yapf

sphinx-build -b html docs/source docs/_build

# Copy 

cp -r docs/_build/ website/docs/

# Can't seem to get this copied over, so doing manually for now.

echo "About to push docs to s3"

python devtools/travis-ci/push-docs-to-s3.py

import tempfile

import subprocess

BUCKET_NAME = 'deepchem.io'

if not any(d.project_name == 's3cmd' for d in pip.get_installed_distributions()):

if not any(d.project_name == 's3cmd'

           for d in pip.get_installed_distributions()):

  raise ImportError('The s3cmd package is required. try $ pip install s3cmd')

# The secret key is available as a secure environment variable

# on travis-ci to push the build documentation to Amazon S3.

#default_mime_type = binary/octet-stream

with tempfile.NamedTemporaryFile('w') as f:

  f.write('''[default]

access_key = {AWS_ACCESS_KEY_ID}

'''.format(**os.environ))

  f.flush()

  template = ('s3cmd -M -H --config {config} '

              'sync website/ s3://{bucket}/')

  cmd = template.format(

          config=f.name,

          bucket=BUCKET_NAME)

  template = ('s3cmd -M -H --config {config} ' 'sync website/ s3://{bucket}/')

  cmd = template.format(config=f.name, bucket=BUCKET_NAME)

  subprocess.call(cmd.split())

  # Perform recursive modification to set css mime types.

  template = ("s3cmd --recursive modify --add-header='content-type':'text/css'"

              "--exclude '' --include '.css' --config {config} s3://{bucket}/")

  cmd = template.format(

          config=f.name,

          bucket=BUCKET_NAME)

  cmd = template.format(config=f.name, bucket=BUCKET_NAME)

  subprocess.call(cmd.split())

  # Perform recursive modification to set js mime types.

  template = ("s3cmd --recursive modify --add-header='content-type':'application/javascript'"

  template = (

      "s3cmd --recursive modify --add-header='content-type':'application/javascript'"

      "--exclude '' --include '.js' --config {config} s3://{bucket}/")

  cmd = template.format(

          config=f.name,

          bucket=BUCKET_NAME)

  cmd = template.format(config=f.name, bucket=BUCKET_NAME)

  subprocess.call(cmd.split())

# extensions coming with Sphinx (named 'sphinx.ext.*') or your custom

# ones.

extensions = [

    'sphinx.ext.autodoc',

    'sphinx.ext.autosummary',

    'sphinx.ext.doctest',

    'sphinx.ext.intersphinx',

    'sphinx.ext.coverage',

    'sphinx.ext.mathjax',

    'sphinx.ext.ifconfig',

    'numpydoc',

    'sphinx.ext.viewcode',

    'sphinx.ext.autodoc', 'sphinx.ext.autosummary', 'sphinx.ext.doctest',

    'sphinx.ext.intersphinx', 'sphinx.ext.coverage', 'sphinx.ext.mathjax',

    'sphinx.ext.ifconfig', 'numpydoc', 'sphinx.ext.viewcode',

    'notebook_sphinxext'

]

# If true, `todo` and `todoList` produce output, else they produce nothing.

todo_include_todos = False

# -- Options for HTML output ----------------------------------------------

# The theme to use for HTML and HTML Help pages.  See the documentation for

}

# Add any paths that contain custom themes here, relative to this directory.

#print("sphinx_bootstrap_theme.get_html_theme_path()")

html_theme_path = sphinx_bootstrap_theme.get_html_theme_path()

#html_theme_path = [

#  #msmb_theme.get_html_theme_path(),

#  sphinx_bootstrap_theme.get_html_theme_path()

#]

# The name for this set of Sphinx documents.  If None, it defaults to

# "<project> v<release> documentation".

# The name of an image file (relative to this directory) to place at the top

# of the sidebar.

html_logo = '_static/logo.png'

#html_logo = '../website/static/img/logo.png' 

# The name of an image file (within the static path) to use as favicon of the

# docs.  This file should be a Windows icon file (.ico) being 16x16 or 32x32

# relative to this directory. They are copied after the builtin static files,

# so a file named "default.css" will overwrite the builtin "default.css".

html_static_path = ['_static']

#html_static_path = ['../website/static/']

# Add any extra paths that contain custom files (such as robots.txt or

# .htaccess) here, relative to this directory. These files are copied

# If false, no module index is generated.

#latex_domain_indices = True

# -- Options for manual page output ---------------------------------------

# One entry per manual page. List of tuples

# (source start file, name, description, authors, manual section).

man_pages = [

    (master_doc, 'deepchem', u'deepchem Documentation',

     ["Stanford University"], 1)

]

man_pages = [(master_doc, 'deepchem', u'deepchem Documentation',

              ["Stanford University"], 1)]

# If true, show URL addresses after external links.

#man_show_urls = False

# -- Options for Texinfo output -------------------------------------------

# Grouping the document tree into Texinfo files. List of tuples

# (source start file, target name, title, author,

#  dir menu entry, description, category)

texinfo_documents = [

  (master_doc, 'deepchem', u'deepchem Documentation',

   'Stanford University', 'deepchem', 'Deep-learning models for drug discovery.',

   'Scientific'),

    (master_doc, 'deepchem', u'deepchem Documentation', 'Stanford University',

     'deepchem', 'Deep-learning models for drug discovery.', 'Scientific'),

]

