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This repository has been archived by the owner on Nov 17, 2023.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

[DO NOT REVIEW] Numpy unary functions #14754

Closed
wants to merge 44 commits into from
Closed

[DO NOT REVIEW] Numpy unary functions #14754

wants to merge 44 commits into from

Conversation

haojin2
Copy link
Contributor

@haojin2 haojin2 commented Apr 21, 2019 •
edited
Loading

Description

As title.

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 numpy interfaces for existing unary functions
  • Corresponding consistency unit tests

Comments

Related Issue #14327

@haojin2 haojin2 added the Numpy label Apr 21, 2019
@haojin2 haojin2 requested a review from reminisce April 21, 2019 23:49
@haojin2 haojin2 self-assigned this Apr 21, 2019
@haojin2 haojin2 force-pushed the numpy_ops branch 3 times, most recently from b1067ec to 5554146 Compare April 22, 2019 20:48
@reminisce reminisce mentioned this pull request Apr 23, 2019
Closed
@perdasilva @marcoabreu
Update base CUDA image for CI to v10.0 cuDNN 7.3.1 (apache#14513)
36c3306 
* Updates Ubuntu GPU CI image base image to cuda10-devel and manually installs cuDNN version 7.3.1.20

* Updates CentOS 7 GPU CI image base image to cuda10-devel and manually installs cuDNN version 7.3.1.20
reminisce and others added 10 commits May 3, 2019 23:28
@reminisce @eric-haibin-lin
Fix sample_multinomial number of outputs bug (apache#14873)
5ba285b 
* Fix sample_multinomial number of outputs bug

* Fix lint
@suyz526 @wkcn
Add the Gluon Implementation of Deformable Convolution (apache#14810)
4d7bae1 
* 	modified:   docs/api/python/gluon/contrib.md
	modified:   python/mxnet/gluon/contrib/__init__.py
	new file:   python/mxnet/gluon/contrib/cnn/__init__.py
	new file:   python/mxnet/gluon/contrib/cnn/conv_layers.py
	new file:   tests/python/gpu/test_gluon_contrib_gpu.py

* 	modified:   python/mxnet/gluon/contrib/cnn/conv_layers.py

* 	modified:   python/mxnet/gluon/contrib/cnn/conv_layers.py

* Update conv_layers.py

* Update conv_layers.py
@ZhennanQin @pengzhao-intel
Refactor ImageRecordIter (apache#14824)
621b391 
* cpu optimized data loader

* Fix CI

* Fix CI

* Fix ci

* Fix doc
@anirudh2290 @wkcn
Prevent crashes for opencv exception and std::exception (apache#14433)
25ba1d1 
* Relax constexpr restriction

* Change imagenet_gen_qsym_mkldnn

* Revert constexpr change

* Add dummy image for test

* Fix unreverted change

* Remove url

* Fix imagenet change

* Catch only std::exception

* Fix const, remove dmlc::Error catch

* Retrigger CI
@lobanov-m @wkcn
[MXNet-1211] Factor and "Like" modes in BilinearResize2D operator (ap…
b30949f 
…ache#13226)

* Added "factor" and "like" modes into BilinearResize2D operator.
Also added tests and some fixes to visualization needed due to added modes.

* Lint fix

* Test fix

* retrigger CI

* retrigger CI

* Retrigger CI

* retrigger CI

* retrigger ci

* retrigger ci again
@gigasquid
[Clojure] Add Fine Tuning Sentence Pair Classification BERT Example (a…
2113cb7 
…pache#14769)

* Rework Bert examples to include QA infer and finetuning

* update notebook example and exported markdown

* add integration test for the classification

* fix tests

* add RAT

* add another RAT

* fix all the typos

* Clojure BERT finetuning example: fix CSV parsing

* update readme and gitignore

* add fix from @davliepmann’s notebook parsing

* feedback from @daveliepmann

* fix running of example and don’t show very first batch on callback speedometer

* rerun the notebook and save results

* remove bert stuff from main .gitignore

* re-putting the license back on after regen

* fix integration test
@sxjscience @eric-haibin-lin
[Bugfix] Fix layer norm for large input shape (apache#14870)
f6ef206 
* fix layer norm for large input shape

* try to fix

* use a larger eps

* try to fix test

* try to fix
@arcadiaphy
rewrite test_custom_op_exc (apache#14878)
42ede50
@stu1130 @lanking520
[Dependency Update] Upgrade openssl to 1.1.1b (apache#14837)
a722db4 
* upgrade openssl to 1.1.0k

* fix the wrong version

* upgrade to the latest version

* avoid lib to be treated as file instead of folder
@stu1130 @szha
[Dependency Update] Upgrade cuDNN & NCCL (apache#14884)
0255dd6 
* upgrade cuDNN & NCCL

* retrigger CI
sandeep-krishnamurthy and others added 20 commits May 9, 2019 16:12
@sandeep-krishnamurthy
Add API documentation for upsampling operator with examples (apache#1…
874fb89 
…4919)

* Add API documentation for upsampling operator with examples

* Update src/operator/nn/upsampling.cc

Co-Authored-By: Aaron Markham <markhama@amazon.com>

* Update src/operator/nn/upsampling.cc

Co-Authored-By: Aaron Markham <markhama@amazon.com>

* Update src/operator/nn/upsampling.cc

Co-Authored-By: Aaron Markham <markhama@amazon.com>

* Make API doc example as pseudocode than code
@perdasilva @marcoabreu
Pins version of scikit-learn for python2 due to drop in support (apac…
f4598e7 
…he#14928)
@zachgk @lanking520
Deprecate NDArrayCollector and instead use ResourceScope (apache#14780)
4796851
@KellenSunderland
[MXNET-857] Add initial NVTX profiler implementation (apache#12328)
b22ee95 
* [MXNET-857] Enable CUDA NVTX extensions for profiler

These extensions mark readable ranges in the NVIDIA Visual Profiler which helps
show correlations between kernel launches and graph node executions.

