Skip to content

Conversation

ForFishes
Copy link
Member

PR types

Bug fixes

PR changes

Others

Description

[Distributed]Add dp/sharding overlap for pipeline

Copy link
Contributor

@ZHUI ZHUI left a comment

Choose a reason for hiding this comment

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

LGTM

@ZHUI ZHUI merged commit 47a71a1 into PaddlePaddle:refactor-training-loop Jul 26, 2023
@ForFishes ForFishes deleted the add_dp_overlap branch August 2, 2023 02:59
ForFishes added a commit to ForFishes/PaddleNLP that referenced this pull request Aug 10, 2023
wawltor pushed a commit that referenced this pull request Aug 11, 2023
* support pp delay_scale_loss and dp_comm_overlap

* [Distributed]Add dp/sharding overlap for pipeline (#6504)

* [Distributed]Support pipelineparallel in accumulation_steps (#6509)

* fix trainer

---------

Co-authored-by: haohongxiang <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants