Summary:
Currently PyTorch forces color output due to #20662. But users should be left an option to turn it off because redirection of the output to a file would be messed if color output is forced.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/20771
Differential Revision: D15495677
Pulled By: ezyang
fbshipit-source-id: 9d89bbed40d0b67368554305394763a54c5ff6f5
Summary:
Let there be color!
Pull Request resolved: https://github.com/pytorch/pytorch/pull/20662
Differential Revision: D15434110
Pulled By: suo
fbshipit-source-id: a317ae72ad72e0b8249f55c9c8d31f420c78c040
Summary:
Fixes#20568.
Looks like CMake is passing `/MD` when we call `add_library`. We need to fix these with C source files too.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/20574
Differential Revision: D15392682
Pulled By: ezyang
fbshipit-source-id: c92034d8725fcec48fd7db6cf5322868e956dc6b
Summary:
First time this was merged it broke master and was reverted. This time I do not add ```set -u``` to the .circleci/scripts/setup* scripts. There's still a chance that ```set -u``` breaks the binary builds on master, but at least those can be fixed in parallel and don't completely eliminate signal from all merges.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/20540
Differential Revision: D15373444
Pulled By: pjh5
fbshipit-source-id: 0203c20865827366ecd8fa07b2db74d255549ed1
Summary:
This PR is an intermediate step toward the ultimate goal of eliminating "caffe2" in favor of "torch". This PR moves all of the files that had constituted "libtorch.so" into the "libcaffe2.so" library, and wraps "libcaffe2.so" with a shell library named "libtorch.so". This means that, for now, `caffe2/CMakeLists.txt` becomes a lot bigger, and `torch/CMakeLists.txt` becomes smaller.
The torch Python bindings (`torch_python.so`) still remain in `torch/CMakeLists.txt`.
The follow-up to this PR will rename references to `caffe2` to `torch`, and flatten the shell into one library.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/17783
Differential Revision: D15284178
Pulled By: kostmo
fbshipit-source-id: a08387d735ae20652527ced4e69fd75b8ff88b05
Summary:
MSVC errors on these flags as they are not supported
Pull Request resolved: https://github.com/pytorch/pytorch/pull/18686
Differential Revision: D14704254
Pulled By: ezyang
fbshipit-source-id: 936d33ed6b7474d7774a49505cdac50dbe8dd99a
Summary:
FBGEMM doesn't work on x86 32bit and prior to this patch, it will
generate x86_64 objects in a build that is supposed to be x86 32bit.
FBGEMM actually relies on registers not available on x86_32, so
we disable it.
This takes of one element of #17901. There are more dependencies
and a separate PR (#17915) regarding AVX detection for the code in the
main repository.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/17922
Differential Revision: D14437340
Pulled By: soumith
fbshipit-source-id: bd9fc98cf607d9b0bc28127fbbc8b04fa10eecbe
Summary:
This PR implements infrastructure for post-processing a model to apply int8 quantization to its `nn.Linear` modules. Highlights of the implementation:
1) Inputs and outputs are `float` (quantized and packed internally), but the weight is quantized and packed ahead of time for efficiency. This implementation performs well in small-batch size GEMM calls. It should not be considered a general-purpose quantized GEMM kernel.
2) Weight packing is dependent on machine architecture (e.g. vector register width), so it is done just-in-time. Concretely, it is done on model load for the weights and it is done during operator execution for the input value.
3) Biases are unquantized
4) We fail loudly if we are attempting to run this on a machine that does not support FBGEMM. This is because we do not want a model's numerics to differ based on which machine it is run on. A model containing these FBGEMM ops *must* be run with FBGEMM
The API can be seen in the added test case. Highlights are:
1) `torch.jit.quantized.quantize_linear_modules` walks the module hierarchy of the passed-in Module and replaces all `nn.Linear` modules with a new `QuantizedLinear` module, which encapsulates the behavior described above.
2) `_pack()` and `_unpack()` script methods are present on `QuantizedLinear` modules. These methods should be called before serialization and after deserialization, respectively. This ensures that the weight matrix is properly packed for the running machine's architecture. Note that in the long term, we would like to move toward a more Pickle-style serialization technique, rather than having these explicit methods that mutate member values. This is blocked on being able to assign attributes in a ScriptMethod, among other things.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13777
Differential Revision: D13383276
Pulled By: jamesr66a
fbshipit-source-id: 00f29c9f34544add2b90107e3cf55a287802c344
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/15322
caffe2 mobile opengl code is not used, deleting it to reduce complications when we perform other changes
Reviewed By: Maratyszcza
Differential Revision: D13499943
fbshipit-source-id: 6479f6b9f50f08b5ae28f8f0bc4a1c4fc3f3c3c2
Summary:
This is broken out of https://github.com/pytorch/pytorch/pull/13733/
We want to install cpp tests so they can ultimately be runnable from that location for Caffe2 tests run from PyTorch builds.
