Containers examples & other updates

README.md

@@ -36,3 +36,17 @@ To build with testing the notebooks (requires GPU):
 ```bash
 make
 ```
+
+## Troubleshooting
+
+You might need to init and update your submodules.
+
+```bash
+git submodule update --init --recursive
+```
+
+You might need to update the environment for the latest modules.
+
+```bash
+conda env update -f environment.yml
+```

api/gluon/containers.rst

@@ -8,7 +8,7 @@ This section covers the classes to construct neural network models.
 Blocks
 ------
 
-The :py:mod:`mxnet.gluon.nn` module provides three classes to construct basc
+The :py:mod:`mxnet.gluon.nn` module provides three classes to construct basic
 blocks for a neural network model:
 
 .. autosummary::
@@ -21,23 +21,25 @@ blocks for a neural network model:
 
 The difference between these three classes:
 
-- :py:class:`Block`: the bass class for any neural nework layers and models.
+- :py:class:`Block`: the base class for any neural network layers and models.
+  `A look under the hood of gluon`_ and
+  `Gluon - Neural network building blocks`_ are tutorials that provide details
+  on how to use this class of container.
 - :py:class:`HybridBlock`: a subclass of :py:class:`Block` that allows to
   hybridize a model. It constraints operations can be run in the ``forward``
-  method, e.g. the `print` function doesn't work any more. Check tutorial XXX
-  for more details.
-
-- :py:class:`SymbolBlock`: a sublcass of :py:class:`Block` that is able to wrap
+  method, e.g. the `print` function doesn't work any more.
+- :py:class:`SymbolBlock`: a subclass of :py:class:`Block` that is able to wrap
   a :py:class:`mxnet.symbol.Symbol` instance into a :py:class:`Block`
-  instance. Check XXX-Symbol tutorials and how to XXX to use this class.
+  instance.
 
 
 Sequential containers
 ---------------------
 
 Besides inheriting :py:class:`mxnet.gluon.nn.Block` to create a neural network
 models, :py:mod:`mxnet.gluon.nn` provides two classes to construct a model by
-stacking layers sequentially. Refer to XXX for tutorials how to use them.
+stacking layers sequentially. Refer to the tutorials list in the
+`Further Reading`_ section on how to use them.
 
 .. currentmodule:: mxnet.gluon.nn
 
@@ -48,6 +50,7 @@ stacking layers sequentially. Refer to XXX for tutorials how to use them.
     Sequential
     HybridSequential
 
+
 Concurrent containers
 ---------------------
 
@@ -65,4 +68,16 @@ ResNet and Inception block in GoogLeNet.
     Concurrent
     HybridConcurrent
 
+
+Further Reading
+---------------
+
+- `A look under the hood of gluon`_
+- `Gluon - Neural network building blocks`_
+
+
 .. disqus::
+
+
+.. _A look under the hood of gluon: https://gluon.mxnet.io/chapter03_deep-neural-networks/plumbing.html
+.. _Gluon - Neural network building blocks: https://mxnet.incubator.apache.org/tutorials/gluon/gluon.html

build/conf.py

@@ -70,9 +70,9 @@
     'sphinx.ext.napoleon',
     # 'sphinx.ext.mathjax',
     # 'sphinx.ext.viewcode',
-    'breathe',
+    'breathe'
 #    'mxdoc'
-    'sphinxcontrib.disqus'
+#    'sphinxcontrib.disqus'
 ]
 
 disqus_shortname = 'mxnet'

develop/tutorials/index.rst

@@ -8,16 +8,16 @@ Tutorials
         :align: left
 
         If you would like to learn deep learning algorithms, please refer to
-        "dive into deep learning" book. It explain every algorithm from the
-        beggining, and implements them from scratch with MXNet.
+        the "Dive into Deep Learning" book. It explains every algorithm from the
+        beginning, and implements them with MXNet from scratch.
 
-In this unit, we will go thourgh how to use MXNet to implement various deep
+In this unit, we will go through how to use MXNet to implement various deep
 learning algorithms.
 
 Getting Started
 ---------------
 
-We recommend you to start with :doc:`crash-course/index`. Gluon is the
+We recommend you start with :doc:`crash-course/index`. Gluon is the
 imperative frontend of MXNet. Once you've completed the course, you should be
 able to understand how to use the major components of Gluon, and define and
 train a basic neural network to classify images.

environment.yml

@@ -11,9 +11,10 @@ dependencies:
   - nbsphinx
   - recommonmark
   - notedown
-  - mxnet-cu92
+  - mxnet
   - pypandoc
   - sphinx_materialdesign_theme
   - breathe
   - mock
   - awscli
+  - sphinxcontrib-disqus