[MXNET-1017] Updating the readme file for cpp-package and adding readme file for example directory. #12737

leleamol · 2018-10-04T18:38:16Z

Description

Update the README file for cpp-package with build instructions and provide new README for example directory.

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

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

remove BaseTestNext and bump Compat

remove usr/setupenv.cmd because it is too invasive

* Fix for symbolic node * wrong indentation * Better condition

normalized mean squared error

add a new metric: NMSE

* Takes eps parameter to prevent log(0) to blow up, * Recognises learning of probability distributions.

…che#223)

ref: apache#407 [ci skip]

docs: update renamed `every_n_batch`

…utput. Fixes apache#213

Check if key stride exists in the node info when producing graphviz output

add link to Jupyter notebooks

fix conflict with Images

fixes bilinear initializer following approach in apache#34

…nts (apache#12711) Default numpy version in The Python Package Index (PyPI) is 1.15.2

* reenable the test * Trigger CI

* update the installation document * fix minor text * update the rename process * fix wording * remind users. of env and vs version * leave only the required dll * fix the link * update the R anchor * refine the description of step 7 * add missing . * fix spelling * update links and fix wording

* [apache#12345] Enabling two tests in the Straight Dope Nightly. Two straight dope notebook tests were disabled due to a timeout so they were disabled. I've updated one of the notebooks (rnn-gluon) to use the gpu instead of the cpu so it takes ~ 5 minutes on a p3.2xl, and verified the other notebook takes a minute and was a false alarm (visual-qa). The PR in the Straight Dope is: zackchase/mxnet-the-straight-dope#540 * Add dependency for IPython update. * Detect errors in notebook execution failure. * Clean up of naming in retry code.

Fixes apache#10977

…ensorcore for fp32 (apache#12722)

…12739) * * Added randn function * Internal SELU function on C++ layer * Predict now accepts ndarray as well * Gluon: Only warn when the blocks are unregistered. * Better sparse support for gluon * Gpu memory info via mxnet api call. * Gluon: Improved block summary. * Added validation docs for MXNet installation for Perl. * Flexible perl env for examples. * Gluon: Custom dtypes for the symbol block * Separate eval metric for the epoch level. * fixed typo.

…apache#12636) * Adding the example to demonstrate the usage of CSVIter * Addressed the review comments to make the example configurable. Moved the unittests folder in 'examples' directory. * Updated the code to address the cpp lint errors. * Removed the author tag. * Fixing the lint errors and usage message. * Update README file for cpp-package and provide README file for example directory. * Revert "Update README file for cpp-package and provide README file for example directory." This reverts commit 02e784a. These files were part of fix for JIRA issue 1017. These files were mistakenly committed in this PR. * Addressed the review comments regarding usage of atoi and avoiding string copy. * Updated to use strtol instead of atoi

* Implement ctc_loss as a normal operator * Update unit test * Update unit test and fix bug in backward * fix lint error * refactoring * Fix compilation error in CUDA * Fix CPU compilation error * Move ctc_include to nn folder and refactor * temporarily disable lint on 3rd party includes * move ctc_include to 3rdparty * remove contrib ctc_loss operator * revert a change by mistake * Fix a bug in kDevCPU * revert change by mistake * add alias to make it backward compatible * add unit test for backward compatibility * linting

* add resnet50-v1 to benchmark_score * rename back and duplicated * rename v2 back to resnet.py

…he#12758) * reflect the PR * add 1 more metric

aaronmarkham

I found a few spelling mistakes that could be caught by a spell checker, so you might want to do that in case I missed something.
Do all of the examples work now?

aaronmarkham · 2018-10-09T16:19:03Z

cpp-package/README.md

+
+###Steps to build the C++ package:
+1.  Building the MXNet C++ package requires building MXNet from source.
+2.  Clone the MXNet github repository **recursively** to ensure the code in submodules is available for building MXNet.


Why not include the full clone command here?

