Release/1.6 (pytorch#1087)

* Add TorchScript fork/join tutorial

* Add note about zipfile format in serialization tutorial

* Profiler recipe (pytorch#1019)

* Profiler recipe

Summary:
Adding a recipe for profiler

Test Plan:
make html-noplot

* [mobile] Mobile Perf Recipe

* Minor syntax edits to mobile perf recipe

* Remove built files

* [android] android native app recipe

* [mobile_perf][recipe] Add ChannelsLast recommendation

* Adding distributed pipeline parallel tutorial

* Add async execution tutorials

* Fix code block in pipeline tutorial

* Adding an Overview Page for PyTorch Distributed (pytorch#1056)

* Adding an Overview Page for PyTorch Distributed

* Let existing PT Distributed tutorials link to the overview page

* Add a link to AMP

* Address Comments

* Remove unnecessary dist.barrier()

* [Mobile Perf Recipe] Add the benchmarking part for iOS (pytorch#1055)

* [Mobile Perf Recipe] Add the benchmarking part for iOS

* [Mobile Perf Recipe] Add the benchmarking part for iOS

Co-authored-by: Jessica Lin <jplin@fb.com>

* RPC profiling recipe (pytorch#1068)

* Initial commit

* Update

* Complete most of recipe

* Add image

* Link image

* Remove extra file

* update

* Update

* update

* Push latest changes from master into release/1.6 (pytorch#1074)

* Update feature classification labels

* Update NVidia -> Nvidia

* Bring back default filename_pattern so that by default we run all galleries.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Add prototype_source directory

* Add prototype directory

* Add prototype

* Remove extra "done"

* Add REAME.txt

* Update for prototype instructions

* Update for prototype feature

* refine torchvision_tutorial doc for windows

* Update neural_style_tutorial.py (pytorch#1059)

Updated the mistake in the Loading Images Section.

* torch_script_custom_ops restructure (pytorch#1057)

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Port custom ops tutorial to new registration API, increase testability.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Kill some other occurrences of RegisterOperators

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Update README.md

* Make torch_script_custom_classes tutorial runnable

I also fixed some warnings in the tutorial, and fixed some minor bitrot
(e.g., torch::script::Module to torch::jit::Module)

I also added some missing quotes around some bash expansions.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Update torch_script_custom_classes to use TORCH_LIBRARY (pytorch#1062)

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

Co-authored-by: Edward Z. Yang <ezyang@fb.com>
Co-authored-by: Yang Gu <yangu@microsoft.com>
Co-authored-by: Hritik Bhandari <bhandari.hritik@gmail.com>

* Tutorial for DDP + RPC (pytorch#1071)

* Update feature classification labels

* Update NVidia -> Nvidia

* Bring back default filename_pattern so that by default we run all galleries.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Tutorial for DDP + RPC.

Summary: Based on example from pytorch/examples#800

* Add to main section

Summary:

Test Plan:

Reviewers:

Subscribers:

Tasks:

Tags:

* Added separate code file and used literalinclude

Summary:

Test Plan:

Reviewers:

Subscribers:

Tasks:

Tags:

Co-authored-by: Jessica Lin <jplin@fb.com>
Co-authored-by: Edward Z. Yang <ezyang@fb.com>
Co-authored-by: pritam <pritam.damania@fb.com>

* Make RPC profiling recipe into prototype tutorial (pytorch#1078)

* Add RPC tutorial

* Update to include recipes

* Add Graph Mode Dynamic Quant tutorial (pytorch#1065)

* Update feature classification labels

* Update NVidia -> Nvidia

* Bring back default filename_pattern so that by default we run all galleries.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Add prototype_source directory

* Add prototype directory

* Add prototype

* Remove extra "done"

* Add REAME.txt

* Update for prototype instructions

* Update for prototype feature

* refine torchvision_tutorial doc for windows

* Update neural_style_tutorial.py (pytorch#1059)

Updated the mistake in the Loading Images Section.

