Added mechanism to invoke nbsphinx to render html versions of example…

… notebooks. Added links to the same from the user guide.

Signed-off-by: Abhi Khobare <[email protected]>

Docs/CMakeLists.txt

@@ -48,7 +48,10 @@ add_custom_target(doc
       "${AIMET_LD_LIBRARY_PATH}"
       "${AIMET_PYTHONPATH}"
       "SW_VERSION=${SW_VERSION}"
-      sphinx-build -v -T -b html ${CMAKE_CURRENT_SOURCE_DIR} ${CMAKE_CURRENT_BINARY_DIR})
+      sphinx-build -v -T -b html ${CMAKE_BINARY_DIR}/Docs_SOURCE ${CMAKE_CURRENT_BINARY_DIR})
+
+add_dependencies(doc
+    copy_doc_source copy_examples)
 
 add_custom_target(copy_doc_source
     COMMAND ${CMAKE_COMMAND} -E copy_directory ${CMAKE_CURRENT_SOURCE_DIR} ${CMAKE_BINARY_DIR}/Docs_SOURCE)

Docs/conf.py

@@ -86,7 +86,8 @@ def setup(app):
     'sphinx.ext.autodoc',
     'sphinx_autodoc_typehints',
     'sphinx.ext.autosectionlabel',
-    'sphinx.ext.ifconfig'
+    'sphinx.ext.ifconfig',
+    'nbsphinx'
     # 'rinoh.frontend.sphinx'
 ]
 
@@ -216,3 +217,6 @@ def setup(app):
 
 # -- Extension configuration -------------------------------------------------
 autoclass_content = 'both'
+
+nbsphinx_allow_errors = True
+nbsphinx_execute = 'never'

Docs/toplevelhidden.rst

@@ -0,0 +1,4 @@
+
+.. toctree::
+    :hidden:
+

Docs/user_guide/examples.rst

@@ -6,8 +6,8 @@
 AIMET Examples
 ==============
 
-AIMET Examples provide reference code (in the form of *scripts* and *Jupyter Notebooks*) to learn how to load models,
-apply AIMET quantization and compression features, fine tune and save your models. It is also a quick way to become
+AIMET Examples provide reference code (in the form of *Jupyter Notebooks*) to learn how to
+apply AIMET quantization and compression features. It is also a quick way to become
 familiar with AIMET usage and APIs.
 
 For more details on each of the features and APIs please refer:
@@ -17,44 +17,60 @@ Installation Instructions
 =========================
 
 - The dataloader, evaluator, and trainer utilized in the examples is for the ImageNet dataset.
-  To run the example, please download it from here: https://www.image-net.org/download.php
+  To run the example, please download the ImageNet dataset from here: https://www.image-net.org/download.php
 - Install AIMET and its dependencies using the instructions in the Installation section'
-- Clone the AIMET repo as follows to any location:
+    - Installing AIMET will also install these AIMET Examples
+
+Examples for PyTorch Models
+===========================
+
+The following table has links to browsable versions of the notebooks for different features.
+
+
+.. list-table:: Quantization Examples
+   :widths: 40 12 12
+   :header-rows: 1
+
+   * - Features
+     - PyTorch
+     - TensorFlow
+   * - Quantization-Aware Training (QAT)
+     - `Link <../Examples/torch/quantization/qat.ipynb>`_
+     - `Link <../Examples/tensorflow/quantization/qat.ipynb>`_
+   * - QAT with Range Learning
+     - `Link <../Examples/torch/quantization/qat_range_learning.ipynb>`_
+     - `Link <../Examples/tensorflow/quantization/qat_range_learning.ipynb>`_
+   * - Cross-Layer Equalization (CLE)
+     - `Link <../Examples/torch/quantization/cle_bc.ipynb>`_
+     - `Link <../Examples/tensorflow/quantization/cle_bc.ipynb>`_
+   * - Adaptive Rounding (AdaRound)
+     - `Link <../Examples/torch/quantization/adaround.ipynb>`_
+     - `Link <../Examples/tensorflow/quantization/adaround.ipynb>`_
+   * - AutoQuant
+     - `Link <../Examples/torch/quantization/autoquant.ipynb>`_
+     - `Link <../Examples/tensorflow/quantization/autoquant.ipynb>`_
+   * - BN Re-estimation
+     - `Link <../Examples/torch/quantization/bn_reestimation.ipynb>`_
+     - `Link <../Examples/tensorflow/quantization/bn_reestimation.ipynb>`_
+
+
+.. list-table:: Compression Examples
+   :widths: 40 12 12
+   :header-rows: 1
+
+   * - Features
+     - PyTorch
+     - TensorFlow
+   * - Channel Pruning
+     - `Link <../Examples/torch/compression/channel_pruning.ipynb>`_
+     - `Link <../Examples/tensorflow/compression/channel_pruning.ipynb>`_
+   * - Spatial SVD
+     - `Link <../Examples/torch/compression/spatial_svd.ipynb>`_
+     - `Link <../Examples/tensorflow/compression/spatial_svd.ipynb>`_
+   * - Spatial SVD + Channel Pruning
+     - `Link <../Examples/torch/compression/spatial_svd_channel_pruning.ipynb>`_
+     - `Link <../Examples/tensorflow/compression/spatial_svd_channel_pruning.ipynb>`_
 
