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

[MXNET-72] Improve sparse sgd on GPU #10293

Merged
eric-haibin-lin merged 2 commits into from
Apr 3, 2018
Merged

[MXNET-72] Improve sparse sgd on GPU #10293

eric-haibin-lin merged 2 commits into from
Apr 3, 2018

Conversation

@eric-haibin-lin
Copy link
Member

@eric-haibin-lin eric-haibin-lin commented Mar 28, 2018

Description

Apply the same optimization in #10062 for sparse sgd.
@ZiyueHuang @haojin2

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

  • The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant JIRA issue created (except PRs with tiny changes)
  • Changes are complete (i.e. I finished coding on this PR)
  • All changes have test coverage:
  • Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
  • Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore)
  • Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL)
  • Code is well-documented:
  • For user-facing API changes, API doc string has been updated.
  • For new C++ functions in header files, their functionalities and arguments are documented.
  • For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable
  • Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html
  • To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

  • Feature1, tests, (and when applicable, API doc)
  • Feature2, tests, (and when applicable, API doc)

Comments

  • If this change is a backward incompatible change, why must this change be made.
  • Interesting edge cases to note here

haojin2
haojin2 reviewed Mar 30, 2018
View reviewed changes
src/operator/optimizer_op-inl.h
mom_data[data_i] = momentum * mom_data[data_i]
- rate * weight_data[data_i]
- lr *
mshadow_op::clip::Map(rescale_grad * grad_data[i],
Copy link
Contributor

@haojin2 haojin2 Mar 30, 2018

Choose a reason for hiding this comment

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

Why a line break here?

haojin2
haojin2 reviewed Mar 30, 2018
View reviewed changes
src/operator/optimizer_op.cc
mom_data[data_i] = momentum * mom_data[data_i]
- rate * weight_data[data_i]
- lr *
mshadow_op::clip::Map(rescale_grad * grad,
Copy link
Contributor

@haojin2 haojin2 Mar 30, 2018

Choose a reason for hiding this comment

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

Why line break here?

Copy link
Member Author

@eric-haibin-lin eric-haibin-lin Mar 30, 2018

Choose a reason for hiding this comment

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

no particular reason. No new line will make it a 200 character line.

Copy link
Contributor

@haojin2 haojin2 Mar 30, 2018

Choose a reason for hiding this comment

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

I mean the extra line break for lr * mshadow_op::clip::Map, these two places are inconsistent with what you have on line 52 of optimizer_op.cu below.

Copy link
Member Author

@eric-haibin-lin eric-haibin-lin Apr 2, 2018 •
edited
Loading

Choose a reason for hiding this comment

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

Don't think it's necessary to add/remove that extra line break. Please provide constructive feedbacks/review comments

@eric-haibin-lin eric-haibin-lin merged commit 5245ef6 into apache:master Apr 3, 2018
rahul003 pushed a commit to rahul003/mxnet that referenced this pull request Jun 4, 2018
@eric-haibin-lin
[MXNET-72] Improve sparse sgd on GPU (apache#10293)
f60aacd 
* gpu kernels

* update warning msg
zheng-da pushed a commit to zheng-da/incubator-mxnet that referenced this pull request Jun 28, 2018
@eric-haibin-lin @zheng-da
[MXNET-72] Improve sparse sgd on GPU (apache#10293)
1d58117 
* gpu kernels

* update warning msg
@eric-haibin-lin eric-haibin-lin deleted the sgd branch November 12, 2018 20:00
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Reviewers

@cjolivier01 cjolivier01 Awaiting requested review from cjolivier01

2 more reviewers

@haojin2 haojin2 haojin2 left review comments

@ZiyueHuang ZiyueHuang ZiyueHuang approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@eric-haibin-lin @haojin2 @ZiyueHuang