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 stage2 overlap #46495

Merged
merged 1 commit into from
Sep 28, 2022
Merged

Sharding stage2 overlap #46495

merged 1 commit into from
Sep 28, 2022

Conversation

FeixLiu
Copy link
Contributor

@FeixLiu FeixLiu commented Sep 26, 2022 •
edited
Loading

PR types

Others

PR changes

Others

Describe

Sharding stage2 overlap

Usage

model, optimizer, scaler = group_sharded_parallel(...)
model._set_comm_overlap(True)

loss compare on 6.7B model

image

speed compare

no overlap overlap gain
33010 40143 +21.6%

@paddle-bot
Copy link

paddle-bot bot commented Sep 26, 2022

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.

sneaxiy
sneaxiy approved these changes Sep 28, 2022
View reviewed changes
Copy link
Collaborator

@sneaxiy sneaxiy left a comment

Choose a reason for hiding this comment

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

LGTM. Excellent job.

@FeixLiu FeixLiu mentioned this pull request Sep 28, 2022
Merged
@FeixLiu FeixLiu merged commit 9c01eae into PaddlePaddle:develop Sep 28, 2022
@FeixLiu FeixLiu deleted the sharding_stage2_overlap branch September 28, 2022 11:39
FeixLiu added a commit to FeixLiu/Paddle that referenced this pull request Oct 17, 2022
@FeixLiu
[dygraph sharding] Overlap the reduce and the caculation for sharding…
a98a306 
… stage 2. (PaddlePaddle#46495)
fuyinno4 pushed a commit that referenced this pull request Oct 18, 2022
@FeixLiu
Cherry pick for sharding (#47061)
5b64214 
* [dygraph sharding] Overlap the reduce and the caculation for sharding stage 2. (#46495)

* [dygraph sharding stage 2] sharding broadcast overlap (#46656)

* Multi groups for broadcast of sharding stage 2 (#46894)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@sneaxiy sneaxiy sneaxiy 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.
2 participants
@FeixLiu @sneaxiy