-  * WORKSPACE="<absolute_path_to_workspace>"
-
-  * mkdir $WORKSPACE && cd $WORKSPACE
-
-  *  Go to https://github.com/quic/aimet/releases and identify the release tag (`<release_tag>`) of the AIMET package that you're working with.
-
-  * git clone https://github.com/quic/aimet.git --branch <release_tag>
-
-  * Update the environment variable as follows: `export PYTHONPATH=$PYTHONPATH:${WORKSPACE}/aimet`
-
-Code Layout
-===========
-
-The code for AIMET Examples shares a common structure:
-
-Examples/
-
-  common/
-
-  torch/
-
-    utils/
-
-    quantization/
-
-    compression/
-
-  tensorflow/
-
-    utils/
-
-    quantization/
-
-    compression/
 
 |
 Running Examples via Jupyter Notebook
@@ -75,15 +91,3 @@ Running Examples via Jupyter Notebook
   - `Examples/tensorflow/compression/`
 - Follow the instructions therein to execute the code.
 
-Running Examples via Command Line
-=================================
-
-To run an AIMET example:
-
-python example_name.py --dataset_dir path/to/dataset/ --use_cuda
-
-For example, to run the channel pruning example run the following:
-
-python channel_pruning.py --dataset_dir path/to/dataset/ --use_cuda --epochs 15 --learning_rate 1e-2 --learning_rate_schedule [5, 10]
-
-Setting the hyperparameters epochs, learning rate, and learning rate scheduler is optional. If the values are not given, the default values will be used.

Examples/CMakeLists.txt

@@ -36,37 +36,11 @@
 
 set(AIMET_EXAMPLES_DEST ${AIMET_INSTALL_DIR}/Examples)
 
-install(DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/common
-      DESTINATION ${AIMET_EXAMPLES_DEST}
-      PATTERN "__pycache*" EXCLUDE
-      PATTERN "CMakeLists.txt" EXCLUDE
-      )
-
-# Copy over the Examples documentation
-install(FILES ${CMAKE_CURRENT_SOURCE_DIR}/README.md
-      DESTINATION ${AIMET_EXAMPLES_DEST}
-      )
-
-if (ENABLE_TORCH)
-      install(DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/torch
-      DESTINATION ${AIMET_EXAMPLES_DEST}
-      PATTERN "__pycache*" EXCLUDE
-      PATTERN "CMakeLists.txt" EXCLUDE
-      )
-
-endif(ENABLE_TORCH)
-
-if (ENABLE_TENSORFLOW)
-      install(DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/tensorflow
-      DESTINATION ${AIMET_EXAMPLES_DEST}
-      PATTERN "__pycache*" EXCLUDE
-      PATTERN "CMakeLists.txt" EXCLUDE
-      )
-
-endif(ENABLE_TENSORFLOW)
+add_custom_target(copy_examples
+    COMMAND ${CMAKE_COMMAND} -E copy_directory ${CMAKE_CURRENT_SOURCE_DIR} ${CMAKE_BINARY_DIR}/Docs_SOURCE/Examples)
 
 # Create a tarball archive of the Examples code
 install(CODE "execute_process(COMMAND
                 ${CMAKE_COMMAND} -E chdir ${AIMET_INSTALL_DIR}
-                ${CMAKE_COMMAND} -E tar cvzf ${AIMET_EXAMPLES_DEST}.tar.gz Examples)"
+                ${CMAKE_COMMAND} -E tar cvzf ${AIMET_EXAMPLES_DEST}.tar.gz ${CMAKE_CURRENT_SOURCE_DIR}/Examples)"
         )