* torch_script_custom_ops restructure (pytorch#1057)

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Port custom ops tutorial to new registration API, increase testability.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Kill some other occurrences of RegisterOperators

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Update README.md

* Make torch_script_custom_classes tutorial runnable

I also fixed some warnings in the tutorial, and fixed some minor bitrot
(e.g., torch::script::Module to torch::jit::Module)

I also added some missing quotes around some bash expansions.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Update torch_script_custom_classes to use TORCH_LIBRARY (pytorch#1062)

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Add Graph Mode Dynamic Quant tutorial

Summary:
Tutorial to demonstrate graph mode dynamic quant on BERT model.
Currently not directly runnable as it requires to download glue dataset and fine-tuned model

Co-authored-by: Jessica Lin <jplin@fb.com>
Co-authored-by: Edward Z. Yang <ezyang@fb.com>
Co-authored-by: Yang Gu <yangu@microsoft.com>
Co-authored-by: Hritik Bhandari <bhandari.hritik@gmail.com>

* Add mobile recipes images

* Update mobile recipe index

* Remove RPC Profiling recipe from index

* 1.6 model freezing tutorial (pytorch#1077)

* Update feature classification labels

* Update NVidia -> Nvidia

* Bring back default filename_pattern so that by default we run all galleries.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Add prototype_source directory

* Add prototype directory

* Add prototype

* Remove extra "done"

* Add REAME.txt

* Update for prototype instructions

* Update for prototype feature

* refine torchvision_tutorial doc for windows

* Update neural_style_tutorial.py (pytorch#1059)

Updated the mistake in the Loading Images Section.

* torch_script_custom_ops restructure (pytorch#1057)

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Port custom ops tutorial to new registration API, increase testability.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Kill some other occurrences of RegisterOperators

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Update README.md

* Make torch_script_custom_classes tutorial runnable

I also fixed some warnings in the tutorial, and fixed some minor bitrot
(e.g., torch::script::Module to torch::jit::Module)

I also added some missing quotes around some bash expansions.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Update torch_script_custom_classes to use TORCH_LIBRARY (pytorch#1062)

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Add Model Freezing in TorchScript

Co-authored-by: Edward Z. Yang <ezyang@fb.com>
Co-authored-by: Yang Gu <yangu@microsoft.com>
Co-authored-by: Hritik Bhandari <bhandari.hritik@gmail.com>

* Update title

* Update recipes_index.rst

Touch for rebuild.

* Update dcgan_faces_tutorial.py

Update labels to be floats to work around torch.full inference change.

Co-authored-by: James Reed <jamesreed@fb.com>
Co-authored-by: ilia-cher <30845429+ilia-cher@users.noreply.github.com>
Co-authored-by: Ivan Kobzarev <ivankobzarev@fb.com>
Co-authored-by: Shen Li <shenli@devfair017.maas>
Co-authored-by: Shen Li <cs.shenli@gmail.com>
Co-authored-by: Tao Xu <taox@fb.com>
Co-authored-by: Rohan Varma <rvarm1@fb.com>
Co-authored-by: Edward Z. Yang <ezyang@fb.com>
Co-authored-by: Yang Gu <yangu@microsoft.com>
Co-authored-by: Hritik Bhandari <bhandari.hritik@gmail.com>
Co-authored-by: Pritam Damania <9958665+pritamdamania87@users.noreply.github.com>
Co-authored-by: pritam <pritam.damania@fb.com>
Co-authored-by: supriyar <supriyar@fb.com>
Co-authored-by: Brian Johnson <brianjo@fb.com>
Co-authored-by: gchanan <gchanan@fb.com>