# Documents to append as an appendix to all manuals.

# If true, do not generate a @detailmenu in the "Top" node's menu.

#texinfo_no_detailmenu = False

# Example configuration for intersphinx: refer to the Python standard library.

intersphinx_mapping = {'https://docs.python.org/': None}

def export_html(nb, f):

  config = {

        'Exporter': {'template_file': 'basic',

                     'template_path': ['./sphinxext/']},

        'ExtractOutputPreprocessor': {'enabled': True},

        'CSSHTMLHeaderPreprocessor': {'enabled': True}

      'Exporter': {

          'template_file': 'basic',

          'template_path': ['./sphinxext/']

      },

      'ExtractOutputPreprocessor': {

          'enabled': True

      },

      'CSSHTMLHeaderPreprocessor': {

          'enabled': True

      }

  }

  exporter = HTMLExporter(config)

    link_rst = "({uneval}; {eval}; {py})".format(

        uneval=formatted_link(f['uneval']),

        eval=formatted_link(f['eval']),

            py=formatted_link(f['py']),

        )

        py=formatted_link(f['py']),)

    rst_file = self.state_machine.document.attributes['source']

    self.state_machine.insert_input([link_rst], rst_file)

    # create notebook node

    nb_node = notebook_node('', html, format='html', source='nb_path')

        nb_node.source, nb_node.line = (self.state_machine

                                        .get_source_and_line(self.lineno))

    nb_node.source, nb_node.line = (

        self.state_machine.get_source_and_line(self.lineno))

    # add dependency

    self.state.document.settings.record_dependencies.add(f['nbpath'])

  setup.config = app.config

  setup.confdir = app.confdir

    app.add_node(notebook_node,

                 html=(visit_notebook_node, depart_notebook_node))

  app.add_node(notebook_node, html=(visit_notebook_node, depart_notebook_node))

  app.add_directive('notebook', NotebookDirective)

<!DOCTYPE html>

<html lang="en">

<head>

    <title>PyTorch | About</title>

    <meta charset="utf-8">

    <meta http-equiv="X-UA-Compatible" content="IE=edge">

    <meta name="viewport" content="width=device-width,initial-scale=1,maximum-scale=1,user-scalable=no">

    <link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Open+Sans:200,300,400,600,700">

    <link rel="stylesheet" 

        href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/css/bootstrap.min.css" 

        integrity="sha384-BVYiiSIFeK1dGmJRAkycuHAHRg32OmUcww7on3RYdg4Va+PmSTsz/K68vbdEjh4u" 

        crossorigin="anonymous">

    <link rel="stylesheet" href="./static/css/main.css">

</head>

<body id="about">

    <header>

    <div class="container">

    <div class="logo"><a href="./index.html" class="pull-left"><img src="./static/img/deepchem_logo.png" height="100"></a></a></div>

    <ul class="primary-nav">

    <li><a  href="/">Get Started</a></li>

    <li><a class="active" href="./about.html">About</a></li>

    <li><a  href="/support/">Support</a></li>

    <li><a class="btn" href="./docs/index.html">Docs</a></li>

</ul>

    </div>

</header>

    <section class="content wrap">

        <div class="container">

            <p>DeepChem is a Python package that provides a tool chain for applying deep-learning to problems in drug-discovery, materials science, quantum chemistry, biology, and more.</p>

<ul>

  <li>Tensor computation (like numpy) with strong GPU acceleration</li>

  <li>Deep Neural Networks built on a tape-based autograd system</li>

</ul>

<p>You can reuse your favorite python packages such as numpy, scipy and Cython to extend PyTorch when needed.</p>

<p>At a granular level, PyTorch is a library that consists of the following components:</p>

<table>

  <thead>

    <tr>

      <th>Package</th>

      <th>Description</th>

    </tr>

  </thead>

  <tbody>

    <tr>

      <td>deepchem</td>

      <td>a deep learning library for scientific applications.</td>

    </tr>

    <tr>

      <td>torch.autograd</td>

      <td>a tape based automatic differentiation library that supports all differentiable Tensor operations in torch</td>

    </tr>

    <tr>

      <td>torch.nn</td>

      <td>a neural networks library deeply integrated with autograd designed for maximum flexibility</td>

    </tr>

    <tr>

      <td>torch.optim</td>

      <td>an optimization package to be used with torch.nn with standard optimization methods such as SGD, RMSProp, LBFGS, Adam etc.</td>

    </tr>

    <tr>

      <td>torch.multiprocessing</td>

      <td>python multiprocessing, but with magical memory sharing of torch Tensors across processes. Useful for data loading and hogwild training.</td>

    </tr>

    <tr>

      <td>torch.utils</td>

      <td>DataLoader, Trainer and other utility functions for convenience</td>

    </tr>

    <tr>

      <td>torch.legacy(.nn/.optim)</td>

      <td>legacy code that has been ported over from torch for backward compatibility reasons</td>

