Merge branch 'master' into tensorboard_profiler

Author: brianjo

_static/img/thumbnails/cropped/generic-pytorch-logo.png

@@ -99,6 +99,13 @@
 import numpy as np
 import matplotlib.pyplot as plt
 
+# NOTE: This is a hack to get around "User-agent" limitations when downloading MNIST datasets
+#       see, https://github.com/pytorch/vision/issues/3497 for more information
+from six.moves import urllib
+opener = urllib.request.build_opener()
+opener.addheaders = [('User-agent', 'Mozilla/5.0')]
+urllib.request.install_opener(opener)
+
 
 ######################################################################
 # Implementation
@@ -413,4 +420,3 @@ def test( model, device, test_loader, epsilon ):
 # it differs from FGSM. Then, try to defend the model from your own
 # attacks.
 # 
-

_static/img/thumbnails/cropped/graph-mode-dynamic-bert.png

@@ -217,13 +217,6 @@ Welcome to PyTorch Tutorials
 
 .. Frontend APIs
 
-.. customcarditem::
-   :header: (prototype) Introduction to Named Tensors in PyTorch
-   :card_description: Learn how to use PyTorch to train a Deep Q Learning (DQN) agent on the CartPole-v0 task from the OpenAI Gym.
-   :image: _static/img/thumbnails/cropped/experimental-Introduction-to-Named-Tensors-in-PyTorch.png
-   :link: intermediate/named_tensor_tutorial.html
-   :tags: Frontend-APIs,Named-Tensor,Best-Practice
-
 .. customcarditem::
    :header: (beta) Channels Last Memory Format in PyTorch
    :card_description: Get an overview of Channels Last memory format and understand how it is used to order NCHW tensors in memory preserving dimensions.
@@ -243,21 +236,21 @@ Welcome to PyTorch Tutorials
    :card_description:  Create a neural network layer with no parameters using numpy. Then use scipy to create a neural network layer that has learnable weights.
    :image: _static/img/thumbnails/cropped/Custom-Cpp-and-CUDA-Extensions.png
    :link: advanced/cpp_extension.html
-   :tags: Frontend-APIs,C++,CUDA
+   :tags: Extending-PyTorch,Frontend-APIs,C++,CUDA
 
 .. customcarditem::
    :header: Extending TorchScript with Custom C++ Operators
    :card_description:  Implement a custom TorchScript operator in C++, how to build it into a shared library, how to use it in Python to define TorchScript models and lastly how to load it into a C++ application for inference workloads.
    :image: _static/img/thumbnails/cropped/Extending-TorchScript-with-Custom-Cpp-Operators.png
    :link: advanced/torch_script_custom_ops.html
-   :tags: Frontend-APIs,TorchScript,C++
+   :tags: Extending-PyTorch,Frontend-APIs,TorchScript,C++
 
 .. customcarditem::
    :header: Extending TorchScript with Custom C++ Classes
    :card_description: This is a continuation of the custom operator tutorial, and introduces the API we’ve built for binding C++ classes into TorchScript and Python simultaneously.
    :image: _static/img/thumbnails/cropped/Extending-TorchScript-with-Custom-Cpp-Classes.png
    :link: advanced/torch_script_custom_classes.html
-   :tags: Frontend-APIs,TorchScript,C++
+   :tags: Extending-PyTorch,Frontend-APIs,TorchScript,C++
 
 .. customcarditem::
    :header: Dynamic Parallelism in TorchScript
@@ -461,6 +454,7 @@ Additional Resources
    :caption: PyTorch Recipes
 
    See All Recipes <recipes/recipes_index>
+   See All Prototype Recipes <prototype/prototype_index>
 
 .. toctree::
    :maxdepth: 2
@@ -534,14 +528,20 @@ Additional Resources
    :hidden:
    :caption: Frontend APIs
 
-   intermediate/named_tensor_tutorial
    intermediate/memory_format_tutorial
    advanced/cpp_frontend
+   advanced/torch-script-parallelism
+   advanced/cpp_autograd
+   
+.. toctree::
+   :maxdepth: 2
+   :includehidden:
+   :hidden:
+   :caption: Extending PyTorch
+
    advanced/cpp_extension
    advanced/torch_script_custom_ops
    advanced/torch_script_custom_classes
-   advanced/torch-script-parallelism
-   advanced/cpp_autograd
    advanced/dispatcher
    advanced/extend_dispatcher
 
