Merge branch 'master' into patch-1

Author: brianjo

.jenkins/build.sh

@@ -12,6 +12,8 @@ export LANG=C.UTF-8
 
 DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null && pwd )"
 
+# Update root certificates by installing new libgnutls30
+sudo apt-get update || sudo apt-get install libgnutls30
 sudo apt-get update
 sudo apt-get install -y --no-install-recommends unzip p7zip-full sox libsox-dev libsox-fmt-all rsync
 

advanced_source/cpp_frontend.rst

@@ -1015,7 +1015,7 @@ probabilities.
 	is called *label smoothing*.
 
 Before evaluating the discriminator, we zero out the gradients of its
-parameters. After computing the loss, we back-propagate through the network by
+parameters. After computing the loss, we back-propagate it through the network by
 calling ``d_loss.backward()`` to compute new gradients. We repeat this spiel for
 the fake images. Instead of using images from the dataset, we let the generator
 create fake images for this by feeding it a batch of random noise. We then

beginner_source/blitz/cifar10_tutorial.py

@@ -22,7 +22,7 @@
 
 Specifically for vision, we have created a package called
 ``torchvision``, that has data loaders for common datasets such as
-Imagenet, CIFAR10, MNIST, etc. and data transformers for images, viz.,
+ImageNet, CIFAR10, MNIST, etc. and data transformers for images, viz.,
 ``torchvision.datasets`` and ``torch.utils.data.DataLoader``.
 
 This provides a huge convenience and avoids writing boilerplate code.

beginner_source/former_torchies/parallelism_tutorial.py

@@ -125,7 +125,7 @@ def forward(self, x):
 #
 # -  :doc:`Train neural nets to play video games </intermediate/reinforcement_q_learning>`
 # -  `Train a state-of-the-art ResNet network on imagenet`_
-# -  `Train an face generator using Generative Adversarial Networks`_
+# -  `Train a face generator using Generative Adversarial Networks`_
 # -  `Train a word-level language model using Recurrent LSTM networks`_
 # -  `More examples`_
 # -  `More tutorials`_
@@ -134,7 +134,7 @@ def forward(self, x):
 #
 # .. _`Deep Learning with PyTorch: a 60-minute blitz`: https://github.com/pytorch/tutorials/blob/master/Deep%20Learning%20with%20PyTorch.ipynb
 # .. _Train a state-of-the-art ResNet network on imagenet: https://github.com/pytorch/examples/tree/master/imagenet
-# .. _Train an face generator using Generative Adversarial Networks: https://github.com/pytorch/examples/tree/master/dcgan
+# .. _Train a face generator using Generative Adversarial Networks: https://github.com/pytorch/examples/tree/master/dcgan
 # .. _Train a word-level language model using Recurrent LSTM networks: https://github.com/pytorch/examples/tree/master/word_language_model
 # .. _More examples: https://github.com/pytorch/examples
 # .. _More tutorials: https://github.com/pytorch/tutorials

beginner_source/ptcheat.rst

@@ -237,7 +237,7 @@ Learning rate scheduling
 .. code-block:: python
 
     scheduler = optim.X(optimizer,...)      # create lr scheduler
-    scheduler.step()                        # update lr at start of epoch
+    scheduler.step()                        # update lr after optimizer updates weights
     optim.lr_scheduler.X                    # where X is LambdaLR, MultiplicativeLR,
                                             # StepLR, MultiStepLR, ExponentialLR,
                                             # CosineAnnealingLR, ReduceLROnPlateau, CyclicLR,

intermediate_source/rpc_async_execution.rst

@@ -14,7 +14,7 @@ This tutorial demonstrates how to build batch-processing RPC applications with
 the `@rpc.functions.async_execution <https://pytorch.org/docs/master/rpc.html#torch.distributed.rpc.functions.async_execution>`__
 decorator, which helps to speed up training by reducing the number of blocked
 RPC threads and consolidating CUDA operations on the callee. This shares the
-same idea as `Batch Inference with TorchServer <https://pytorch.org/serve/batch_inference_with_ts.html>`__.
+same idea as `Batch Inference with TorchServe <https://pytorch.org/serve/batch_inference_with_ts.html>`__.
 
 .. note:: This tutorial requires PyTorch v1.6.0 or above.
 
@@ -199,7 +199,7 @@ Batch-Processing CartPole Solver
 
 This section uses CartPole-v1 from `OpenAI Gym <https://gym.openai.com/>`__ as
 an example to show the performance impact of batch processing RPC. Please note
-that the goal is to demonstrate the usage of
+that since the goal is to demonstrate the usage of
 `@rpc.functions.async_execution <https://pytorch.org/docs/master/rpc.html#torch.distributed.rpc.functions.async_execution>`__
 instead of building the best CartPole solver or solving most different RL
 problems, we use very simple policies and reward calculation strategies and

intermediate_source/torchvision_tutorial.rst

@@ -220,7 +220,7 @@ way of doing it:
    import torchvision
    from torchvision.models.detection.faster_rcnn import FastRCNNPredictor
 
-   # load a model pre-trained pre-trained on COCO
+   # load a model pre-trained on COCO
    model = torchvision.models.detection.fasterrcnn_resnet50_fpn(pretrained=True)
 
    # replace the classifier with a new one, that has
@@ -290,7 +290,7 @@ be using Mask R-CNN:
 
 
    def get_model_instance_segmentation(num_classes):
-       # load an instance segmentation model pre-trained pre-trained on COCO
+       # load an instance segmentation model pre-trained on COCO
        model = torchvision.models.detection.maskrcnn_resnet50_fpn(pretrained=True)
 
        # get number of input features for the classifier

recipes_source/recipes/loading_data_recipe.py

@@ -14,9 +14,9 @@
 `torch.utils.data.Dataset <https://pytorch.org/docs/stable/data.html#torch.utils.data.Dataset>`__.
 These datasets are currently available in:
 
-* `torchvision <https://pytorch.org/docs/stable/torchvision/datasets.html>`__
-* `torchaudio <https://pytorch.org/audio/datasets.html>`__
-* `torchtext <https://pytorch.org/text/datasets.html>`__
+* `torchvision <https://pytorch.org/vision/stable/datasets.html>`__
+* `torchaudio <https://pytorch.org/audio/stable/datasets.html>`__
+* `torchtext <https://pytorch.org/text/stable/datasets.html>`__
 
 with more to come.
 Using the Yesno dataset from ``torchaudio.datasets.YESNO``, we will