Example shown here: https://user-images.githubusercontent.com/7443219/33946110-34296d18-e021-11e7-8d18-6d40b797405c.png
The additional information enabled is in the 'Markers and Ranges' row.

* [MXNET-857] Add initial NVTX profiler implementation

This commit removes NVTX headers from the Amalgamation build process,
but this is a CUDA/CMake only feature, so it's not relevant to
Amalagamation builds.

* [MXNET-857] Use macro for NVTX specific code

* [MXNET-857] Add integration test.

* Turn on NVTX by default in Unix.

* Fixed typos and added NTVX info to profiler.md

* Add NVTX example to profiling tutorial

* Add NVTX flags for make
@stu1130 @szha
upgrade the version to 2.0.2 (apache#14621)
8a4ad9f
@perdasilva @wkcn
Updates to cudnn package installation (apache#14923)
669ab2c 
* Generalizes centos7 cudnn download and install script

* Updates setting of cudnn version to a position in the Dockerfile that will have the least impact on caching
@junrushao @zheng-da
[MXNET-1352] Allow dynamic shape in while_loop and if conditionals (a…
d577b6f 
…pache#14393)

* Initial commit

* Rebase

* WIP for fixing rebase issues

* WIP for fixing rebase issues

* fix wip

* wip fix

* wip fix

* wip fix

* wip fix

* wip fix

* wip fix

* should be good to go

* wip remove debug info

* wip remove debug info

* linter

* linter

* Retrigger

* Address comments from Da
@arcadiaphy
Add numpy linspace (apache#14927)
f67d067 
* add linspace operator

* add test

* fix bug

* register gpu op

* fix lint
@TaoLv @pengzhao-intel
Fix reshape to add in-place back (apache#14903)
13d6ee6 
* fix reshape

* save

* fix reshape

* clean code

* fix memory allocation & lint

* add unit test

* req type

* add comments to describe the logic
@larroy @wkcn
use mx.context.num_gpus instead of mx.test_utils.list_gpus in MF reco…
3122548 
…mmender example (apache#14926)
@ZhennanQin @pengzhao-intel
Re-enable static cached_op optimization (apache#14931)
1eba37a
@perdasilva @marcoabreu
Disables TensorRT build step (apache#14958)
99f5f66
@perdasilva @eric-haibin-lin
Fixed and re-enables TensorRT steps (apache#14960)
f7b7163 
* Pins libnvinfer versions

* Sets cudnn to version 7.5.0 in tensorrt environment

* Re-enables TensorRT stages
@haojin2
[Do not review] [Do not merge] New numpy-compatible sum (apache#14739)
b1a9b79 
* Add numpy namespace and initial impl of np.sum (not complete)

* Clean up

* Fix import error

* numpy sum

* add test and backward data type support

* add license to test_numpy_op.py

* improve test to reduce flakiness

* fix sanity build

* extra numeric test and imperative test

* add error message for initial argument
@reminisce @haojin2
[numpy] Infra for supporting numpy ops in imperative mode and Gluon A…
831fe6c 
…PIs (apache#14758)

* Infra of new ndarray and symbol types for numpy operators

* Rename

* Fix import problem

* Refactor

* Remove redundant code

* Add docstring

* More on numpy ndarray and symbol

* Override unimplemented methdos for ndarray and _NumpySymbol

* Fix built-in methods of ndarray and _NumpySymbol

* Fix test and sanity check

* Fix pylint

* Address cr comments

* Add unit tests for ndarray and _NumpySymbol

* Add _true_divide

* Fix gpu build

* Add future import division

* More correct way of checking if an output is from a np compat op

* Fix gpu build

* Fix output ndarray/symbol types with at least one new ndarray/symbol

* Modify true_divide doc

* Fix flaky copying zero-size arrays via gpus

* Fix zero size in gluon hybridize and zeros/ones symbol not creating new symbol type

* Fix doc
@reminisce @haojin2
Enable np op compat check with name prefix (apache#14897)
3e93422
@haojin2
[numpy] Numpy dot (apache#14831)
647751c 
* Numpy Dot case 1-4 + case 3.5 forward and 0.5 backward

* Backward computation and test coverage
@haojin2
add numpy interface for all unary funcs
3bd9027
@haojin2
add numpy.clip and test_np_clip
dd214e4
@haojin2
move to numpy folder and re-declare
0408628
@haojin2 haojin2 closed this May 15, 2019
@haojin2 haojin2 deleted the numpy_ops branch May 15, 2019 23:19
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Reviewers

@reminisce reminisce Awaiting requested review from reminisce

@anirudh2290 anirudh2290 Awaiting requested review from anirudh2290

@gigasquid gigasquid Awaiting requested review from gigasquid

@nswamy nswamy Awaiting requested review from nswamy

@szha szha Awaiting requested review from szha

@yzhliu yzhliu Awaiting requested review from yzhliu

Assignees

@haojin2 haojin2

Labels
Numpy
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.