Skip to content

Replace Sequence Parallel to Paddle Sequence Parallel #7966

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.

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

Merged
merged 12 commits into from
Mar 21, 2024

Conversation

iosmers
Copy link
Contributor

@iosmers iosmers commented Feb 5, 2024

PR types

Others

PR changes

Others

Description

本PR用于将PaddleNLP中的SP并行统一替换为Paddle框架提供的SP并行的代码

Copy link

paddle-bot bot commented Feb 5, 2024

Thanks for your contribution!

@iosmers iosmers force-pushed the support_sp_overlap branch from d751a66 to 2c693cc Compare February 6, 2024 11:24
Copy link

codecov bot commented Feb 6, 2024

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 55.47%. Comparing base (b6dcb4e) to head (8a72ac1).
Report is 1 commits behind head on develop.

Additional details and impacted files
@@             Coverage Diff             @@
##           develop    #7966      +/-   ##
===========================================
+ Coverage    55.37%   55.47%   +0.10%     
===========================================
  Files          596      595       -1     
  Lines        91622    91419     -203     
===========================================
- Hits         50732    50715      -17     
+ Misses       40890    40704     -186     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

# and sp_configs.sp_fused_linear_param_grad_add
# )

self.sp_asyn_reduce_scatter = True
Copy link
Contributor Author

Choose a reason for hiding this comment

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

为了测试,这里先手动写死

@iosmers iosmers closed this Mar 18, 2024
@iosmers iosmers reopened this Mar 18, 2024
@iosmers iosmers changed the title SP support overlap Replace Sequence Parallel to Paddle Sequence Parallel Mar 18, 2024
@iosmers iosmers closed this Mar 19, 2024
@iosmers iosmers reopened this Mar 19, 2024
@ZHUI ZHUI requested a review from DesmonDay March 19, 2024 03:42
@@ -29,7 +29,7 @@
from .feature_extraction_utils import BatchFeature, FeatureExtractionMixin
from .image_processing_utils import ImageProcessingMixin
from .attention_utils import create_bigbird_rand_mask_idx_list
from .sequence_parallel_utils import (
Copy link
Collaborator

Choose a reason for hiding this comment

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

需要paddle的版本是?2.6.0 不行吧?

这里需要加 try except吧

Copy link
Collaborator

@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

Copy link
Contributor

@DesmonDay DesmonDay 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 a092775 into PaddlePaddle:develop Mar 21, 2024
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.

3 participants