@@ -581,6 +581,6 @@ Additional Resources
    :hidden:
    :caption: Mobile
 
-   beginner/deeplabv3_on_ios.html
-   beginner/deeplabv3_on_android.html
+   beginner/deeplabv3_on_ios
+   beginner/deeplabv3_on_android
 

beginner_source/fgsm_tutorial.py

@@ -450,19 +450,20 @@ the feature extractor is quantized).
 .. note:: Because of the random initialization your results might differ from
    the results shown in this tutorial.
 
+.. code:: python
 
-# notice `quantize=False`
-model = models.resnet18(pretrained=True, progress=True, quantize=False)
-num_ftrs = model.fc.in_features
-
-# Step 1
-model.train()
-model.fuse_model()
-# Step 2
-model_ft = create_combined_model(model)
-model_ft[0].qconfig = torch.quantization.default_qat_qconfig  # Use default QAT configuration
-# Step 3
-model_ft = torch.quantization.prepare_qat(model_ft, inplace=True)
+      # notice `quantize=False`
+      model = models.resnet18(pretrained=True, progress=True, quantize=False)
+      num_ftrs = model.fc.in_features
+
+      # Step 1
+      model.train()
+      model.fuse_model()
+      # Step 2
+      model_ft = create_combined_model(model)
+      model_ft[0].qconfig = torch.quantization.default_qat_qconfig  # Use default QAT configuration
+      # Step 3
+      model_ft = torch.quantization.prepare_qat(model_ft, inplace=True)
 
 
 Finetuning the model

index.rst

@@ -47,6 +47,11 @@
 # standard convolutional network augmented with a spatial transformer
 # network.
 
+from six.moves import urllib
+opener = urllib.request.build_opener()
+opener.addheaders = [('User-agent', 'Mozilla/5.0')]
+urllib.request.install_opener(opener)
+
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
 # Training dataset