Sure will do.

aaronmarkham · 2018-10-09T16:20:24Z

cpp-package/README.md

+
+## Usage
+
+In order to consume the C++ API please follow the steps below


end with . or :

aaronmarkham · 2018-10-09T16:20:59Z

cpp-package/README.md

+	```
+	#include <mxnet-cpp/MxNetCpp.h>
+	```
+3. While building the program, ensure that the correct paths to the directories containing header files and MxNet shared library.


aaronmarkham · 2018-10-09T16:29:36Z

cpp-package/README.md

+	#include <mxnet-cpp/MxNetCpp.h>
+	```
+3. While building the program, ensure that the correct paths to the directories containing header files and MxNet shared library.
+4. The program links MxNet shared library dynamically. Hence the library needs to be accessible to the program during the runtime. This can be achieved by including the path to shared library to environment variable such as LD_LIBRARY_PATH.


links the MXNet...
during runtime...
the shared library in the environment variable LD_LIBRARY_PATH.

Also, doesn't this end up in a known place? Can't we tell them the path for linux, mac, windows?

Yes we can suggest platform specific variables.

aaronmarkham · 2018-10-09T16:32:05Z

cpp-package/example/README.md

+
+## Building C++ examples
+
+The examples are built while building the MXNet library and cpp-package from source . However, they can be built manually as follows


Really? I had to build them manually.
...from source. However,

Yes. I noticed it when I built it on mac and under ubuntu docker. The executables are kept under ./build/cpp-package/example directory.

aaronmarkham · 2018-10-09T16:50:00Z

cpp-package/example/README.md

+
+### [resnet.cpp](<https://github.com/apache/incubator-mxnet/blob/master/cpp-package/example/resnet.cpp>)
+
+The code implements resnet model using C++ API. The model is used to train MNIST data. The number of epochs for training the model can be specified on the command line. By default, model is trained for 100 epochs.


implements a ResNet... using the C++...
End with:
For example, to train with 10 epochs use the following:

aaronmarkham · 2018-10-09T16:50:57Z

cpp-package/example/README.md

+
+### [lenet.cpp](<https://github.com/apache/incubator-mxnet/blob/master/cpp-package/example/lenet.cpp>)
+
+The code implements lenet model using C++ API. It uses MNIST training data in CSV format to train the network. The example does not use built-in CSVIter to read the data from CSV file. The number of epochs can be specified on the command line. By default, the mode is trained for 100000 epochs.


implements a LeNet... using the C++...
the model... 100,000 (use a comma)

aaronmarkham · 2018-10-09T16:52:31Z

cpp-package/example/README.md

+```
+### [lenet\_with\_mxdataiter.cpp](<https://github.com/apache/incubator-mxnet/blob/master/cpp-package/example/mlp_cpu.cpp>)
+
+The code implements lenet model using C++ API. It uses MNIST training data to train the network. The example uses built-in MNISTIter to read the data. The number of epochs can be specified on the command line. By default, the mode is trained for 100 epochs.


implements a LeNet... using the C++...
the model... 100,000 (use a comma)
For example, to train with 10 epochs use the following:

Also, why backslashes?

aaronmarkham · 2018-10-09T16:53:03Z

cpp-package/example/README.md

+./lenet\_with\_mxdataiter 10
+```
+
+In addition, there is `run_lenet_with_mxdataiter.sh` that downloads the mnist data and run `lenet_with_mxdataiter` example.


MNIST
and runs the...

aaronmarkham · 2018-10-09T16:53:49Z

cpp-package/example/README.md

+
+###[inception_bn.cpp](<https://github.com/apache/incubator-mxnet/blob/master/cpp-package/example/inception_bn.cpp>)
+
+The code implements Inception network using C++ API with batch normalization. The example uses MNIST data to train the network. The model trains for 100 epochs.


an Inception... using the C++...
