[Don't Merge] Tutorials Refresh. Creating for preview.

.jenkins/build.sh

@@ -76,6 +76,17 @@ if [[ "${JOB_BASE_NAME}" == *worker_* ]]; then
       FILES_TO_RUN+=($(basename $filename .py))
     fi
     count=$((count+1))
+  done
+    done
+  for filename in $(find recipes_source/ -name '*.py' -not -path '*/data/*'); do
+    if [ $(($count % $NUM_WORKERS)) != $WORKER_ID ]; then
+      echo "Removing runnable code from "$filename
+      python $DIR/remove_runnable_code.py $filename $filename
+    else
+      echo "Keeping "$filename
+      FILES_TO_RUN+=($(basename $filename .py))
+    fi
+    count=$((count+1))
   done
   echo "FILES_TO_RUN: " ${FILES_TO_RUN[@]}
 

beginner_source/blitz/tensor_tutorial.py

@@ -3,101 +3,147 @@
 What is PyTorch?
 ================
 
-It’s a Python-based scientific computing package targeted at two sets of
+It is a open source machine learning framework that accelerates the 
+path from research prototyping to production deployment.
+
+PyTorch is built as a Python-based scientific computing package targeted at two sets of
 audiences:
 
--  A replacement for NumPy to use the power of GPUs
--  a deep learning research platform that provides maximum flexibility
-   and speed
+-  Those who are looking for a replacement for NumPy to use the power of GPUs.
+-  Researchers who want to build with a deep learning platform that provides maximum flexibility
+   and speed.
 
 Getting Started
 ---------------
 
+In this section of the tutorial, we will introduce the concept of a tensor in PyTorch, and its operations.
+
 Tensors
 ^^^^^^^
 
-Tensors are similar to NumPy’s ndarrays, with the addition being that
-Tensors can also be used on a GPU to accelerate computing.
+A tensor is a generic n-dimensional array. Tensors in PyTorch are similar to NumPy’s ndarrays, 
+with the addition being that tensors can also be used on a GPU to accelerate computing.
+
+To see the behavior of tensors, we will first need to import PyTorch into our program.
 """
 
 from __future__ import print_function
 import torch
 
-###############################################################
-# .. note::
-#     An uninitialized matrix is declared,
-#     but does not contain definite known
-#     values before it is used. When an
-#     uninitialized matrix is created,
-#     whatever values were in the allocated
-#     memory at the time will appear as the initial values.
+"""
+We import ``future`` here to help port our code from Python 2 to Python 3.
+For more details, see the `Python-Future technical documentation <https://python-future.org/quickstart.html>`_. 
+
+Let's take a look at how we can create tensors.
+"""
 
 ###############################################################
-# Construct a 5x3 matrix, uninitialized:
+# First, construct a 5x3 empty matrix:
 
 x = torch.empty(5, 3)
 print(x)
+
+"""
+``torch.empty`` creates an uninitialized matrix of type tensor.
+When an empty tensor is declared, it does not contain definite known values
+before you populate it. The values in the empty tensor are those that were in 
+the allocated memory at the time of initialization.
+"""
 
 ###############################################################
-# Construct a randomly initialized matrix:
+# Now, construct a randomly initialized matrix:
 
 x = torch.rand(5, 3)
 print(x)
 
+"""
+``torch.rand`` creates an initialized matrix of type tensor with a random 
+sampling of values. 
+"""
+
 ###############################################################
 # Construct a matrix filled zeros and of dtype long:
 
 x = torch.zeros(5, 3, dtype=torch.long)
 print(x)
 
+"""
+``torch.zeros`` creates an initialized matrix of type tensor with every
+index having a value of zero.
+"""
+
 ###############################################################
-# Construct a tensor directly from data:
+# Let's construct a tensor with data that we define ourselves:
 
 x = torch.tensor([5.5, 3])
 print(x)
 
+"""
+Our tensor can represent all types of data. This data can be an audio waveform, the
+pixels of an image, even entities of a language.
+
+PyTorch has packages that support these specific data types. For additional learning, see:
+-  `torchvision <https://pytorch.org/docs/stable/torchvision/index.html>`_
+-  `torchtext <https://pytorch.org/text/>`_
+-  `torchaudio <https://pytorch.org/audio/>`_
+"""
+
 ###############################################################
-# or create a tensor based on an existing tensor. These methods
-# will reuse properties of the input tensor, e.g. dtype, unless
-# new values are provided by user
+# You can create a tensor based on an existing tensor. These methods 
+# reuse the properties of the input tensor, e.g. ``dtype``, unless
+# new values are provided by the user.
+#
 
