Skip to content
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

Refine the activation type getting in the LSTM operator to speed. #6996

Merged
merged 2 commits into from
Dec 26, 2017
Merged

Refine the activation type getting in the LSTM operator to speed. #6996

merged 2 commits into from
Dec 26, 2017

Conversation

qingqing01
Copy link
Contributor

Fix #6841

  • Time: collected by yep + pprof based the config of sentiment analysis

    • forward: LSTMUnitFunctor: 9.52s -> 7.95s
    • backward: LSTMGradUnitFunctor: 5.60s -> 3.68s

  • Refine the code style for the activation type

    • But the old code is not removed, since it is used in GRU operator. This will be fixed in next PR.

  • Still use std::string type in the attribute definition.

    • By using the int type in the attribute definition, the time is close with the std::string type.

reyoung
reyoung previously approved these changes Dec 25, 2017
View reviewed changes
Copy link
Collaborator

@reyoung reyoung left a comment

Choose a reason for hiding this comment

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

Cool

reyoung
reyoung approved these changes Dec 26, 2017
View reviewed changes
Copy link
Collaborator

@reyoung reyoung left a comment

Choose a reason for hiding this comment

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

Excellent

@qingqing01 qingqing01 merged commit f839154 into PaddlePaddle:develop Dec 26, 2017
@qingqing01 qingqing01 mentioned this pull request Dec 27, 2017
Closed
@qingqing01 qingqing01 deleted the lstm_active_type branch November 14, 2019 05:26
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@reyoung reyoung reyoung approved these changes

@chengduoZH chengduoZH Awaiting requested review from chengduoZH

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

Bad performance of ActiveType in lstm_op
2 participants
@qingqing01 @reyoung