advanced_source/rpc_ddp_tutorial.rst

@@ -0,0 +1,159 @@
+Combining Distributed DataParallel with Distributed RPC Framework
+=================================================================
+**Author**: `Pritam Damania <https://github.com/pritamdamania87>`_
+
+
+This tutorial uses a simple example to demonstrate how you can combine 
+`DistributedDataParallel <https://pytorch.org/docs/stable/nn.html#torch.nn.parallel.DistributedDataParallel>`__ (DDP)
+with the `Distributed RPC framework <https://pytorch.org/docs/master/rpc.html>`__ 
+to combine distributed data parallelism with distributed model parallelism to 
+train a simple model. Source code of the example can be found `here <https://github.com/pytorch/examples/tree/master/distributed/rpc/ddp_rpc>`__.
+
+Previous tutorials,
+`Getting Started With Distributed Data Parallel <https://pytorch.org/tutorials/intermediate/ddp_tutorial.html>`__
+and `Getting Started with Distributed RPC Framework <https://pytorch.org/tutorials/intermediate/rpc_tutorial.html>`__,
+described how to perform distributed data parallel and distributed model 
+parallel training respectively. Although, there are several training paradigms 
+where you might want to combine these two techniques. For example:
+
+1) If we have a model with a sparse part (large embedding table) and a dense 
+   part (FC layers), we might want to put the embedding table on a parameter 
+   server and replicate the FC layer across multiple trainers using `DistributedDataParallel <https://pytorch.org/docs/stable/nn.html#torch.nn.parallel.DistributedDataParallel>`__.
+   The `Distributed RPC framework <https://pytorch.org/docs/master/rpc.html>`__ 
+   can be used to perform embedding lookups on the parameter server.
+2) Enable hybrid parallelism as described in the `PipeDream <https://arxiv.org/abs/1806.03377>`__ paper.
+   We can use the `Distributed RPC framework <https://pytorch.org/docs/master/rpc.html>`__ 
+   to pipeline stages of the model across multiple workers and replicate each 
+   stage (if needed) using `DistributedDataParallel <https://pytorch.org/docs/stable/nn.html#torch.nn.parallel.DistributedDataParallel>`__.
+
+|
+In this tutorial we will cover case 1 mentioned above. We have a total of 4 
+workers in our setup as follows:
+
+
+1) 1 Master, which is responsible for creating an embedding table 
+   (nn.EmbeddingBag) on the parameter server. The master also drives the 
+   training loop on the two trainers.
+2) 1 Parameter Server, which basically holds the embedding table in memory and 
+   responds to RPCs from the Master and Trainers.
+3) 2 Trainers, which store an FC layer (nn.Linear) which is replicated amongst 
+   themselves using `DistributedDataParallel <https://pytorch.org/docs/stable/nn.html#torch.nn.parallel.DistributedDataParallel>`__.
+   The trainers are also responsible for executing the forward pass, backward 
+   pass and optimizer step.
+
+|
+The entire training process is executed as follows:
+
+1) The master creates an embedding table on the Parameter Server and holds an 
+   `RRef <https://pytorch.org/docs/master/rpc.html#rref>`__ to it.
+2) The master, then kicks off the training loop on the trainers and passes the 
+   embedding table RRef to the trainers.
+3) The trainers create a ``HybridModel`` which first performs an embedding lookup 
+   using the embedding table RRef provided by the master and then executes the 
+   FC layer which is wrapped inside DDP.
+4) The trainer executes the forward pass of the model and uses the loss to 
+   execute the backward pass using `Distributed Autograd <https://pytorch.org/docs/master/rpc.html#distributed-autograd-framework>`__.
+5) As part of the backward pass, the gradients for the FC layer are computed 
+   first and synced to all trainers via allreduce in DDP.
+6) Next, Distributed Autograd propagates the gradients to the parameter server, 
+   where the gradients for the embedding table are updated.
+7) Finally, the `Distributed Optimizer <https://pytorch.org/docs/master/rpc.html#module-torch.distributed.optim>`__ is used to update all the parameters.
+
+
+.. attention::
+
+  You should always use `Distributed Autograd <https://pytorch.org/docs/master/rpc.html#distributed-autograd-framework>`__ 
+  for the backward pass if you're combining DDP and RPC.
+
+
+Now, let's go through each part in detail. Firstly, we need to setup all of our 
+workers before we can perform any training. We create 4 processes such that 
+ranks 0 and 1 are our trainers, rank 2 is the master and rank 3 is the 
+parameter server.
+
+We initialize the RPC framework on all 4 workers using the TCP init_method. 