-x = x.new_ones(5, 3, dtype=torch.double)      # new_* methods take in sizes
+x = x.new_ones(5, 3, dtype=torch.double)
 print(x)
 
 x = torch.randn_like(x, dtype=torch.float)    # override dtype!
 print(x)                                      # result has the same size
 
+"""
+``tensor.new_*`` methods take in the size of the tensor and a ``dtype``, 
+returning a tensor filled with ones.
+
+In this example,``torch.randn_like`` creates a new tensor based upon the 
+input tensor, and overrides the ``dtype`` to be a float. The output of 
+this method is a tensor of the same size and different ``dtype``.
+"""
+
 ###############################################################
-# Get its size:
+# We can get the size of a tensor as a tuple:
 
 print(x.size())
 
 ###############################################################
 # .. note::
-#     ``torch.Size`` is in fact a tuple, so it supports all tuple operations.
+#     Since ``torch.Size`` is a tuple, it supports all tuple operations.
 #
 # Operations
 # ^^^^^^^^^^
-# There are multiple syntaxes for operations. In the following
-# example, we will take a look at the addition operation.
+# There are multiple syntaxes for operations that can be performed on tensors.
+# In the following example, we will take a look at the addition operation.
 #
-# Addition: syntax 1
+# First, let's try using the ``+`` operator.
+
 y = torch.rand(5, 3)
 print(x + y)
 
 ###############################################################
-# Addition: syntax 2
+# Using the ``+`` operator should have the same output as using the 
+# ``add()`` method. 
 
 print(torch.add(x, y))
 
 ###############################################################
-# Addition: providing an output tensor as argument
+# You can also provide a tensor as an argument to the ``add()``
+# method that will contain the data of the output operation.
+
 result = torch.empty(5, 3)
 torch.add(x, y, out=result)
 print(result)
 
 ###############################################################
-# Addition: in-place
+# Finally, you can perform this operation in-place.
 
 # adds x to y
 y.add_(x)
@@ -107,21 +153,29 @@
 # .. note::
 #     Any operation that mutates a tensor in-place is post-fixed with an ``_``.
 #     For example: ``x.copy_(y)``, ``x.t_()``, will change ``x``.
-#
-# You can use standard NumPy-like indexing with all bells and whistles!
+
+###############################################################
+# Similar to NumPy, tensors can be indexed using the standard
+# Python ``x[i]`` syntax, where ``x`` is the array and ``i`` is the selection.
+# 
+# That said, you can use NumPy-like indexing with all its bells and whistles!
 
 print(x[:, 1])
 
 ###############################################################
-# Resizing: If you want to resize/reshape tensor, you can use ``torch.view``:
+# Resizing your tensors might be necessary for your data.
+# If you want to resize or reshape tensor, you can use ``torch.view``:
+
 x = torch.randn(4, 4)
 y = x.view(16)
 z = x.view(-1, 8)  # the size -1 is inferred from other dimensions
 print(x.size(), y.size(), z.size())
 
 ###############################################################
-# If you have a one element tensor, use ``.item()`` to get the value as a
-# Python number
+# You can access the Python number-value of a one-element tensor using ``.item()``.
+# If you have a multidimensional tensor, see the 
+# `tolist() <https://pytorch.org/docs/stable/tensors.html#torch.Tensor.tolist>`_ method.
+
 x = torch.randn(1)
 print(x)
 print(x.item())
@@ -130,43 +184,55 @@
 # **Read later:**
 #
 #
-#   100+ Tensor operations, including transposing, indexing, slicing,
-#   mathematical operations, linear algebra, random numbers, etc.,
-#   are described
-#   `here <https://pytorch.org/docs/torch>`_.
+#   This was just a sample of the 100+ Tensor operations you have 
+#   access to in PyTorch. There are many others, including transposing,
+#   indexing, slicing, mathematical operations, linear algebra,
+#   random numbers, and more. Read and explore more about them in our 
+#   `technical documentation <https://pytorch.org/docs/torch>`_.
 #
 # NumPy Bridge
 # ------------
 #
-# Converting a Torch Tensor to a NumPy array and vice versa is a breeze.
+# As mentioned earlier, one of the benefits of using PyTorch is that it 
+# is built to provide a seemless transition from NumPy.
+# 
+# For example, converting a Torch Tensor to a NumPy array (and vice versa)
+# is a breeze.
 #
 # The Torch Tensor and NumPy array will share their underlying memory
-# locations (if the Torch Tensor is on CPU), and changing one will change
+# locations (if the Torch Tensor is on CPU). That means, changing one will change
 # the other.
 #
+# Let's see this in action.
+#
 # Converting a Torch Tensor to a NumPy Array
 # ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+# First, construct a 1-dimensional tensor populated with ones.
 