cc pjh5 yf225 anderspapitto
Pull Request resolved: https://github.com/pytorch/pytorch/pull/15000
Reviewed By: pjh5
Differential Revision: D13416253
Pulled By: orionr
fbshipit-source-id: 51280be0a22557a742f90c9f303c58c35cbd4a38
Summary:
Previously if it unconditionally tries to load rocm cmake files, so there was no way to disable rocm build. After this change, USE_ROCM=0 will disable rocm build.
Should fix#14025
soumith
Pull Request resolved: https://github.com/pytorch/pytorch/pull/14261
Differential Revision: D13242090
Pulled By: bddppq
fbshipit-source-id: 652ec7d49dce9b357778bfa53a8e04b7079787ab
Summary:
Hi guys,
I'd like to build Caffe2 with more supported options in Windows with Microsoft Visual Studios.
This is the first pull request.
Running scripts/build_windows_shared.bat is able to build Caffe2 with both CMAKE_BUILD_TYPE=Debug and CMAKE_BUILD_TYPE=Release with Visual Studio 14 2015.
CUDA is 9.0, cudnn is 7.0.5, glog, gflags and lmdb are supported on my system.
Python is 3.5, Detectron works from python interface as well.
It was even possible to debug detectron code and step into caffe2_gpu.dll with pdbs built.
What is disappointing, that c10/experimental ops don't build with this Visual Studio generator, I added special option INCLUDE_EXPERIMENTAL_C10_OPS (default ON) to deal with it in build_windows_shared.bat.
After this pull request the next step is to add Visual Studio 2017 support in the script.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13550
Reviewed By: ezyang
Differential Revision: D13042597
Pulled By: orionr
fbshipit-source-id: f313f909f599cd582a1d000eff766eef3a9fc4fc
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/14048
Setting USE_FBGEMM to OFF by default until we figure out properly separating avx2 code. See [this issue](https://github.com/pytorch/pytorch/issues/13993). Pytorch can still be compiled with fbgemm by using USE_FBGEMM=ON.
Reviewed By: jspark1105
Differential Revision: D13090454
fbshipit-source-id: 6e0e92612e4362a306e376df3dc33e8edeb066e9
Summary:
Following up #11488 conversation with orionr
And our brief conversation at PTDC about ATen with soumith and apaszke
This PR enables a very slim build focused on ATen particularly without caffe2 and protobuf among other dependencies.
WIth this PR NimTorch tests pass fully, including AD, convolutions, wasm, etc.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12443
Reviewed By: mingzhe09088
Differential Revision: D10249313
Pulled By: orionr
fbshipit-source-id: 4f50503f08b79f59e7717fca2b4a1f420d908707
Summary:
this removes a bunch of spam output from the build. This is
(1) cleaner
(2) a couple seconds faster in some cases, e.g. my slow-rendering emacs-based shell
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12392
Differential Revision: D10225340
Pulled By: anderspapitto
fbshipit-source-id: 477ee76d24f8db50084b1e261db8c22733de923b
Summary:
All usages of the `ndarray` construct have now been guarded with `USE_NUMPY`. This eliminates the requirement of NumPy while building PyTorch from source.
Fixes#11757
Reviewed By: Yangqing
Differential Revision: D10031862
Pulled By: SsnL
fbshipit-source-id: 32d84fd770a7714d544e2ca1895a3d7c75b3d712
Summary:
At long last, we will have clang-tidy enabled in CI. For a while I thought I could clean up the project enough to enable clang-tidy with all checks enabled, but I figure it's smarter to set up the minimal checks and at least have those in CI. We can fix more going forward.
ezyang apaszke
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12213
Differential Revision: D10183069
Pulled By: goldsborough
fbshipit-source-id: 7ecd2d368258f46efe23a2449c0a206d10f3a769
Summary:
This unifies our versions across setup.py, libtorch, and libcaffe2. CMake has a default version (bumped to 1.0.0) that can be overridden by setup.py. The versions are also printed as a part of cmake/Summary.cmake to make sure they are correct.
cc Yangqing ezyang soumith goldsborough pjh5
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12053
Differential Revision: D10041878
Pulled By: orionr
fbshipit-source-id: a98a01771f6c008d1016ab63ab785c3a88c3ddb0
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
Summary:
I am working on unifying the C++ extensions and C++ API, and one constraint for this is that we will want to be able to build the C++ API without cereal, since we won't want to ship it with the Python `torch` package.