intermediate_source/quantized_transfer_learning_tutorial.rst

@@ -0,0 +1,147 @@
+PyTorch Prototype Recipes
+---------------------------------------------
+Prototype features are not available as part of binary distributions like PyPI or Conda (except maybe behind run-time flags). To test these features we would, depending on the feature, recommend building from master or using the nightly wheels that are made available on `pytorch.org <https://pytorch.org>`_.
+
+*Level of commitment*: We are committing to gathering high bandwidth feedback only on these features. Based on this feedback and potential further engagement between community members, we as a community will decide if we want to upgrade the level of commitment or to fail fast.
+
+
+.. raw:: html
+
+        </div>
+    </div>
+
+    <div id="tutorial-cards-container">
+
+    <nav class="navbar navbar-expand-lg navbar-light tutorials-nav col-12">
+        <div class="tutorial-tags-container">
+            <div id="dropdown-filter-tags">
+                <div class="tutorial-filter-menu">
+                    <div class="tutorial-filter filter-btn all-tag-selected" data-tag="all">All</div>
+                </div>
+            </div>
+        </div>
+    </nav>
+
+    <hr class="tutorials-hr">
+
+    <div class="row">
+
+    <div id="tutorial-cards">
+    <div class="list">
+
+.. Add prototype tutorial cards below this line
+
+.. Quantization
+
+.. customcarditem::
+   :header: FX Graph Mode Quantization User Guide
+   :card_description: Learn about FX Graph Mode Quantization.
+   :image: ../_static/img/thumbnails/cropped/generic-pytorch-logo.PNG
+   :link: ../prototype/fx_graph_mode_quant_guide.html
+   :tags: FX,Quantization
+
+.. customcarditem::
+   :header: FX Graph Mode Post Training Dynamic Quantization
+   :card_description: Learn how to do post training dynamic quantization in graph mode based on torch.fx. 
+   :image: ../_static/img/thumbnails/cropped/generic-pytorch-logo.PNG
+   :link: ../prototype/fx_graph_mode_ptq_dynamic.html
+   :tags: FX,Quantization
+
+.. customcarditem::
+   :header: FX Graph Mode Post Training Static Quantization
+   :card_description: Learn how to do post training static quantization in graph mode based on torch.fx. 
+   :image: ../_static/img/thumbnails/cropped/generic-pytorch-logo.PNG
+   :link: ../prototype/fx_graph_mode_ptq_static.html
+   :tags: FX,Quantization
+
+.. customcarditem::
+   :header: Graph Mode Dynamic Quantization on BERT
+   :card_description: Learn how to do post training dynamic quantization with graph mode quantization on BERT models.
+   :image: ../_static/img/thumbnails/cropped/graph-mode-dynamic-bert.PNG
+   :link: ../prototype/graph_mode_dynamic_bert_tutorial.html
+   :tags: Text,Quantization
+
+.. customcarditem::
+   :header: PyTorch Numeric Suite Tutorial
+   :card_description: Learn how to use the PyTorch Numeric Suite to support quantization debugging efforts.
+   :image: ../_static/img/thumbnails/cropped/generic-pytorch-logo.PNG
+   :link: ../prototype/numeric_suite_tutorial.html
+   :tags: Debugging,Quantization
+
+.. Mobile
+
+.. customcarditem::
+   :header: Use iOS GPU in PyTorch
+   :card_description: Learn how to run your models on iOS GPU.
+   :image: ../_static/img/thumbnails/cropped/ios.PNG
+   :link: ../prototype/ios_gpu_workflow.html
+   :tags: Mobile
+
+.. customcarditem::
+   :header: Convert MobileNetV2 to NNAPI
+   :card_description: Learn how to prepare a computer vision model to use Android’s Neural Networks API (NNAPI).
+   :image: ../_static/img/thumbnails/cropped/android.PNG
+   :link: ../prototype/nnapi_mobilenetv2.html
+   :tags: Mobile
+
+.. customcarditem::
+   :header: PyTorch Vulkan Backend User Workflow
+   :card_description: Learn how to use the Vulkan backend on mobile GPUs.
+   :image: ../_static/img/thumbnails/cropped/android.PNG
+   :link: ../prototype/vulkan_workflow.html
+   :tags: Mobile
+
+.. customcarditem::
+   :header: Lite Interpreter Workflow in Android and iOS
+   :card_description: Learn how to use the lite interpreter on iOS and Andriod devices.
+   :image: ../_static/img/thumbnails/cropped/mobile.PNG
+   :link: ../prototype/lite_interpreter.html
+   :tags: Mobile
+
+.. TorchScript
+
+.. customcarditem::
+   :header: Model Freezing in TorchScript
+   :card_description: Freezing is the process of inlining Pytorch module parameters and attributes values into the TorchScript internal representation.
+   :image: ../_static/img/thumbnails/cropped/generic-pytorch-logo.PNG
+   :link: ../prototype/torchscript_freezing.html
+   :tags: TorchScript
+
+.. vmap
+
+.. customcarditem::
+   :header: Using torch.vmap
+   :card_description: Learn about torch.vmap, an autovectorizer for PyTorch operations.
+   :image: ../_static/img/thumbnails/cropped/generic-pytorch-logo.PNG
+   :link: ../prototype/vmap_recipe.html
+   :tags: vmap
+
+.. End of tutorial card section
+
+.. raw:: html
+
+    </div>
+
+    <div class="pagination d-flex justify-content-center"></div>
+
+    </div>
+
+    </div>
+
+.. -----------------------------------------
+.. Page TOC
+.. -----------------------------------------
+.. toctree::
+   :hidden:
+
+   prototype/fx_graph_mode_quant_guide.html
+   prototype/fx_graph_mode_ptq_dynamic.html
+   prototype/fx_graph_mode_ptq_static.html
+   prototype/graph_mode_dynamic_bert_tutorial.html
+   prototype/ios_gpu_workflow.html
+   prototype/nnapi_mobilenetv2.html
+   prototype/numeric_suite_tutorial.html
+   prototype/torchscript_freezing.html
+   prototype/vmap_recipe.html
+   prototype/vulkan_workflow.html
+   prototype/lite_interpreter.html