14:37
@github-actions github-actions bot added the category: GPU OpenVINO GPU plugin label Nov 29, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@ahnyoung-paul ahnyoung-paul ahnyoung-paul approved these changes

Assignees
No one assigned
Labels
category: GPU OpenVINO GPU plugin
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@yeonbok @ahnyoung-paul