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

sharding stage1 V1 support Broadcast overlap Forward #63945

Merged
merged 13 commits into from
May 11, 2024
Merged

sharding stage1 V1 support Broadcast overlap Forward #63945

merged 13 commits into from
May 11, 2024

Conversation

iosmers
Copy link
Contributor

@iosmers iosmers commented Apr 28, 2024 •
edited
Loading

PR Category

Performance Optimization

PR Types

Performance

Description

1、本PR主要是针对shrding Stage V1的param的broadcast和前向计算进行重叠，以实现性能优化
2、正确性验证，优化前和优化后逐位对齐实验结果

image 3、llama7B sharding 8,性能提升
no_overlap overlap speedup
6877 7053.3278 2.6%

4、timeline 分析
image

card-13678

@paddle-bot Paddle-bot
Copy link

paddle-bot bot commented Apr 28, 2024

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-ci-bot Paddle-ci-bot
Copy link

paddle-ci-bot bot commented May 6, 2024

Sorry to inform you that eb36ebe's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually.

@iosmers iosmers marked this pull request as draft May 7, 2024 06:34
@iosmers iosmers marked this pull request as ready for review May 7, 2024 06:34
@iosmers iosmers marked this pull request as draft May 7, 2024 08:41
@iosmers iosmers marked this pull request as ready for review May 7, 2024 08:41
@iosmers iosmers changed the title sharding v1 overlap sharding stage1 V1 support Broadcast overlap Forward May 9, 2024
deepllz
deepllz approved these changes May 10, 2024
View reviewed changes
Copy link
Contributor

@deepllz deepllz left a comment

Choose a reason for hiding this comment

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

LGTM

@sneaxiy sneaxiy merged commit 3789d83 into PaddlePaddle:develop May 11, 2024
31 checks passed
co63oc pushed a commit to co63oc/Paddle that referenced this pull request May 12, 2024
@iosmers @co63oc
sharding stage1 V1 support Broadcast overlap Forward (PaddlePaddle#63945
06cdc1e 
)

* sharding v1 overlap

* delete pybind

* add txt

* add b.txt

* delete test file

* add pybind

* add test case for stage1 v1 overlap

* add test case for stage1 v1 overlap

* update test case

* delete print optimizer

* update

* update models to layers

* update mlp1
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@FeixLiu FeixLiu FeixLiu approved these changes

@deepllz deepllz deepllz approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
4 participants
@iosmers @FeixLiu @deepllz @sneaxiy