[Dy2St] Add a hook to replace run impl for partial program layer by SigureMo · Pull Request #73081 · PaddlePaddle/Paddle
Navigation Menu
{{ message }}
PaddlePaddle / Paddle Public
- Notifications You must be signed in to change notification settings
- Fork 6k
Merged
Merged
Conversation
Copy link
Member
SigureMo
commented
Jun 4, 2025
SigureMo
commented
PR Category
Execute Infrastructure
PR Types
New features
Description
添加一个 hook 用来允许替换掉 run_program_op 逻辑,为动转静流程中插入 cuda graph 实现做准备
初步设计如下:
- 前几轮 warmup,将
run_impl替换为 capture 实现,用于收集动态 shape 的 CUDA Graph,以 batch size 为 cache key,存入 graph cache - 后续将
run_impl替换为 replay 实现,查找 cache,调用 CUDA Graph 的 replay
Copy link
paddle-bot
bot
commented
Jun 4, 2025
paddle-bot bot commented
Jun 4, 2025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.
SigureMo added 2 commits
June 9, 2025 10:55…artial-program-layer
zhangbo9674 approved these changes Jun 12, 2025
zyfncg approved these changes Jun 12, 2025
SigureMo
deleted the
dy2st/add-a-hook-to-replace-run-impl-of-partial-program-layer
branch
DrRyanHuang pushed a commit to DrRyanHuang/Paddle that referenced this pull request
Jun 14, 2025This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters. Learn more about bidirectional Unicode characters
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment