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This repository has been archived by the owner on Nov 17, 2023.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

Silence excessive mkldnn logging output on tests. #14947

Merged
merged 2 commits into from
May 18, 2019

Conversation

larroy
Copy link
Contributor

@larroy larroy commented May 14, 2019

Description

Silenced excessive logging output:

http:https://jenkins.mxnet-ci.amazon-ml.com/blue/rest/organizations/jenkins/pipelines/mxnet-validation/pipelines/unix-cpu/branches/PR-14940/runs/1/nodes/283/steps/749/log/?start=0

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

  • The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant JIRA issue created (except PRs with tiny changes)
  • Changes are complete (i.e. I finished coding on this PR)
  • All changes have test coverage:
  • Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
  • Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore)
  • Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL)
  • Code is well-documented:
  • For user-facing API changes, API doc string has been updated.
  • For new C++ functions in header files, their functionalities and arguments are documented.
  • For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable
  • Check the API doc at http:https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$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

Copy link
Contributor

@pengzhao-intel pengzhao-intel left a comment

Choose a reason for hiding this comment

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

LGTM

Minor question, is this only for CI purpose?

@@ -553,8 +553,9 @@ void OpCheck::Run(mxnet::FCompute fn, const nnvm::NodeAttrs &attrs,
for (size_t i = 0; i < out_blobs.size(); i++)
out_blobs[i] = outputs[i].data();
fn(attrs, ctx, in_blobs, req, out_blobs);

#ifndef NDEBUG
Copy link
Member

Choose a reason for hiding this comment

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

Thank you for fixing this. How about changing to check the environmental variable MXNET_MKLDNN_DEBUG? We used this variable in other places.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

@TaoLv sure, will do that.

Copy link
Member

Choose a reason for hiding this comment

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

Thank you @larroy . We're thinking about adding verbose levels to MXNET_MKLDNN_DEBUG in the future.

@larroy
Copy link
Contributor Author

larroy commented May 15, 2019

@pengzhao-intel I saw this excessive log output and had a 3h timeout on tests, wanted to see if the PR validation time goes down because of this, probably not related.

@TaoLv
Copy link
Member

TaoLv commented May 16, 2019

@larroy Sorry just notice that you're trying to save time of CI? I remember MXNET_MKLDNN_DEBUG is turned on explicitly in CI. So my suggestion might not help for this case.

@pengzhao-intel
Copy link
Contributor

@larroy Sorry just notice that you're trying to save time of CI? I remember MXNET_MKLDNN_DEBUG is turned on explicitly in CI. So my suggestion might not help for this case.

How about turning it off because we have enough test cases in the CI for MKLDNN now?

@pengzhao-intel
Copy link
Contributor

Thank you for your improvement. Merge now.

@pengzhao-intel pengzhao-intel merged commit 3d6646a into apache:master May 18, 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

3 participants