[MXNET-121] Docs page for ONNX module. #10140

anirudhacharya · 2018-03-16T22:47:43Z

Description

Documentation page for ONNX module.

Checklist

Essentials

Passed code style checking (make lint)
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
To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

Documentation page for onnx module in contrib.

Comments

@reminisce @nswamy @Roshrini @spidydev

nswamy · 2018-03-16T22:56:27Z

docs/api/python/contrib/onnx.md

@@ -0,0 +1,111 @@
+# ONNX API


ONNX-MXNet Converter API ?
Please check the rendered output and fix the markdown

I think this can be created as a notebook, which users can use as a tutorial

nswamy · 2018-03-16T22:56:29Z

docs/api/python/contrib/onnx.md

+```
+
+```eval_rst
+.. note:: **Install ONNX** which needs protobuf compiler to be installed separately. Please **follow the instructions to install ONNX** [here](https://github.com/onnx/onnx).


mark down render check?

nswamy · 2018-03-16T22:56:43Z

docs/api/python/contrib/onnx.md

+.. note:: **Install ONNX** which needs protobuf compiler to be installed separately. Please **follow the instructions to install ONNX** [here](https://github.com/onnx/onnx).
+```
+
+This document describes the ONNX APIs in mxnet.


ONNX-MXNet API

This seems a bit confusing as we are not describing ONNX API's . May be something like
"This document describes APIs to support ONNX in MXNet"

nswamy · 2018-03-16T22:57:15Z

docs/api/python/contrib/onnx.md

+    mxnet.contrib.onnx.import_model
+```
+
+## Import ONNX model into Mxnet


Using ONNX Model in MXNet

nswamy · 2018-03-16T23:00:31Z

docs/api/python/contrib/onnx.md

+You can download the converted ONNX model from
+[here](https://s3.amazonaws.com/onnx-mxnet/examples/super_resolution.onnx).
+
+A pre-trained model in MXNet contains two elements: a symbolic graph, containing the model's network definition, and a binary file containing the model weights. You can import the ONNX model and get the symbol and parameters objects using "import_model" API as shown below:


bulleted points

Symbolic graph(with the extension .json) - containing the model's network definition

Binary file(with the extension .params) - containing model weights.
..

nswamy · 2018-03-16T23:01:39Z

docs/api/python/contrib/onnx.md

+
+```
+# By default, 'input_0' is an input of the imported model.
+mod = mx.mod.Module(symbol=sym, data_names=['input_0'], context=mx.cpu(), label_names=None)


link to Module API/tutorial page, similarly others where useful

ThomasDelteil · 2018-03-19T17:15:48Z

I am writing a tutorial on ONNX with MXNet, it would be nice if the docs show an example on how to load an ONNX model in Gluon using a symbolic block.

anirudhacharya · 2018-03-19T17:24:52Z

@ThomasDelteil Will include it. As discussed on github, we will also eventually have an API for gluon import and that API will have its own documentation and example illustration.

anirudhacharya · 2018-03-26T23:12:02Z

@aaronmarkham

anirudhacharya · 2018-03-29T01:28:46Z

This is how the page renders - http://34.227.67.197/api/python/contrib/onnx.html
Please review @aaronmarkham @sandeep-krishnamurthy @nswamy

aaronmarkham · 2018-03-29T16:49:30Z

docs/api/python/contrib/onnx.md

+
+## Overview
+
+[ONNX](https://onnx.ai/) is an open format to represent deep learning models. With ONNX as an intermediate representation, it is easier to move models between state-of-the-art tools and frameworks for training and inference..


