Skip to content
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.

fix OOM error during resource allocation #14444

Merged
merged 1 commit into from
Mar 16, 2019
Merged

Conversation

szha
Copy link
Member

@szha szha commented Mar 16, 2019

Description

fix OOM error during resource allocation for dropout state

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

  • Changes are complete (i.e. I finished coding on this PR)
  • Code is well-documented:
  • To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

  • alloc and direct free for cudnn dropout state space

Comments

Copy link
Member

@wkcn wkcn left a comment

Choose a reason for hiding this comment

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

Thank you for the fix!
LGTM

@wkcn
Copy link
Member

wkcn commented Mar 16, 2019

Is there any other state_space which need to reserve memory?
Could we add a function to reserve memory for the memory pool?

@szha
Copy link
Member Author

szha commented Mar 16, 2019

@wkcn I think most other cases are already handled manually (e.g. cudnn CNN/RNN). It might be worth a refactor for this though it's not urgent.

@szha
Copy link
Member Author

szha commented Mar 16, 2019

And thanks for the quick review!

@szha szha merged commit f602b0d into apache:master Mar 16, 2019
@szha szha deleted the cudnn_reserve branch March 16, 2019 21:24
vdantu pushed a commit to vdantu/incubator-mxnet that referenced this pull request Mar 31, 2019
haohuanw pushed a commit to haohuanw/incubator-mxnet that referenced this pull request Jun 23, 2019
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants