Protobuf serialization (#11619)

Summary:
This PR serves two purposes:

1. Design an abstraction over a serialization scheme for C++ modules, optimizers and tensors in general,
2. Add serialization to the ONNX/PyTorch proto format.

This is currently a rough prototype I coded up today, to get quick feedback.

For this I propose the following serialization interface within the C++ API:

```cpp
namespace torch { namespace serialize {
class Reader {
 public:
  virtual ~Reader() = default;
  virtual void read(const std::string& key, Tensor& tensor, bool is_buffer = false) = 0;
  virtual void finish() { }
};

class Writer {
 public:
  virtual ~Reader() = default;
  virtual void writer(const std::string& key, const Tensor& tensor, bool is_buffer = false) = 0;
  virtual void finish() { }
};
}} // namespace torch::serialize
```

There are then subclasses of these two for (1) Cereal and (2) Protobuf (called the "DefaultWriter" and "DefaultReader" to hide the implementation details). See `torch/serialize/cereal.h` and `torch/serialize/default.h`. This abstraction and subclassing for these two allows us to:

1. Provide a cereal-less serialization forward that we can ship and iterate on going forward,
2. Provide no-friction backwards compatibility with existing C++ API uses, mainly StarCraft.

The user-facing API is (conceptually):

```cpp
void torch::save(const Module& module, Writer& writer);
void torch::save(const Optimizer& optimizer, Writer& writer);
void torch::read(Module& module, Reader& reader);
void torch::read(Optimizer& optimizer, Reader& reader);
```

with implementations for both optimizers and modules that write into the `Writer` and read from the `Reader`

ebetica ezyang zdevito dzhulgakov
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11619

Differential Revision: D9984664

Pulled By: goldsborough

fbshipit-source-id: e03afaa646221546e7f93bb8dfe3558e384a5847

2 files changed, 0 insertions, 8 deletions

diff --git a/tools/build_libtorch.py b/tools/build_libtorch.py
index db698a2412..1f7d709d31 100644
--- a/tools/build_libtorch.py
+++ b/tools/build_libtorch.py
@@ -9,7 +9,6 @@ from setup_helpers.cuda import USE_CUDA
 if __name__ == '__main__':
     # Placeholder for future interface. For now just gives a nice -h.
     parser = argparse.ArgumentParser(description='Build libtorch')
-    parser.add_argument('--use-cereal', action='store_true')
     options = parser.parse_args()
 
     os.environ['BUILD_TORCH'] = 'ON'
@@ -25,8 +24,6 @@ if __name__ == '__main__':
         command.append('--use-cuda')
         if os.environ.get('USE_CUDA_STATIC_LINK', False):
             command.append('--cuda-static-link')
-    if options.use_cereal:
-        command.append('--use-cereal')
     command.append('caffe2')
 
     sys.stdout.flush()
diff --git a/tools/build_pytorch_libs.sh b/tools/build_pytorch_libs.sh
index 37d816775f..01cb82f49c 100755
--- a/tools/build_pytorch_libs.sh
+++ b/tools/build_pytorch_libs.sh
@@ -22,7 +22,6 @@ USE_NNPACK=0
 USE_MKLDNN=0
 USE_GLOO_IBVERBS=0
 CAFFE2_STATIC_LINK_CUDA=0
-TORCH_USE_CEREAL=0
 RERUN_CMAKE=1
 while [[ $# -gt 0 ]]; do
     case "$1" in
@@ -47,9 +46,6 @@ while [[ $# -gt 0 ]]; do
       --cuda-static-link)
           CAFFE2_STATIC_LINK_CUDA=1
           ;;
-      --use-cereal)
-          TORCH_USE_CEREAL=1
-          ;;
       *)
           break
           ;;
@@ -194,7 +190,6 @@ function build() {
 		       -DTHCUNN_SO_VERSION=1 \
 		       -DTHD_SO_VERSION=1 \
 		       -DUSE_CUDA=$USE_CUDA \
-		       -DTORCH_USE_CEREAL=$TORCH_USE_CEREAL \
 		       -DBUILD_EXAMPLES=OFF \
 		       -DBUILD_TEST=$BUILD_TEST \
 		       -DNO_NNPACK=$((1-$USE_NNPACK)) \