For this I introduce a `TORCH_WITH_CEREAL` option to CMake. If on, the C++ API will be built with cereal and thus serialization support. If off, serialization functions will throw exceptions, but the library will otherwise still compile the same. __This option is on by default, so for regular C++ API users nothing will change__. However, from C++ extensions, we'll be able to turn it off. This effectively means we won't be searching for any cereal headers from C++ API headers, which wouldn't be installed in the Python package.
ebetica ezyang soumith
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11498
Differential Revision: D9784803
Pulled By: goldsborough
fbshipit-source-id: 5d0a1f2501993012d28cf3d730f45932b483abc4
Summary:
The previous error was caused by mpi_test not depending on MPI_CXX_LIBRARIES. This might solve the problem.
Not tested locally - waiting for CI test.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11416
Reviewed By: mingzhe09088
Differential Revision: D9771694
Pulled By: Yangqing
fbshipit-source-id: 53e7b4f64eadc88313bc4dd9b8e3f7931cda6e91
Summary:
Add flags for LMDB and LevelDB, default `OFF`. These can be enabled with
```
USE_LMDB=1 USE_LEVELDB=1 python setup.py build_deps
```
Also add a flag to build Caffe2 ops, which is default `ON`. Disable with
```
NO_CAFFE2_OPS=1 python setup.py build_deps
```
cc Yangqing soumith pjh5 mingzhe09088
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11462
Reviewed By: soumith
Differential Revision: D9758156
Pulled By: orionr
fbshipit-source-id: 95fd206d72fdf44df54fc5d0aeab598bff900c63
Summary:
Continuing pjh5's work to remove FULL_CAFFE2 flag completely.
With these changes you'll be able to also do something like
```
NO_TEST=1 python setup.py build_deps
```
and this will skip building tests in caffe2, aten, and c10d. By default the tests are built.
cc mingzhe09088 Yangqing
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11321
Reviewed By: mingzhe09088
Differential Revision: D9694950
Pulled By: orionr
fbshipit-source-id: ff5c4937a23d1a263378a196a5eda0cba98af0a8
Summary:
This is an experimental build on top of what orionr and mingzhe09088 built.
Essentially, the idea is that we will need separate *_API versions for different shared libraries. If this theory is right, I'll try to clean up the design a bit and document it properly.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11266
Reviewed By: orionr
Differential Revision: D9682942
Pulled By: Yangqing
fbshipit-source-id: c79653199e67a1500c9174f39f8b0357324763f3
Summary:
this is a fix that's needed for building extensions with a
pre-packaged pytorch. Consider the scenario where
(1) pytorch is compiled and packaged on machine A
(2) the package is downloaded and installed on machine B
(3) an extension is compiled on machine B, using the downloaded package
Before this patch, stage (1) would embed absolute paths to the system
installation of mkl into the generated Caffe2Config.cmake, leading to
failures in stage (3) if mkl was not at the same location on B as on
A. After this patch, only a reference to the wrapper library is
embedded, which is re-resolved on machine B.
We are already using a similar approach for cuda.
Testing: built a package on jenkins, downloaded locally and compiled an extension. Works with this patch, fails without.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11298
Differential Revision: D9683150
Pulled By: anderspapitto
fbshipit-source-id: 06a80c3cd2966860ce04f76143b358de15f94aa4
Summary:
Now that we're building everything together, making all distributed flags conditional of USE_DISTRIBUTED being set.
cc pietern The controller you requested could not be found. cpuhrsch
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11221
Reviewed By: Yangqing
Differential Revision: D9664267
Pulled By: orionr
fbshipit-source-id: a296cda5746ad150028c97160f8beacba955ff73
Summary:
TSIA. apaszke pointed out that it might be better to use third party folder in default, since system Eigen may often be out of date and does not have the version we need to compile successfully.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11020
Differential Revision: D9562548
Pulled By: Yangqing
fbshipit-source-id: d8ab8a6ebe1f3d9eec638ef726cf5dc4dcf777b5
Summary:
This completely removes BUILD_CAFFE2 from CMake. There is still a little bit of "full build" stuff in setup.py that enables USE_CUDNN and BUILD_PYTHON, but otherwise everything should be enabled for PyTorch as well as Caffe2. This gets us a lot closer to full unification.
cc mingzhe09088, pjh5, ezyang, smessmer, Yangqing
Pull Request resolved: https://github.com/pytorch/pytorch/pull/8338
Reviewed By: mingzhe09088
Differential Revision: D9600513
Pulled By: orionr
fbshipit-source-id: 9f6ca49df35b920d3439dcec56e7b26ad4768b7d
