Merge branch 'main' into patch-3

Author: Svetlana Karslioglu

.github/PULL_REQUEST_TEMPLATE.md

@@ -8,4 +8,4 @@ Fixes #ISSUE_NUMBER
 - [ ] The issue that is being fixed is referred in the description (see above "Fixes #ISSUE_NUMBER")
 - [ ] Only one issue is addressed in this pull request
 - [ ] Labels from the issue that this PR is fixing are added to this pull request
-- [ ] No unnessessary issues are included into this pull request.
+- [ ] No unnecessary issues are included into this pull request.

.github/scripts/docathon-label-sync.py

@@ -14,6 +14,9 @@ def main():
     repo = g.get_repo(f'{repo_owner}/{repo_name}')
     pull_request = repo.get_pull(pull_request_number)
     pull_request_body = pull_request.body
+    # PR without description
+    if pull_request_body is None:
+        return
 
     # get issue number from the PR body
     if not re.search(r'#\d{1,5}', pull_request_body):

beginner_source/finetuning_torchvision_models_tutorial.rst

@@ -0,0 +1,10 @@
+Finetuning Torchvision Models
+=============================
+
+This tutorial has been moved to https://pytorch.org/tutorials/intermediate/torchvision_tutorial.html
+
+It will redirect in 3 seconds.
+
+.. raw:: html
+
+   <meta http-equiv="Refresh" content="3; url='https://pytorch.org/tutorials/intermediate/torchvision_tutorial.html'" />

beginner_source/introyt/tensorboardyt_tutorial.py

@@ -64,6 +64,13 @@
 # PyTorch TensorBoard support
 from torch.utils.tensorboard import SummaryWriter
 
+# In case you are using an environment that has TensorFlow installed,
+# such as Google Colab, uncomment the following code to avoid
+# a bug with saving embeddings to your TensorBoard directory
+
+# import tensorflow as tf
+# import tensorboard as tb
+# tf.io.gfile = tb.compat.tensorflow_stub.io.gfile
 
 ######################################################################
 # Showing Images in TensorBoard

beginner_source/nn_tutorial.py

@@ -795,8 +795,7 @@ def __len__(self):
         return len(self.dl)
 
     def __iter__(self):
-        batches = iter(self.dl)
-        for b in batches:
+        for b in self.dl:
             yield (self.func(*b))
 
 train_dl, valid_dl = get_data(train_ds, valid_ds, bs)

beginner_source/transformer_tutorial.py

@@ -149,7 +149,7 @@ def forward(self, x: Tensor) -> Tensor:
 # into ``batch_size`` columns. If the data does not divide evenly into
 # ``batch_size`` columns, then the data is trimmed to fit. For instance, with
 # the alphabet as the data (total length of 26) and ``batch_size=4``, we would
-# divide the alphabet into 4 sequences of length 6:
+# divide the alphabet into sequences of length 6, resulting in 4 of such sequences.
 #
 # .. math::
 #   \begin{bmatrix}

intermediate_source/mario_rl_tutorial.py

@@ -711,17 +711,18 @@ def record(self, episode, epsilon, step):
                 f"{datetime.datetime.now().strftime('%Y-%m-%dT%H:%M:%S'):>20}\n"
             )
 
-        for metric in ["ep_rewards", "ep_lengths", "ep_avg_losses", "ep_avg_qs"]:
-            plt.plot(getattr(self, f"moving_avg_{metric}"))
-            plt.savefig(getattr(self, f"{metric}_plot"))
+        for metric in ["ep_lengths", "ep_avg_losses", "ep_avg_qs", "ep_rewards"]:
             plt.clf()
+            plt.plot(getattr(self, f"moving_avg_{metric}"), label=f"moving_avg_{metric}")
+            plt.legend()
+            plt.savefig(getattr(self, f"{metric}_plot"))
 
 
 ######################################################################
 # Let’s play!
 # """""""""""""""
 #
-# In this example we run the training loop for 10 episodes, but for Mario to truly learn the ways of
+# In this example we run the training loop for 40 episodes, but for Mario to truly learn the ways of
 # his world, we suggest running the loop for at least 40,000 episodes!
 #
 use_cuda = torch.cuda.is_available()
@@ -735,7 +736,7 @@ def record(self, episode, epsilon, step):
 
 logger = MetricLogger(save_dir)
 
-episodes = 10
+episodes = 40
 for e in range(episodes):
 
     state = env.reset()

intermediate_source/tensorboard_profiler_tutorial.py

@@ -18,7 +18,7 @@
 -----
 To install ``torch`` and ``torchvision`` use the following command:
 
-::
+.. code-block::
 
    pip install torch torchvision
 
@@ -160,23 +160,23 @@ def train(data):
 #
 # Install PyTorch Profiler TensorBoard Plugin.
 #
-# ::
+# .. code-block::
 #
 #     pip install torch_tb_profiler
 #
 
 ######################################################################
 # Launch the TensorBoard.
 #
-# ::
+# .. code-block::
 #
 #     tensorboard --logdir=./log
 #
 