+Once RPC initialization is done, the master creates an `EmbeddingBag <https://pytorch.org/docs/master/generated/torch.nn.EmbeddingBag.html>`__ 
+on the Parameter Server using `rpc.remote <https://pytorch.org/docs/master/rpc.html#torch.distributed.rpc.remote>`__.
+The master then loops through each trainer and kicks of the training loop by 
+calling ``_run_trainer`` on each trainer using `rpc_async <https://pytorch.org/docs/master/rpc.html#torch.distributed.rpc.rpc_async>`__.
+Finally, the master waits for all training to finish before exiting.
+
+The trainers first initialize a ``ProcessGroup`` for DDP with world_size=2 
+(for two trainers) using `init_process_group <https://pytorch.org/docs/stable/distributed.html#torch.distributed.init_process_group>`__.
+Next, they initialize the RPC framework using the TCP init_method. Note that 
+the ports are different in RPC initialization and ProcessGroup initialization. 
+This is to avoid port conflicts between initialization of both frameworks. 
+Once the initialization is done, the trainers just wait for the ``_run_trainer`` 
+RPC from the master.
+
+The parameter server just initializes the RPC framework and waits for RPCs from 
+the trainers and master.
+
+
+.. literalinclude:: ../advanced_source/rpc_ddp_tutorial/main.py
+  :language: py
+  :start-after: BEGIN run_worker
+  :end-before: END run_worker
+
+Before we discuss details of the Trainer, let's introduce the ``HybridModel`` that 
+the trainer uses. As described below, the ``HybridModel`` is initialized using an 
+RRef to the embedding table (emb_rref) on the parameter server and the ``device`` 
+to use for DDP. The initialization of the model wraps an 
+`nn.Linear <https://pytorch.org/docs/master/generated/torch.nn.Linear.html>`__ 
+layer inside DDP to replicate and synchronize this layer across all trainers.
+
+The forward method of the model is pretty straightforward. It performs an 
+embedding lookup on the parameter server using an 
+`RRef helper <https://pytorch.org/docs/master/rpc.html#torch.distributed.rpc.RRef.rpc_sync>`__ 
+and passes its output onto the FC layer.
+
+
+.. literalinclude:: ../advanced_source/rpc_ddp_tutorial/main.py
+  :language: py
+  :start-after: BEGIN hybrid_model
+  :end-before: END hybrid_model
+
+Next, let's look at the setup on the Trainer. The trainer first creates the 
+``HybridModel`` described above using an RRef to the embedding table on the 
+parameter server and its own rank.
+
+Now, we need to retrieve a list of RRefs to all the parameters that we would 
+like to optimize with `DistributedOptimizer <https://pytorch.org/docs/master/rpc.html#module-torch.distributed.optim>`__. 
+To retrieve the parameters for the embedding table from the parameter server, 
+we define a simple helper function ``_retrieve_embedding_parameters``, which 
+basically walks through all the parameters for the embedding table and returns 
+a list of RRefs. The trainer calls this method on the parameter server via RPC 
+to receive a list of RRefs to the desired parameters. Since the 
+DistributedOptimizer always takes a list of RRefs to parameters that need to 
+be optimized, we need to create RRefs even for the local parameters for our 
+FC layers. This is done by walking ``model.parameters()``, creating an RRef for 
+each parameter and appending it to a list. Note that ``model.parameters()`` only 
+returns local parameters and doesn't include ``emb_rref``.
+
+Finally, we create our DistributedOptimizer using all the RRefs and define a 
+CrossEntropyLoss function.
+
+.. literalinclude:: ../advanced_source/rpc_ddp_tutorial/main.py
+  :language: py
+  :start-after: BEGIN setup_trainer 
+  :end-before: END setup_trainer
+
+Now we're ready to introduce the main training loop that is run on each trainer. 
+``get_next_batch`` is just a helper function to generate random inputs and 
+targets for training. We run the training loop for multiple epochs and for each 
+batch:
+
+1) Setup a `Distributed Autograd Context <https://pytorch.org/docs/master/rpc.html#torch.distributed.autograd.context>`__ 
+   for Distributed Autograd.
+2) Run the forward pass of the model and retrieve its output.
+3) Compute the loss based on our outputs and targets using the loss function.
+4) Use Distributed Autograd to execute a distributed backward pass using the loss.
+5) Finally, run a Distributed Optimizer step to optimize all the parameters.
+
+.. literalinclude:: ../advanced_source/rpc_ddp_tutorial/main.py
+  :language: py
+  :start-after: BEGIN run_trainer
+  :end-before: END run_trainer
+.. code:: python
+
+Source code for the entire example can be found `here <https://github.com/pytorch/examples/tree/master/distributed/rpc/ddp_rpc>`__.