Remove extra period

aaronmarkham · 2018-03-29T16:49:57Z

docs/api/python/contrib/onnx.md

+
+[ONNX](https://onnx.ai/) is an open format to represent deep learning models. With ONNX as an intermediate representation, it is easier to move models between state-of-the-art tools and frameworks for training and inference..
+
+The `mxnet.contrib.onnx` package refers to the APIs and interfaces that implements ONNX model format support for Apache MXNet.


aaronmarkham · 2018-03-29T16:51:25Z

docs/api/python/contrib/onnx.md

+
+The `mxnet.contrib.onnx` package refers to the APIs and interfaces that implements ONNX model format support for Apache MXNet.
+
+With ONNX format support for MXNet, developers can build and train models with PyTorch, CNTK, or Caffe2, and import these models into MXNet to run them for inference and training using MXNet’s highly optimized engine.


Simplify.
...train models with a variety of deep learning frameworks...
Then list the currently supported frameworks with bullets. (easier to maintain too, when we add more later)

How about this -
... train models with a variety of deep learning frameworks....
I have linked it to the supported tools page on the ONNX website. That way our page never gets outdated.

aaronmarkham · 2018-03-29T16:57:25Z

docs/api/python/contrib/onnx.md

+```
+
+```eval_rst
+.. note:: To use this module developers need to **install ONNX**, which needs protobuf compiler, to be installed separately. Please follow the instructions to install ONNX - https://github.com/onnx/onnx


Make this installation instructions.

aaronmarkham · 2018-03-29T16:58:38Z

docs/tutorials/onnx/super_resolution.md

+- learn how to load a pre-trained ONNX model file into MXNet.
+- run inference in MXNet.
+
+## Pre-requisite


plural, no dash

aaronmarkham · 2018-03-29T16:59:13Z

docs/tutorials/onnx/super_resolution.md

+- run inference in MXNet.
+
+## Pre-requisite
+The code demonstration assumes that the following python packages are installed:


This example assumes...

aaronmarkham · 2018-03-29T16:59:55Z

docs/tutorials/onnx/super_resolution.md

+sym, arg, aux = onnx_mxnet.import_model(onnx_model_file)
+```
+
+We can now visualize the imported model( graphviz needs to be installed)


space before ( not after

aaronmarkham · 2018-03-29T17:00:21Z

docs/tutorials/onnx/super_resolution.md

+mod.set_params(arg_params=arg, aux_params=aux, allow_missing=True, allow_extra=True)
+```
+
+Module API's forward method requires Batch of data as input. We will prepare the data in that format and feed it to the forward method.


lower case b

aaronmarkham

GTG!

sandeep-krishnamurthy

LGTM. Thanks for this documentation and tutorial.

@nswamy - Are your requested changes addressed?

sandeep-krishnamurthy · 2018-03-30T01:14:27Z

docs/api/python/contrib/onnx.md

+
+The `mxnet.contrib.onnx` package refers to the APIs and interfaces that implement ONNX model format support for Apache MXNet.
+
+With ONNX format support for MXNet, developers can build and train models with a [variety of deep learning frameworks](http://onnx.ai/supported-tools), and import these models into MXNet to run them for inference and training using MXNet’s highly optimized engine.


also for re training and transfer learning usecases right?

* ONNX Documentation * Fix rendering issues. * lint issues * fixing image links * Fix Aaron's comments * rerun CI due to unrelated failures.

anirudhacharya requested a review from szha as a code owner March 16, 2018 22:47

nswamy suggested changes Mar 16, 2018

View reviewed changes

anirudhacharya mentioned this pull request Mar 16, 2018

[MXNET-34] Onnx Module to import onnx models into mxnet #9963

Merged

8 tasks

anirudhacharya force-pushed the docs branch from ed72a3a to 4c58058 Compare March 27, 2018 23:38

Acharya added 2 commits March 27, 2018 19:46

ONNX Documentation

6e27a71

Fix rendering issues.

ea11c48

anirudhacharya force-pushed the docs branch from f43d3d0 to ea11c48 Compare March 28, 2018 23:08

Acharya added 2 commits March 28, 2018 16:26

lint issues

1bdb64c

fixing image links

9c3c03c

anirudhacharya changed the title ~~Docs page for ONNX module.~~ [MXNET-121] Docs page for ONNX module. Mar 29, 2018

aaronmarkham suggested changes Mar 29, 2018

View reviewed changes

aaronmarkham approved these changes Mar 29, 2018

View reviewed changes

Fix Aaron's comments

d70407a

anirudhacharya force-pushed the docs branch from cbeca05 to d70407a Compare March 29, 2018 22:21

sandeep-krishnamurthy approved these changes Mar 30, 2018

View reviewed changes

rerun CI due to unrelated failures.

