Skip to content

Support tpdp-ep sharding reshard #10568

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

Conversation

sneaxiy
Copy link
Collaborator

@sneaxiy sneaxiy commented May 8, 2025

Before submitting

  • Lint code. If there are lint issues, please format the code first.
# Install and register `pre-commit` in the project folder
pip install pre-commit && pre-commit install

# Process previous code files separately
pre-commit run --file XXXX.py
  • Add test cases into tests folder. If there are codecov issues, please add tests cases first.

PR types

New features

PR changes

APIs

Description

Support tpdp-ep sharding reshard.

Copy link

paddle-bot bot commented May 8, 2025

Thanks for your contribution!

@sneaxiy sneaxiy force-pushed the support_ep_reshard branch from 7985e41 to d447d0c Compare May 8, 2025 08:14
Copy link

codecov bot commented May 8, 2025

Codecov Report

Attention: Patch coverage is 13.58025% with 210 lines in your changes missing coverage. Please review.

Project coverage is 48.96%. Comparing base (51009f9) to head (7985e41).
Report is 1 commits behind head on incubate/paddlenlp-fleety_20250421.

Current head 7985e41 differs from pull request most recent head c2c0537

Please upload reports for the commit c2c0537 to get more accurate results.

Files with missing lines Patch % Lines
paddlenlp/trainer/utils/sharding_io.py 7.69% 120 Missing ⚠️
paddlenlp/trainer/utils/reshard/common.py 16.04% 68 Missing ⚠️
paddlenlp/trainer/utils/reshard/sharding_v2.py 18.75% 13 Missing ⚠️
paddlenlp/trainer/training_args.py 41.66% 7 Missing ⚠️
paddlenlp/trainer/utils/reshard/sharding_v1.py 50.00% 2 Missing ⚠️

❌ Your patch status has failed because the patch coverage (13.58%) is below the target coverage (80.00%). You can increase the patch coverage or adjust the target coverage.
❌ Your project status has failed because the head coverage (48.96%) is below the target coverage (58.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files
@@                          Coverage Diff                           @@
##           incubate/paddlenlp-fleety_20250421   #10568      +/-   ##
======================================================================
+ Coverage                               48.94%   48.96%   +0.02%     
======================================================================
  Files                                     768      765       -3     
  Lines                                  126279   126281       +2     
======================================================================
+ Hits                                    61806    61833      +27     
+ Misses                                  64473    64448      -25     

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

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@sneaxiy sneaxiy force-pushed the support_ep_reshard branch from d447d0c to c2c0537 Compare May 12, 2025 02:06
@sneaxiy sneaxiy requested a review from ZHUI May 12, 2025 07:22
@@ -1230,6 +1230,9 @@ def __post_init__(self):

if expert_parallel_degree > 1:
moe_sharding_parallel_degree = world_size // (pipeline_parallel_degree * expert_parallel_degree)
assert (
self.expert_tensor_parallel_degree <= 1
Copy link
Collaborator

Choose a reason for hiding this comment

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

expert_tensor_parallel_degree 现在这几种 ep 的模式有空,最好统一一下,之前 gexiao 合入的时候说这个 expert_tensor_parallel_degree 好像一般不启用。

@sneaxiy sneaxiy merged commit 4213c30 into PaddlePaddle:incubate/paddlenlp-fleety_20250421 May 12, 2025
4 of 5 checks passed
@sneaxiy sneaxiy deleted the support_ep_reshard branch May 12, 2025 07:40
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