    </tr>

  </tbody>

</table>

<p>Usually one uses PyTorch either as:</p>

<ul>

  <li>A replacement for numpy to use the power of GPUs.</li>

  <li>a deep learning research platform that provides maximum flexibility and speed</li>

</ul>

<p>Elaborating further:</p>

<h3 id="a-gpu-ready-tensor-library">A GPU-ready Tensor library</h3>

<p>If you use numpy, then you have used Tensors (a.k.a ndarray).</p>

<p><img src="./static/img/tensor_illustration.png" alt="tensor_illustration" /></p>

<p>PyTorch provides Tensors that can live either on the CPU or the GPU, and accelerate

compute by a huge amount.</p>

<p>We provide a wide variety of tensor routines to accelerate and fit your scientific computation needs

such as slicing, indexing, math operations, linear algebra, reductions.

And they are fast!</p>

<h3 id="dynamic-neural-networks-tape-based-autograd">Dynamic Neural Networks: Tape based Autograd</h3>

<p>PyTorch has a unique way of building neural networks: using and replaying a tape recorder.</p>

<p>Most frameworks such as <code class="highlighter-rouge">TensorFlow</code>, <code class="highlighter-rouge">Theano</code>, <code class="highlighter-rouge">Caffe</code> and <code class="highlighter-rouge">CNTK</code> have a static view of the world.

One has to build a neural network, and reuse the same structure again and again.

Changing the way the network behaves means that one has to start from scratch.</p>

<p>With PyTorch, we use a technique called Reverse-mode auto-differentiation, which allows you to

change the way your network behaves arbitrarily with zero lag or overhead. Our inspiration comes

from several research papers on this topic, as well as current and past work such as

<a href="https://github.com/twitter/torch-autograd">autograd</a>,

<a href="https://github.com/HIPS/autograd">autograd</a>,

<a href="http://chainer.org">Chainer</a>, etc.</p>

<p>While this technique is not unique to PyTorch, it’s one of the fastest implementations of it to date.

You get the best of speed and flexibility for your crazy research.</p>

<p><img src="./static/img/dynamic_graph.gif" alt="dynamic_graph" /></p>

<h3 id="python-first">Python first</h3>

<p>PyTorch is not a Python binding into a monolothic C++ framework.

It is built to be deeply integrated into Python.

You can use it naturally like you would use numpy / scipy / scikit-learn etc.

You can write your new neural network layers in Python itself, using your favorite libraries

and use packages such as Cython and Numba.

Our goal is to not reinvent the wheel where appropriate.</p>

<h3 id="imperative-experiences">Imperative experiences</h3>

<p>PyTorch is designed to be intuitive, linear in thought and easy to use.

When you execute a line of code, it gets executed. There isn’t an asynchronous view of the world.

When you drop into a debugger, or receive error messages and stack traces, understanding them is straight-forward.

The stack-trace points to exactly where your code was defined.

We hope you never spend hours debugging your code because of bad stack traces or asynchronous and opaque execution engines.</p>

<h3 id="fast-and-lean">Fast and Lean</h3>

<p>PyTorch has minimal framework overhead. We integrate acceleration libraries 

such as Intel MKL and NVIDIA (CuDNN, NCCL) to maximize speed. 

At the core, it’s CPU and GPU Tensor and Neural Network backends 

(TH, THC, THNN, THCUNN) are written as independent libraries with a C99 API.<br />

They are mature and have been tested for years.</p>

<p>Hence, PyTorch is quite fast – whether you run small or large neural networks.</p>

<p>The memory usage in PyTorch is extremely efficient compared to Torch or some of the alternatives.

We’ve written custom memory allocators for the GPU to make sure that

your deep learning models are maximally memory efficient.

This enables you to train bigger deep learning models than before.</p>

<h3 id="extensions-without-pain">Extensions without pain</h3>

<p>Writing new neural network modules, or interfacing with PyTorch’s Tensor API was designed to be straight-forward

and with minimal abstractions.</p>

<p>You can write new neural network layers in Python using the torch API

<a href="http://pytorch.org/tutorials/advanced/numpy_extensions_tutorial.html">or your favorite numpy based libraries such as SciPy</a></p>

<p>If you want to write your layers in C/C++, we provide an extension API based on

<a href="http://cffi.readthedocs.io/en/latest/">cffi</a> that is efficient and with minimal boilerplate.<br />

There is no wrapper code that needs to be written. <a href="https://github.com/pytorch/extension-ffi">You can see an example here</a>.</p>

        </div>

    </section>

    <footer>

    <div class="container">

    <div class="left">

        <div class="logo"><a href="/"></a></div>

        <p>

            Maintained by the PyTorch core team. <br>

            ©2017 PyTorch

        </p>

    </div>

    <ul class="primary-nav">

    <li><a  href="/">Get Started</a></li>

    <li><a class="active" href="/about/">About</a></li>

    <li><a  href="/blog/">Blog</a></li>

    <li><a  href="/support/">Support</a></li>

    <li><a  href="https://discuss.pytorch.org">Discuss</a></li>

</ul>

    </div>

</footer>

</body>

</html>