 a = torch.ones(5)
 print(a)
 
 ###############################################################
-#
+# Now, let's construct a NumPy array based off of that tensor.
 
 b = a.numpy()
 print(b)
 
 ###############################################################
-# See how the numpy array changed in value.
+# Let's see how they share their memory locations. Add ``1`` to the torch tensor.
 
 a.add_(1)
 print(a)
 print(b)
 
+###############################################################
+# Take note how the numpy array also changed in value.
+
 ###############################################################
 # Converting NumPy Array to Torch Tensor
 # ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-# See how changing the np array changed the Torch Tensor automatically
+# Try the same thing for NumPy to Torch Tensor.
+# See how changing the NumPy array changed the Torch Tensor automatically as well.
 
 import numpy as np
 a = np.ones(5)
@@ -176,15 +242,17 @@
 print(b)
 
 ###############################################################
-# All the Tensors on the CPU except a CharTensor support converting to
+# All the Tensors on the CPU (except a CharTensor) support converting to
 # NumPy and back.
 #
 # CUDA Tensors
 # ------------
 #
 # Tensors can be moved onto any device using the ``.to`` method.
+# The following code block can be run by changing the runtime in
+# your notebook to "GPU" or greater.
 
-# let us run this cell only if CUDA is available
+# This cell will run only if CUDA is available
 # We will use ``torch.device`` objects to move tensors in and out of GPU
 if torch.cuda.is_available():
     device = torch.device("cuda")          # a CUDA device object
@@ -193,3 +261,7 @@
     z = x + y
     print(z)
     print(z.to("cpu", torch.double))       # ``.to`` can also change dtype together!
+
+###############################################################
+# Now that you have had time to experiment with Tensors in PyTorch, let's take
+# a look at Automatic Differentiation.

conf.py

@@ -34,15 +34,15 @@
 import torch
 import glob
 import shutil
-from custom_directives import IncludeDirective, GalleryItemDirective, CustomGalleryItemDirective
+from custom_directives import IncludeDirective, GalleryItemDirective, CustomGalleryItemDirective, CustomCalloutItemDirective, CustomCardItemDirective
 
 
 try:
     import torchvision
 except ImportError:
     import warnings
     warnings.warn('unable to load "torchvision" package')
-import pytorch_sphinx_theme
+#import pytorch_sphinx_theme
 
 # -- General configuration ------------------------------------------------
 
@@ -61,8 +61,8 @@
 
 sphinx_gallery_conf = {
     'examples_dirs': ['beginner_source', 'intermediate_source',
-                      'advanced_source'],
-    'gallery_dirs': ['beginner', 'intermediate', 'advanced'],
+                      'advanced_source', 'recipes_source'],
+    'gallery_dirs': ['beginner', 'intermediate', 'advanced', 'recipes'],
     'filename_pattern': os.environ.get('GALLERY_PATTERN', r'tutorial.py'),
     'backreferences_dir': False
 }
@@ -162,7 +162,7 @@
 
 
 html_theme = 'pytorch_sphinx_theme'
-html_theme_path = [pytorch_sphinx_theme.get_html_theme_path()]
+html_theme_path = ['../../pytorch_sphinx_theme-1']
 html_logo = '_static/img/pytorch-logo-dark.svg'
 html_theme_options = {
     'pytorch_project': 'tutorials',
@@ -237,3 +237,5 @@ def setup(app):
     app.add_directive('includenodoc', IncludeDirective)
     app.add_directive('galleryitem', GalleryItemDirective)
     app.add_directive('customgalleryitem', CustomGalleryItemDirective)
+    app.add_directive('customcarditem', CustomCardItemDirective)
+    app.add_directive('customcalloutitem', CustomCalloutItemDirective)