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This repository has been archived by the owner on Nov 17, 2023.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

Add Intel MKL blas to Jenkins #13607

Merged
merged 4 commits into from
Dec 11, 2018
Merged

Add Intel MKL blas to Jenkins #13607

merged 4 commits into from
Dec 11, 2018

Conversation

xinyu-intel
Copy link
Contributor

@xinyu-intel xinyu-intel commented Dec 11, 2018 •
edited
Loading

Description

This is a PR to add Intel MKL to Jenkins.

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

  • Add mkl to ubuntu_cpu_mkldnn
  • Add mkl to ubuntu_cpu

Comments

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

@pengzhao-intel @marcoabreu @TaoLv
#13150

@pengzhao-intel
Copy link
Contributor

Thanks, Xinyu, Add more people in the loop to aware of the change.

@azai91 @lupesko @larroy @KellenSunderland

pengzhao-intel
pengzhao-intel reviewed Dec 11, 2018
View reviewed changes
ci/jenkins/Jenkinsfile_unix_cpu Outdated
@@ -45,6 +46,8 @@ core_logic: {
custom_steps.test_unix_python3_debug_cpu(),
custom_steps.test_unix_python2_mkldnn_cpu(),
custom_steps.test_unix_python3_mkldnn_cpu(),
custom_steps.test_unix_python2_mkldnn_mkl_cpu(),
Copy link
Contributor

Choose a reason for hiding this comment

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

Please add a test for non-MKLDNN with MKL, thanks.

marcoabreu
marcoabreu reviewed Dec 11, 2018
View reviewed changes
Copy link
Contributor

@marcoabreu marcoabreu left a comment

Choose a reason for hiding this comment

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

Great job, thank you!

ci/jenkins/Jenkins_steps.groovy
}]
}

def test_unix_python3_mkldnn_mkl_cpu() {
Copy link
Contributor

Choose a reason for hiding this comment

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

I don't think we need python 2 and 3

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Thanks, I'll leave python3 only. How to resolve the cache problem?

Copy link
Contributor

Choose a reason for hiding this comment

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

What cache problem?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

I find it will redownload and install mkl in unittest Python2: MKLDNN-MKL-CPU http://jenkins.mxnet-ci.amazon-ml.com/blue/rest/organizations/jenkins/pipelines/mxnet-validation/pipelines/unix-cpu/branches/PR-13607/runs/3/nodes/193/steps/479/log/?start=0

Copy link
Contributor

Choose a reason for hiding this comment

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

Yeah that's expected. The Docker cache will only activate after this PR has been merged into master. We do that to avoid growing the cache too heavily and instead only use the master state for caching.

In your case, ubuntu_mkl has been added by you and thus the cache diverges and a non-cached version gets used at that point. It works as intended :)

@xinyu-intel xinyu-intel changed the title [WIP]Add Intel MKL blas to Jenkins Add Intel MKL blas to Jenkins Dec 11, 2018
@marcoabreu marcoabreu merged commit e25e18f into apache:master Dec 11, 2018
@marcoabreu
Copy link
Contributor

Thanks a lot for your contribution!

@TaoLv TaoLv mentioned this pull request Dec 12, 2018
Merged
zhaoyao73 added a commit to zhaoyao73/incubator-mxnet that referenced this pull request Dec 13, 2018
@zhaoyao73
Merge remote-tracking branch 'upstream/master'
76e901a 
* upstream/master: (38 commits)
  Feature/mkldnn static (apache#13628)
  Fix the bug of BidirectionalCell (apache#13575)
  Set install path for libmxnet.so dynamic lib on Mac OS (apache#13629)
  add batch norm test (apache#13625)
  Scripts for building dependency libraries of MXNet (apache#13282)
  fix quantize pass error when the quantization supported Op are excluded in the model (apache#13596)
  Optimize C++ API (apache#13496)
  Fix warning in waitall doc (apache#13618)
  [MXNET-1225] Always use config.mk in make install instructions (apache#13364)
  [MXNET-1224]: improve scala maven jni build and packing. (apache#13493)
  [MXNET-1155] Add scala packageTest utility (apache#13046)
  fix the Float not showing correctly problem (apache#13617)
  apache#13385 [Clojure] - Turn examples into integration tests (apache#13554)
  Add Intel MKL blas to Jenkins (apache#13607)
  Revert "[MXNET-1198] MXNet Java API (apache#13162)"
  Reducing the length of setup tutorial (apache#13306)
  [MXNET-1182] Predictor example (apache#13237)
  [MXNET-1187] Added Java SSD Inference Tutorial for website (apache#13201)
  add defaults and clean up the tests (apache#13295)
  [MXNET-1181] Added command line alternative to IntelliJ in install instructions (apache#13267)
  ...
@anirudh2290 anirudh2290 mentioned this pull request Dec 13, 2018
Closed
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Reviewers

@pengzhao-intel pengzhao-intel pengzhao-intel left review comments

@marcoabreu marcoabreu marcoabreu left review comments

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
@xinyu-intel @pengzhao-intel @marcoabreu