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

Refine self_attention_fuse pass with removing the size restriction #56909

Merged
merged 1 commit into from Sep 11, 2023
Merged

Refine self_attention_fuse pass with removing the size restriction #56909

merged 1 commit into from Sep 11, 2023

Conversation

abenmao
Copy link
Contributor

@abenmao abenmao commented Sep 2, 2023 •
edited by xinyu-intel

PR types

Bug fixes

PR changes

Others

Description

remove the size restriction in self_attention_fuse pass. Fix PaddleOCR v4 on AVX512 CPU.

@paddle-bot
Copy link

paddle-bot bot commented Sep 2, 2023

你的PR提交成功，感谢你对开源项目的贡献!
请关注后续CI自动化测试结果，详情请参考Paddle-CI手册。
Your PR has been submitted. Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

@paddle-bot paddle-bot bot added the contributor External developers label Sep 2, 2023
YangQun1
YangQun1 previously approved these changes Sep 5, 2023
View reviewed changes
xinyu-intel
xinyu-intel previously approved these changes Sep 5, 2023
View reviewed changes
@xinyu-intel
Copy link
Contributor

@tianshuo78520a Hi, we've added test cases for this PR but it seems cannot pass the c++ converage check. Can you give some suggestions on that?

@xinyu-intel
Copy link
Contributor

@zoooo0820 @jeff41404 Hi, this is an enhancement for the current fluid operator. Can you please check on it?

zoooo0820
zoooo0820 reviewed Sep 7, 2023
View reviewed changes
Copy link
Contributor

@zoooo0820 zoooo0820 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi， there are some fluid APIs, which will be deprecated very soon. Could you please modify them to our recommended APIs？

By the way, we are planning to remove fluid APIs in this week, and this PRs may be conflicted. Is it possible to merge this PR in next week, thanks.

test/mkldnn/test_fused_vit_attention.py Outdated Show resolved Hide resolved
test/mkldnn/test_fused_vit_attention.py Outdated Show resolved Hide resolved
@xinyu-intel
Copy link
Contributor

Hi， there are some fluid APIs, which will be deprecated very soon. Could you please modify them to our recommended APIs？

By the way, we are planning to remove fluid APIs in this week, and this PRs may be conflicted. Is it possible to merge this PR in next week, thanks.

Thanks for your suggestions, we'll hold on this PR.

@zoooo0820
Copy link
Contributor

Hi, @xinyu-intel @abenmao
Sorry for blocking this PR before. We have already removed the fluid APIs, now you can just replace these APIs according to the comments, resolve the conflict and continue this PR directly.

@abenmao abenmao dismissed stale reviews from xinyu-intel and YangQun1 via cb83604 September 8, 2023 08:35
@abenmao
Copy link
Contributor Author

abenmao commented Sep 10, 2023

Hi, @xinyu-intel @abenmao Sorry for blocking this PR before. We have already removed the fluid APIs, now you can just replace these APIs according to the comments, resolve the conflict and continue this PR directly.

Done~ But The PR-CI-Api-Benchmark seems to be hung for some reason.

vivienfanghuagood
vivienfanghuagood approved these changes Sep 11, 2023
View reviewed changes
Copy link
Contributor

@vivienfanghuagood vivienfanghuagood left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@xinyu-intel xinyu-intel merged commit 4998c32 into PaddlePaddle:develop Sep 11, 2023
26 of 27 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@zoooo0820 zoooo0820 zoooo0820 left review comments

@xinyu-intel xinyu-intel xinyu-intel approved these changes

@YangQun1 YangQun1 YangQun1 approved these changes

@vivienfanghuagood vivienfanghuagood vivienfanghuagood approved these changes

Assignees

@vivienfanghuagood vivienfanghuagood

Labels
contributor External developers Intel
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
5 participants
@abenmao @xinyu-intel @zoooo0820 @YangQun1 @vivienfanghuagood