Merge branch 'master' into SethHWeidman-patch-1

mthrok · Sep 12, 2019 · 95df504 · 95df504
2 parents 6bf6bb0 + d214b12
commit 95df504
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 4 deletions.
diff --git a/intermediate_source/dist_tuto.rst b/intermediate_source/dist_tuto.rst
@@ -1,4 +1,4 @@
-Writing Distributed Applications with PyTorch
+3. Writing Distributed Applications with PyTorch
 =============================================
 **Author**: `Séb Arnold <https://seba1511.com>`_
 

diff --git a/intermediate_source/model_parallel_tutorial.py b/intermediate_source/model_parallel_tutorial.py
@@ -28,7 +28,7 @@
 applications.
 
 Basic Usage
-================================
+-----------
 """
 
 ######################################################################
@@ -75,7 +75,7 @@ def forward(self, x):
 
 ######################################################################
 # Apply Model Parallel to Existing Modules
-# =======================
+# ----------------------------------------
 #
 # It is also possible to run an existing single-GPU module on multiple GPUs
 # with just a few lines of changes. The code below shows how to decompose
@@ -235,7 +235,7 @@ def plot(means, stds, labels, fig_name):
 
 ######################################################################
 # Speed Up by Pipelining Inputs
-# =======================
+# -----------------------------
 #
 # In the following experiments, we further divide each 120-image batch into
 # 20-image splits. As PyTorch launches CUDA operations asynchronizely, the
@@ -350,3 +350,4 @@ def forward(self, x):
 # for your environment, a proper approach is to first generate the curve to
 # figure out the best split size, and then use that split size to pipeline
 # inputs.
+#