Skip to content
Branch: master
Find file History
yf225 and facebook-github-bot Use c10::variant-based enums for Nonlinearity and FanMode
Summary: Pull Request resolved: #27933

Test Plan: Imported from OSS

Differential Revision: D18009044

Pulled By: yf225

fbshipit-source-id: e88229ee30badf7a699f62af61d1e88debc0dc7d
Latest commit eb4bb00 Oct 19, 2019
Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
autograd FunctionEventAvg implements __iadd__ interface (#27498) Oct 8, 2019
backends Remove fbgemm_is_cpu_supported in favor of torch.backends.quantized.s… Sep 27, 2019
contrib Remove `torch.contrib._graph_vis` (#24874) Aug 21, 2019
csrc Use c10::variant-based enums for Nonlinearity and FanMode Oct 19, 2019
cuda Bag of documentation fixes; fix more sphinx warnings (#27850) Oct 15, 2019
distributed fix python rpc handler exit crash (#27251) Oct 16, 2019
distributions Fix most documentation warnings (#27782) Oct 13, 2019
for_onnx Turn on F401: Unused import warning. (#18598) Mar 30, 2019
jit Add doc copy-edits from review (#26322) Oct 17, 2019
legacy Remove torch/legacy (#11823) Sep 20, 2018
lib improve error handling in getNCCLVersion in NCCLUtils (#27883) Oct 16, 2019
multiprocessing Rename torch.namedtensor -> torch._namedtensor_internals (#26349) Sep 18, 2019
nn Fix RNN nonlinearity (#28058) Oct 17, 2019
onnx Move Method::lowered_graph to a separate pass out of the Method class. ( Oct 18, 2019
optim Bag of documentation fixes; fix more sphinx warnings (#27850) Oct 15, 2019
quantization Add default arg to `prepare_qat` mapping. (#28193) Oct 17, 2019
sparse add some sparse tensor ops support in TorchScript (#24967) Aug 23, 2019
testing Fix get_all_math_dtypes for device='cuda' retuning None (#23028) Jul 19, 2019
utils adding IterableDataset to dataset.pyi (#27966) Oct 15, 2019
CMakeLists.txt Fix early expansion of CUDA_TOOLKIT_ROOT_DIR in libtorch builds (#27887) Oct 16, 2019
README.txt Make all of TH and THC C++. (#6913) Apr 28, 2018
__config__.py Allow a non-OpenMP based build (#19749) May 7, 2019
__future__.py Add `torch.__future__._overwrite_module_params_on_conversion` global … Jun 19, 2019
__init__.py Rename _intrinsic to intrinsic Oct 3, 2019
__init__.pyi.in Add data field to Tensor pyi. (#26093) Sep 13, 2019
_classes.py Initial torchbind prototype (#21098) Aug 3, 2019
_jit_internal.py Add doc copy-edits from review (#26322) Oct 17, 2019
_namedtensor_internals.py Allow `align_to` to take in partially named tensors (#27308) Oct 9, 2019
_ops.py Initial torchbind prototype (#21098) Aug 3, 2019
_six.py module dedupe (#26666) Oct 12, 2019
_storage_docs.py Enabled BFloat16 storage (#21523) Jul 10, 2019
_tensor_docs.py Add Polygamma to the docs (#27696) Oct 15, 2019
_tensor_str.py Per-channel quantized tensor to have only a single axis (#26675) Sep 24, 2019
_torch_docs.py cdist performance improvement for euclidean distance (#25799) Oct 17, 2019
_utils.py Implement pickle support for sparse tensors and torch.layout instances ( Oct 4, 2019
_utils_internal.py Add a wrapper for inspect in JIT to produce better error message (#25415 Sep 15, 2019
abi-check.cpp Fixes for Torch Script C++ API (#11682) Sep 17, 2018
custom_class.h search class type for methods (#23689) Aug 13, 2019
extension.h Remove deprecated variable_tensor_functions (#15003) Dec 12, 2018
functional.py cdist performance improvement for euclidean distance (#25799) Oct 17, 2019
hub.py Fix hub when branch name contains slash. (#27960) Oct 18, 2019
py.typed More type stubs (#18511) Apr 1, 2019
quasirandom.py Make SobolEngine use random seed if not specified (#24884) Aug 20, 2019
random.py Refactor Random Number Generators in ATen (#21555) Jun 19, 2019
script.h turn off autograd mode in android JNI wrapper (#26477) Sep 20, 2019
serialization.py Implement pickle support for sparse tensors and torch.layout instances ( Oct 4, 2019
storage.py Enabled BFloat16 storage (#21523) Jul 10, 2019
tensor.py Allow `align_to` to take in partially named tensors (#27308) Oct 9, 2019

README.txt

Note [TH abstraction violation]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

TH/THC provide some hpp headers, which are proper C++ headers rather than
C headers.  These headers serve double duty as *internal implementation
detail* headers, whose contents should largely not be used by external
clients.

Ideally, we would not install these headers at all; instead, you should
use public functions (in headers like `THTensor.h`, NOT `THTensor.hpp`)
to manipulate these structs.  However, there are a few places
in torch/csrc where we violate this abstraction.  They are marked with
a pointer to this note.  Each of those sites will have to be refactored
when we refactor the guts of THTensor and related structures.
You can’t perform that action at this time.