[docs] add a Pipeline tutorial (#82)

.isort.cfg

 [settings]
-known_third_party =numpy,pytest,pytorch_sphinx_theme,recommonmark,setuptools,torch,torchtext,torchvision
+known_third_party =numpy,pytest,recommonmark,setuptools,torch,torchtext,torchvision

docs/source/api/index.rst

+API Reference
+=============
+
+.. toctree::
+   :maxdepth: 1
+
+   optim/oss
+   nn/pipe

docs/source/index.rst

@@ -7,16 +7,7 @@ Welcome to fairscale's documentation!
 =====================================
 
 .. toctree::
-   :maxdepth: 1
-   :caption: API Reference
+   :maxdepth: 2
 
-   api/optim/oss
-   api/nn/pipe
-
-
-Indices and tables
-==================
-
-* :ref:`genindex`
-* :ref:`modindex`
-* :ref:`search`
+   api/index
+   tutorials/index

docs/source/tutorials/index.rst

+Tutorials
+=========
+
+.. toctree::
+   :maxdepth: 1
+
+   pipe

docs/source/tutorials/pipe.rst

+Pipeline Parallel
+=================
+
+Let us start with a toy model that contains two linear layers.
+
+.. code-block:: default
+
+
+    import torch
+    import torch.nn as nn
+
+    class ToyModel(nn.Module):
+        def __init__(self):
+            super(ToyModel, self).__init__()
+            self.net1 = torch.nn.Linear(10, 10)
+            self.relu = torch.nn.ReLU()
+            self.net2 = torch.nn.Linear(10, 5)
+
+        def forward(self, x):
+            x = self.relu(self.net1(x))
+            return self.net2(x)
+
+    model = ToyModel()
+
+To run this model on 2 GPUs we need to convert the model
+to ``torch.nn.Sequential`` and then wrap it with ``fairscale.nn.Pipe``.
+
+.. code-block:: default
+
+
+    import fairscale
+    import torch
+    import torch.nn as nn
+
+    model = nn.Sequential(
+                torch.nn.Linear(10, 10),
+                torch.nn.ReLU(),
+                torch.nn.Linear(10, 5)
+            )
+
+    model = fairscale.nn.Pipe(model, balance=[2, 1])
+
+This will run the first two layers on ``cuda:0`` and the last
+layer on ``cuda:1``. To learn more, visit the `Pipe <../api/nn/pipe.html>`_ documentation.