Adapt fully_shard
PR Category
User Experience
PR Types
Others
Description
- Add support for the
fully_shardAPI in PaddleFleet - Support both
manual_parallelandauto_parallel. - The
fully_shardAPI no longer accepts theoptimizer. - Check on GLM (manual_parallel) and llama3 (auto_parallel) using
fully_shardAPI in PaddleFormers, check same loss, memory, performance compared to the original sharding stage3.
card-92763
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.
Codecov Report
:x: Patch coverage is 89.97429% with 39 lines in your changes missing coverage. Please review.
:warning: Please upload report for BASE (develop@5e27f87). Learn more about missing BASE report.
:x: Your patch status has failed because the patch coverage (89.97%) is below the target coverage (90.00%). You can increase the patch coverage or adjust the target coverage.
Additional details and impacted files
@@ Coverage Diff @@
## develop #76770 +/- ##
==========================================
Coverage ? 89.97%
==========================================
Files ? 4
Lines ? 389
Branches ? 0
==========================================
Hits ? 350
Misses ? 39
Partials ? 0
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
:rocket: New features to boost your workflow:
- :snowflake: Test Analytics: Detect flaky tests, report on failures, and find test suite problems.