The example can be run by executing the following command:

* Implement mkldnn convolution fusion. Implement mkldnn convolution quantization. * Fix lint * Fix performance regression caused by mkldnn fallback. * clean up include * Fix msbuild on openmp pragma. * Fix quantization test, allow to use original op names as exclude layer for quantization. * Fix unittest. * Fix unittest * fix lint * Add post quantize fusion * add test case * add head license in test case * Remove GetBoolHash() * Remove mkldnn fallback change. * Address Haibin's comments. * Add TIsMKLDNN for _sg_mkldnn_conv temporarily. * Address reminisce's comments. * Handle the case that inplace fail. * pass unit test. * Add symbol api get_backend_symbol() * Retrigger ci * update the test case * Check subgraph index. * Use index as FAvoidQuantizeInput's parameter. * Add mkldnn_hwigo support as quantizaiton needs. * Address KellenSunderland's comments. * Handle input order change after subgraph pass. * Fix ci test

* Core Java API class commit * Update ScalaStyle max line length to 132 instead of 100

…xample directory.

leleamol · 2018-10-09T22:12:29Z

Something went wrong while merging the latest code from the branch. It resulted in including undesired files into this PR.
Closing this PR.

vchuravy and others added 30 commits January 28, 2017 15:52

Merge pull request apache#192 from dmlc/vc/review_comments

d59b214

remove BaseTestNext and bump Compat

remove usr/setupenv.cmd because it is too invasive

0ee926e

Merge pull request apache#196 from dmlc/vc/fix_win

79d80b6

remove usr/setupenv.cmd because it is too invasive

Fix for symbolic node (apache#199)

75060bc

* Fix for symbolic node * wrong indentation * Better condition

get_name (apache#206)

f33f1dc

fix data URL error (apache#204)

17a5442

fix for issue apache#207 (apache#208)

c58eb46

metric: intro NMSE

fa25ad6

normalized mean squared error

doc: update CDN link of MathJax

f23b0a9

metric: add test cases for NMSE

96191ea

docs: add an explanation for normalized MSE

ba97d93

Merge pull request apache#211 from iblis17/nmse

09d8f27

add a new metric: NMSE

Fixed data attribute

cf7f32b

switch to Plots.jl from pyplot (apache#217)

c9eb71b

Update ACE metric (apache#220)

1781290

* Takes eps parameter to prevent log(0) to blow up, * Recognises learning of probability distributions.

provide an optional context in mx.load_checkpoint (apache#221)

9474d95

Added initializers for Spatial Transformer Networks (STN) layers (apa…

46440e3

…che#223)

docs: update renamed every_n_batch

9dd3291

ref: apache#407 [ci skip]

Merge pull request apache#225 from iblis17/docs

fc2974f

docs: update renamed `every_n_batch`

Check if key stride exists in the node info when producing graphviz o…

5c02fcb

…utput. Fixes apache#213

Merge pull request apache#226 from facundoq/master

867b1c5

Check if key stride exists in the node info when producing graphviz output

Update mnist.md (apache#227)

31b95fb

add link to Jupyter notebooks

fix conflict with Images

1be1b87

Merge pull request apache#228 from dmlc/vc/images_subtype

7a6120a

fix conflict with Images

fixes bilinear initializer following approach in apache#34

bfa966a

adds test for bilinear initializer

8947ead

Merge pull request apache#229 from dmlc/vc/bilinear_initializer

44a6c36

fixes bilinear initializer following approach in apache#34

add SeqMetric to apply different metrics for multi-output

558adb7

add NullMetric

b4ece81

fix pointer(NDArray) for Float64

ced3f73

apivovarov and others added 13 commits October 5, 2018 14:10

Change numpy version to 1.15.2 in python and docker install requireme…

d6dfd41

…nts (apache#12711) Default numpy version in The Python Package Index (PyPI) is 1.15.2

Reenable test_gluon.test_conv (apache#12718)

857cea1

* reenable the test * Trigger CI

Fix failing GPU test on single GPU host (kvstore) (apache#12726)

836ba78