advanced_source/rpc_ddp_tutorial/main.py

@@ -0,0 +1,191 @@
+import os
+from functools import wraps
+
+import random
+import torch
+import torch.distributed as dist
+import torch.distributed.autograd as dist_autograd
+import torch.distributed.rpc as rpc
+from torch.distributed.rpc import ProcessGroupRpcBackendOptions
+import torch.multiprocessing as mp
+import torch.optim as optim
+from torch.distributed.optim import DistributedOptimizer
+from torch.distributed.rpc import RRef
+from torch.nn.parallel import DistributedDataParallel as DDP
+
+NUM_EMBEDDINGS = 100
+EMBEDDING_DIM = 16
+
+# BEGIN hybrid_model
+class HybridModel(torch.nn.Module):
+    r"""
+    The model consists of a sparse part and a dense part. The dense part is an
+    nn.Linear module that is replicated across all trainers using
+    DistributedDataParallel. The sparse part is an nn.EmbeddingBag that is
+    stored on the parameter server.
+
+    The model holds a Remote Reference to the embedding table on the parameter
+    server.
+    """
+
+    def __init__(self, emb_rref, device):
+        super(HybridModel, self).__init__()
+        self.emb_rref = emb_rref
+        self.fc = DDP(torch.nn.Linear(16, 8).cuda(device), device_ids=[device])
+        self.device = device
+
+    def forward(self, indices, offsets):
+        emb_lookup = self.emb_rref.rpc_sync().forward(indices, offsets)
+        return self.fc(emb_lookup.cuda(self.device))
+# END hybrid_model
+
+# BEGIN setup_trainer
+def _retrieve_embedding_parameters(emb_rref):
+    param_rrefs = []
+    for param in emb_rref.local_value().parameters():
+        param_rrefs.append(RRef(param))
+    return param_rrefs
+
+
+def _run_trainer(emb_rref, rank):
+    r"""
+    Each trainer runs a forward pass which involves an embedding lookup on the
+    parameter server and running nn.Linear locally. During the backward pass,
+    DDP is responsible for aggregating the gradients for the dense part
+    (nn.Linear) and distributed autograd ensures gradients updates are
+    propagated to the parameter server.
+    """
+
+    # Setup the model.
+    model = HybridModel(emb_rref, rank)
+
+    # Retrieve all model parameters as rrefs for DistributedOptimizer.
+
+    # Retrieve parameters for embedding table.
+    model_parameter_rrefs = rpc.rpc_sync(
+            "ps", _retrieve_embedding_parameters, args=(emb_rref,))
+
+    # model.parameters() only includes local parameters.
+    for param in model.parameters():
+        model_parameter_rrefs.append(RRef(param))
+
+    # Setup distributed optimizer
+    opt = DistributedOptimizer(
+        optim.SGD,
+        model_parameter_rrefs,
+        lr=0.05,
+    )
+
+    criterion = torch.nn.CrossEntropyLoss()
+    # END setup_trainer
+
+    # BEGIN run_trainer
+    def get_next_batch(rank):
+        for _ in range(10):
+            num_indices = random.randint(20, 50)
+            indices = torch.LongTensor(num_indices).random_(0, NUM_EMBEDDINGS)
+
+            # Generate offsets.
+            offsets = []
+            start = 0
+            batch_size = 0
+            while start < num_indices:
+                offsets.append(start)
+                start += random.randint(1, 10)
+                batch_size += 1