559e17e

anirudhacharya force-pushed the docs branch from 9017d3d to 559e17e Compare March 31, 2018 05:56

sandeep-krishnamurthy merged commit 10c6af4 into apache:master Apr 2, 2018

anirudhacharya deleted the docs branch April 16, 2018 18:05

rahul003 pushed a commit to rahul003/mxnet that referenced this pull request Jun 4, 2018

[MXNET-121] Docs page for ONNX module. (apache#10140)

28ef19c

* ONNX Documentation * Fix rendering issues. * lint issues * fixing image links * Fix Aaron's comments * rerun CI due to unrelated failures.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[MXNET-121] Docs page for ONNX module. #10140

[MXNET-121] Docs page for ONNX module. #10140

anirudhacharya commented Mar 16, 2018 •

edited

Loading

nswamy Mar 16, 2018

nswamy Mar 16, 2018

nswamy Mar 16, 2018

rajanksin Mar 28, 2018 •

edited

Loading

nswamy Mar 16, 2018

nswamy Mar 16, 2018

nswamy Mar 16, 2018

ThomasDelteil commented Mar 19, 2018

anirudhacharya commented Mar 19, 2018

anirudhacharya commented Mar 26, 2018

anirudhacharya commented Mar 29, 2018

aaronmarkham Mar 29, 2018

aaronmarkham Mar 29, 2018

aaronmarkham Mar 29, 2018

anirudhacharya Mar 29, 2018

aaronmarkham Mar 29, 2018

aaronmarkham Mar 29, 2018

aaronmarkham Mar 29, 2018

aaronmarkham Mar 29, 2018

aaronmarkham Mar 29, 2018

aaronmarkham left a comment

sandeep-krishnamurthy left a comment

sandeep-krishnamurthy Mar 30, 2018

anirudhacharya Mar 30, 2018


		## Overview

		[ONNX](https://onnx.ai/) is an open format to represent deep learning models. With ONNX as an intermediate representation, it is easier to move models between state-of-the-art tools and frameworks for training and inference..


		[ONNX](https://onnx.ai/) is an open format to represent deep learning models. With ONNX as an intermediate representation, it is easier to move models between state-of-the-art tools and frameworks for training and inference..

		The `mxnet.contrib.onnx` package refers to the APIs and interfaces that implements ONNX model format support for Apache MXNet.


		The `mxnet.contrib.onnx` package refers to the APIs and interfaces that implements ONNX model format support for Apache MXNet.

		With ONNX format support for MXNet, developers can build and train models with PyTorch, CNTK, or Caffe2, and import these models into MXNet to run them for inference and training using MXNet’s highly optimized engine.

[MXNET-121] Docs page for ONNX module. #10140

[MXNET-121] Docs page for ONNX module. #10140

Conversation

anirudhacharya commented Mar 16, 2018 • edited Loading

Description

Checklist

Essentials

Changes

Comments

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rajanksin Mar 28, 2018 • edited Loading

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

ThomasDelteil commented Mar 19, 2018

anirudhacharya commented Mar 19, 2018

anirudhacharya commented Mar 26, 2018

anirudhacharya commented Mar 29, 2018

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

aaronmarkham left a comment

Choose a reason for hiding this comment

sandeep-krishnamurthy left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

anirudhacharya commented Mar 16, 2018 •

edited

Loading

rajanksin Mar 28, 2018 •

edited

Loading