Skip to content

[Eager Hook] Support GradientHook before running separate GradNode #39638

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.

Sign up for GitHub

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

Jump to bottom
Merged
merged 6 commits into from
Feb 17, 2022
Merged

[Eager Hook] Support GradientHook before running separate GradNode #39638

merged 6 commits into from
Feb 17, 2022

Conversation

@veyron95
Copy link
Contributor

@veyron95 veyron95 commented Feb 16, 2022 •
edited
Loading

PR types

Function optimization

PR changes

Others

Describe

  1. Remove RunBackwardHooks() in backward.cc.
  2. Modify the auto code generator logic to support GradientHook before running separate GradeNode.
  3. Add related unite tests to make sure the intermediate GradNode in node.cc are real called and meet our expected hook result.
@paddle-bot-old
Copy link

paddle-bot-old bot commented Feb 16, 2022

Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.
jim19930609
jim19930609 previously approved these changes Feb 16, 2022
View reviewed changes
@veyron95 veyron95 changed the title [Eager] Support GradientHook before running seperate GradNode [Eager] Support GradientHook before running separate GradNode Feb 16, 2022
jim19930609
jim19930609 previously approved these changes Feb 17, 2022
View reviewed changes
@veyron95 veyron95 merged commit adf4b98 into PaddlePaddle:develop Feb 17, 2022
@veyron95 veyron95 changed the title [Eager] Support GradientHook before running separate GradNode [Eager Hook] Support GradientHook before running separate GradNode Feb 18, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

3 participants

@veyron95 @zyfncg @jim19930609