Skip to content

Conversation

@xiaoguoguo626807
Copy link
Contributor

PR Category

Execute Infrastructure

PR Types

Improvements

Description

pcard-67164

解决PR65357 modulus模型性能下降问题anti_derivative-physics_informed/N1C1

该pr 解决其他模型由于静态分支和动态分支合并导致的性能问题。
本pr 将部分静态shape 分支和动态shape 分支合并,平衡命中各个分支的模型都能达到更好的性能。

同时修改了by_pass(非kernel输入) -> set_output

@paddle-bot
Copy link

paddle-bot bot commented Oct 11, 2024

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.

@xiaoguoguo626807 xiaoguoguo626807 merged commit b89eb26 into PaddlePaddle:develop Oct 14, 2024
26 checks passed
@xiaoguoguo626807 xiaoguoguo626807 deleted the modulus_perform branch October 14, 2024 06:10
changeyoung98 pushed a commit to changeyoung98/Paddle that referenced this pull request Oct 15, 2024
…ePaddle#68629)

* modify old ir put_along_axis api

* upate

* modify paddletest

* modify fetch input name

* combine reduce_as branch of dynamic and staic shape

* combine reduce_as branch of dynamic and staic shape

* modify divice_grad as phi kernel

* delete program check

---------

Co-authored-by: phlrain <[email protected]>
XiaoguangHu01 pushed a commit that referenced this pull request Oct 16, 2024
… (#68715)

* modify old ir put_along_axis api

* upate

* modify paddletest

* modify fetch input name

* combine reduce_as branch of dynamic and staic shape

* combine reduce_as branch of dynamic and staic shape

* modify divice_grad as phi kernel

* delete program check

---------

Co-authored-by: xiaoguoguo626807 <[email protected]>
Co-authored-by: phlrain <[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.

3 participants