Fixes apache#10977

Add option for automatic downcasting dtype for cudnn to allow using T…

5314cf4

…ensorcore for fp32 (apache#12722)

fix benchmark on control flow operators. (apache#12693)

077253d

Add new name to CONTRIBUTORS.md (apache#12763)

167355b

Add resnet50-v1 to benchmark_score (apache#12595)

3c3506f

* add resnet50-v1 to benchmark_score * rename back and duplicated * rename v2 back to resnet.py

[MXNET-716][MIRROR apache#12723] Scala Benchmark Extension pack (apac…

f0140b3

…he#12758) * reflect the PR * add 1 more metric

aaronmarkham suggested changes Oct 9, 2018

View reviewed changes

ZhennanQin and others added 7 commits October 9, 2018 10:38

Introduction to Clojure-MXNet video link. (apache#12754)

4968900

[MXNET-915] Java Inference API core wrappers and tests (apache#12757)

bcbac41

* Core Java API class commit * Update ScalaStyle max line length to 132 instead of 100

Disabled flaky test: test_mkldnn.test_Deconvolution (apache#12770)

c98b19e

Updating the readme file for cpp-package and adding readme file for e…

90fbbd7

…xample directory.

Addressed the review comments.

9ee38de

Meged the latest chages.

2776712

leleamol requested review from anirudh2290, gigasquid, marcoabreu, sergeykolychev, szha and yzhliu as code owners October 9, 2018 22:08

leleamol closed this Oct 9, 2018

leleamol mentioned this pull request Oct 10, 2018

[MXNET-1017] Updating the readme file for cpp-package and adding readme file for example directory. #12773

Merged

4 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[MXNET-1017] Updating the readme file for cpp-package and adding readme file for example directory. #12737

[MXNET-1017] Updating the readme file for cpp-package and adding readme file for example directory. #12737

leleamol commented Oct 4, 2018

aaronmarkham left a comment

aaronmarkham Oct 9, 2018

leleamol Oct 9, 2018

aaronmarkham Oct 9, 2018

aaronmarkham Oct 9, 2018

aaronmarkham Oct 9, 2018

aaronmarkham Oct 9, 2018

leleamol Oct 9, 2018

aaronmarkham Oct 9, 2018

leleamol Oct 9, 2018

aaronmarkham Oct 9, 2018

aaronmarkham Oct 9, 2018

aaronmarkham Oct 9, 2018

aaronmarkham Oct 9, 2018

aaronmarkham Oct 9, 2018

leleamol commented Oct 9, 2018


		## Usage

		In order to consume the C++ API please follow the steps below


		## Building C++ examples

		The examples are built while building the MXNet library and cpp-package from source . However, they can be built manually as follows


		### [resnet.cpp](<https://github.com/apache/incubator-mxnet/blob/master/cpp-package/example/resnet.cpp>)

		The code implements resnet model using C++ API. The model is used to train MNIST data. The number of epochs for training the model can be specified on the command line. By default, model is trained for 100 epochs.


		### [lenet.cpp](<https://github.com/apache/incubator-mxnet/blob/master/cpp-package/example/lenet.cpp>)

		The code implements lenet model using C++ API. It uses MNIST training data in CSV format to train the network. The example does not use built-in CSVIter to read the data from CSV file. The number of epochs can be specified on the command line. By default, the mode is trained for 100000 epochs.


		###[inception_bn.cpp](<https://github.com/apache/incubator-mxnet/blob/master/cpp-package/example/inception_bn.cpp>)

		The code implements Inception network using C++ API with batch normalization. The example uses MNIST data to train the network. The model trains for 100 epochs.

[MXNET-1017] Updating the readme file for cpp-package and adding readme file for example directory. #12737

[MXNET-1017] Updating the readme file for cpp-package and adding readme file for example directory. #12737

Conversation

leleamol commented Oct 4, 2018

Description

Checklist

Essentials

Changes

Comments

aaronmarkham left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

leleamol commented Oct 9, 2018