fix missed code-blocks because of syntax error (#1496)

advanced_source/extend_dispatcher.rst

@@ -53,6 +53,7 @@ You can choose any of keys above to prototype your customized backend.
 To create a Tensor on ``PrivateUse1`` backend, you need to set dispatch key in ``TensorImpl`` constructor.
 
 .. code-block:: cpp
+
   /* Example TensorImpl constructor */
   TensorImpl(
       Storage&& storage,

advanced_source/torch-script-parallelism.rst

@@ -207,6 +207,7 @@ Let's use the profiler along with the Chrome trace export functionality to
 visualize the performance of our parallelized model:
 
 .. code-block:: python
+
     with torch.autograd.profiler.profile() as prof:
         ens(x)
     prof.export_chrome_trace('parallel.json')

advanced_source/torch_script_custom_ops.rst

@@ -605,7 +605,7 @@ Along with a small ``CMakeLists.txt`` file:
 
 At this point, we should be able to build the application:
 
-.. code-block::
+.. code-block:: shell
 
   $ mkdir build
   $ cd build
@@ -645,7 +645,7 @@ At this point, we should be able to build the application:
 
 And run it without passing a model just yet:
 
-.. code-block::
+.. code-block:: shell
 
   $ ./example_app
   usage: example_app <path-to-exported-script-module>
@@ -672,7 +672,7 @@ The last line will serialize the script function into a file called
 "example.pt". If we then pass this serialized model to our C++ application, we
 can run it straight away:
 
-.. code-block::
+.. code-block:: shell
 
   $ ./example_app example.pt
   terminate called after throwing an instance of 'torch::jit::script::ErrorReport'

beginner_source/hyperparameter_tuning_tutorial.py

@@ -431,7 +431,7 @@ def main(num_samples=10, max_num_epochs=10, gpus_per_trial=2):
 ######################################################################
 # If you run the code, an example output could look like this:
 #
-# .. code-block::
+# ::
 #
 #     Number of trials: 10 (10 TERMINATED)
 #     +-----+------+------+-------------+--------------+---------+------------+--------------------+

prototype_source/fx_graph_mode_ptq_static.rst

@@ -10,7 +10,8 @@ we'll have a separate tutorial to show how to make the part of the model we want
 We also have a tutorial for `FX Graph Mode Post Training Dynamic Quantization <https://pytorch.org/tutorials/prototype/fx_graph_mode_ptq_dynamic.html>`_.
 tldr; The FX Graph Mode API looks like the following:
 
-.. code:: python    
+.. code:: python
+
   import torch    
   from torch.quantization import get_default_qconfig  
   # Note that this is temporary, we'll expose these functions to torch.quantization after official releasee   

recipes_source/android_native_app_with_custom_op.rst

@@ -704,7 +704,7 @@ If you check the android logcat:
 
 You should see logs with tag 'PyTorchNativeApp' that prints x, y, and the result of the model forward, which we print with ``log`` function in ``NativeApp/app/src/main/cpp/pytorch_nativeapp.cpp``.
 
-.. code-block::
+::
 
   I/PyTorchNativeApp(26968): x: -0.9484 -1.1757 -0.5832  0.9144  0.8867  1.0933 -0.4004 -0.3389
   I/PyTorchNativeApp(26968): -1.0343  1.5200 -0.7625 -1.5724 -1.2073  0.4613  0.2730 -0.6789