Skip to content

Conversation

@A-nnonymous
Copy link
Contributor

PR Category

Operator Mechanism

PR Types

Improvements

Description

cherry-pick linear_v2 alignment into release 3.3
devPR:#77039
pcard-91067

* aligned linear & matmul_with_bias, need check incompatible cases.

* aligned high-dim matmul operation, adding flags control and controlling failed cases

* aligned matmul, start aligning einsum

* Add flag

* polish

* fix shape related issues.

* Finish crash handling

* revert redundant diff

* revert redundant diff

* restrict influence to only CUDA

* Optimized CPU overhead, bypass windows.

* optimize branch cost

* disable dist tensor case

* add GPUPlace check

* fix matmul diff

* add flags related logic

* polish

* polish

* polish

* polish

* polish

* polish

* polish

* polish

* bypass win32

* stash

* stash

* align fwd in gpu

* fix fwd miscs

* fix shape miscs

* clean code

* fix grad

* recover redundant diff

* using legacy linear

* add multi-platform support, polish

* refractor

* fix flag and amp rules

* fix CI

* polish

* fix miscs

* add infersymbolics

* Add metaconfig

* fix symbolic, move flags

* fix bwd infermeta

* Add prim linear_v2_grad

* fix fwd decomp

* add proper fallback to fulfill legacy promise

* tmp restrict prim

* Add inferSPMD, fix CI

* fix ci

* fix auto parallel.

* fix TRT and prec test

* add infer_symbolic instance, remove glog including in header

* fix reduncant DtoD cpy

* coverage linear_v2 symbolics
@paddle-bot
Copy link

paddle-bot bot commented Jan 15, 2026

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.

@A-nnonymous A-nnonymous changed the title Align linear op with torch v2.9.1 (#77039) Cherry-Pick Align linear op with torch v2.9.1 (#77039) Jan 15, 2026
@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

3 similar comments
@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

@codecov-commenter
Copy link

Codecov Report

❌ Patch coverage is 98.57820% with 3 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (release/3.3@d00b579). Learn more about missing BASE report.

Files with missing lines Patch % Lines
paddle/phi/infermeta/spmd_rules/linear_v2.cc 98.68% 1 Missing ⚠️
paddle/phi/kernels/cpu/linear_v2_kernel.cc 90.90% 1 Missing ⚠️
python/paddle/nn/functional/common.py 95.65% 1 Missing ⚠️
Additional details and impacted files
@@              Coverage Diff               @@
##             release/3.3   #77356   +/-   ##
==============================================
  Coverage               ?   98.57%           
==============================================
  Files                  ?        9           
  Lines                  ?      211           
  Branches               ?        0           
==============================================
  Hits                   ?      208           
  Misses                 ?        3           
  Partials               ?        0           

☔ 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.

Copy link
Contributor

@wanghuancoder wanghuancoder left a comment

Choose a reason for hiding this comment

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

LGTM

@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

5 similar comments
@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

@A-nnonymous
Copy link
Contributor Author

/re-run all-failed

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