+
+            offsets_tensor = torch.LongTensor(offsets)
+            target = torch.LongTensor(batch_size).random_(8).cuda(rank)
+            yield indices, offsets_tensor, target
+
+    # Train for 100 epochs
+    for epoch in range(100):
+        # create distributed autograd context
+        for indices, offsets, target in get_next_batch(rank):
+            with dist_autograd.context() as context_id:
+                output = model(indices, offsets)
+                loss = criterion(output, target)
+
+                # Run distributed backward pass
+                dist_autograd.backward(context_id, [loss])
+
+                # Tun distributed optimizer
+                opt.step(context_id)
+
+                # Not necessary to zero grads as each iteration creates a different
+                # distributed autograd context which hosts different grads
+        print("Training done for epoch {}".format(epoch))
+    # END run_trainer
+
+
+# BEGIN run_worker
+def run_worker(rank, world_size):
+    r"""
+    A wrapper function that initializes RPC, calls the function, and shuts down
+    RPC.
+    """
+    os.environ['MASTER_ADDR'] = 'localhost'
+    os.environ['MASTER_PORT'] = '29500'
+
+
+    rpc_backend_options = ProcessGroupRpcBackendOptions()
+    rpc_backend_options.init_method='tcp://localhost:29501'
+
+    # Rank 2 is master, 3 is ps and 0 and 1 are trainers.
+    if rank == 2:
+        rpc.init_rpc(
+                "master",
+                rank=rank,
+                world_size=world_size,
+                rpc_backend_options=rpc_backend_options)
+
+        # Build the embedding table on the ps.
+        emb_rref = rpc.remote(
+                "ps",
+                torch.nn.EmbeddingBag,
+                args=(NUM_EMBEDDINGS, EMBEDDING_DIM),
+                kwargs={"mode": "sum"})
+
+        # Run the training loop on trainers.
+        futs = []
+        for trainer_rank in [0, 1]:
+            trainer_name = "trainer{}".format(trainer_rank)
+            fut = rpc.rpc_async(
+                    trainer_name, _run_trainer, args=(emb_rref, rank))
+            futs.append(fut)
+
+        # Wait for all training to finish.
+        for fut in futs:
+            fut.wait()
+    elif rank <= 1:
+        # Initialize process group for Distributed DataParallel on trainers.
+        dist.init_process_group(
+                backend="gloo", rank=rank, world_size=2)
+
+        # Initialize RPC.
+        trainer_name = "trainer{}".format(rank)
+        rpc.init_rpc(
+                trainer_name,
+                rank=rank,
+                world_size=world_size,
+                rpc_backend_options=rpc_backend_options)
+
+        # Trainer just waits for RPCs from master.
+    else:
+        rpc.init_rpc(
+                "ps",
+                rank=rank,
+                world_size=world_size,
+                rpc_backend_options=rpc_backend_options)
+        # parameter server do nothing
+        pass
+
+    # block until all rpcs finish
+    rpc.shutdown()
+
+
+if __name__=="__main__":
+    # 2 trainers, 1 parameter server, 1 master.
+    world_size = 4
+    mp.spawn(run_worker, args=(world_size, ), nprocs=world_size, join=True)
+# END run_worker