 ######################################################################
 # Open the TensorBoard profile URL in Google Chrome browser or Microsoft Edge browser.
 #
-# ::
+# .. code-block::
 #
 #     http://localhost:6006/#pytorch_profiler
 #
@@ -287,7 +287,7 @@ def train(data):
 # In this example, we follow the "Performance Recommendation" and set ``num_workers`` as below,
 # pass a different name such as ``./log/resnet18_4workers`` to ``tensorboard_trace_handler``, and run it again.
 #
-# ::
+# .. code-block::
 #
 #     train_loader = torch.utils.data.DataLoader(train_set, batch_size=32, shuffle=True, num_workers=4)
 #
@@ -316,7 +316,7 @@ def train(data):
 #
 # You can try it by using existing example on Azure
 #
-# ::
+# .. code-block::
 #
 #     pip install azure-storage-blob
 #     tensorboard --logdir=https://torchtbprofiler.blob.core.windows.net/torchtbprofiler/demo/memory_demo_1_10
@@ -366,7 +366,7 @@ def train(data):
 #
 # You can try it by using existing example on Azure:
 #
-# ::
+# .. code-block::
 #
 #     pip install azure-storage-blob
 #     tensorboard --logdir=https://torchtbprofiler.blob.core.windows.net/torchtbprofiler/demo/distributed_bert

prototype_source/README.txt

@@ -1,8 +1,8 @@
 Prototype Tutorials
 ------------------
 1. distributed_rpc_profiling.rst
-     Profiling PyTorch RPC-Based Workloads
-     https://github.com/pytorch/tutorials/blob/release/1.6/prototype_source/distributed_rpc_profiling.rst
+           Profiling PyTorch RPC-Based Workloads
+           https://github.com/pytorch/tutorials/blob/main/prototype_source/distributed_rpc_profiling.rst
 
 2. graph_mode_static_quantization_tutorial.py
 	   Graph Mode Post Training Static Quantization in PyTorch
@@ -21,8 +21,8 @@ Prototype Tutorials
 	   https://github.com/pytorch/tutorials/blob/main/prototype_source/torchscript_freezing.py
 
 6. vulkan_workflow.rst
-     Vulkan Backend User Workflow
-     https://pytorch.org/tutorials/intermediate/vulkan_workflow.html
+           Vulkan Backend User Workflow
+           https://pytorch.org/tutorials/intermediate/vulkan_workflow.html
 
 7. fx_graph_mode_ptq_static.rst
 	   FX Graph Mode Post Training Static Quantization

prototype_source/fx_graph_mode_quant_guide.rst

@@ -4,7 +4,7 @@
 **Author**: `Jerry Zhang <https://github.com/jerryzh168>`_
 
 FX Graph Mode Quantization requires a symbolically traceable model.
-We use the FX framework (TODO: link) to convert a symbolically traceable nn.Module instance to IR,
+We use the FX framework to convert a symbolically traceable nn.Module instance to IR,
 and we operate on the IR to execute the quantization passes.
 Please post your question about symbolically tracing your model in `PyTorch Discussion Forum <https://discuss.pytorch.org/c/quantization/17>`_
 
@@ -22,16 +22,19 @@ You can use any combination of these options:
     b. Write your own observed and quantized submodule
 
 
-####################################################################
 If the code that is not symbolically traceable does not need to be quantized, we have the following two options
 to run FX Graph Mode Quantization:
-1.a. Symbolically trace only the code that needs to be quantized
+
+
+Symbolically trace only the code that needs to be quantized
 -----------------------------------------------------------------
 When the whole model is not symbolically traceable but the submodule we want to quantize is
 symbolically traceable, we can run quantization only on that submodule.
+
 before:
 
 .. code:: python
+
   class M(nn.Module):
       def forward(self, x):
           x = non_traceable_code_1(x)
@@ -42,6 +45,7 @@ before:
 after:
 
 .. code:: python
+    
   class FP32Traceable(nn.Module):
       def forward(self, x):
           x = traceable_code(x)
@@ -69,8 +73,7 @@ Note if original model needs to be preserved, you will have to
 copy it yourself before calling the quantization APIs.
 
 
-#####################################################
-1.b. Skip symbolically trace the non-traceable code
+Skip symbolically trace the non-traceable code
 ---------------------------------------------------
 When we have some non-traceable code in the module, and this part of code doesn’t need to be quantized,
 we can factor out this part of the code into a submodule and skip symbolically trace that submodule.
@@ -134,8 +137,7 @@ quantization code:
 
 If the code that is not symbolically traceable needs to be quantized, we have the following two options:
 
-##########################################################
-2.a Refactor your code to make it symbolically traceable
+Refactor your code to make it symbolically traceable
 --------------------------------------------------------
 If it is easy to refactor the code and make the code symbolically traceable,
 we can refactor the code and remove the use of non-traceable constructs in python.
@@ -167,15 +169,10 @@ after:
       return x.permute(0, 2, 1, 3)
 
 
-quantization code:
-
 This can be combined with other approaches and the quantization code
 depends on the model.
 
-
-
-#######################################################
-2.b. Write your own observed and quantized submodule
+Write your own observed and quantized submodule
 -----------------------------------------------------
 
 If the non-traceable code can’t be refactored to be symbolically traceable,
@@ -207,8 +204,8 @@ non-traceable logic, wrapped in a module
   class FP32NonTraceable:
       ...
 
-
-2. Define observed version of FP32NonTraceable
+2. Define observed version of
+FP32NonTraceable
 
 .. code:: python