pytorch/torch
David Riazati ac00a0cd47 Dict mutability (#16884)
Summary:
Adds `aten::_set_item` for `dict[key]` calls
Pull Request resolved: https://github.com/pytorch/pytorch/pull/16884

Differential Revision: D14000488

Pulled By: driazati

fbshipit-source-id: ea1b46e0a736d095053effb4bc52753f696617b2
2019-02-21 16:24:17 -08:00
..
_thnn try to get rid of tmp_install (#16414) 2019-01-29 17:29:40 -08:00
autograd Return namedtuples from torch.* function with multiple return arguments for C++ operators (#15429) 2019-01-22 11:12:18 -08:00
backends Add torch.backends.openmp.is_available(); fix some cmake messages (#16425) 2019-01-31 16:15:46 -08:00
contrib Remove stages from IR, they are not longer used 2018-10-05 13:58:15 -07:00
csrc Dict mutability (#16884) 2019-02-21 16:24:17 -08:00
cuda Bool tensor. Part 0: Boolean storage implementation (#16810) 2019-02-19 08:22:13 -08:00
distributed Pass torch.distributed launch process local rank as environment variable instead of argument (#16360) 2019-02-15 14:52:55 -08:00
distributions Fix the slowness of mvn's log_prob (#17294) 2019-02-21 08:04:33 -08:00
for_onnx
jit Partial support for kwarg_only arguments in script (#17339) 2019-02-21 15:27:06 -08:00
legacy Remove torch/legacy (#11823) 2018-09-20 14:00:54 -07:00
lib Trim libshm deps, move tempfile.h to c10 (#17019) 2019-02-13 19:38:35 -08:00
multiprocessing Move all Stream and Event Python implementation to C++ (#15937) 2019-01-17 07:29:22 -08:00
nn Fix #17218 by updating documentation (#17258) 2019-02-21 14:17:35 -08:00
onnx ONNX Export Maxpool Indices 2019-02-19 21:10:14 -08:00
optim Redefine scheduler to set learning rate using recursive formula (#14010) 2018-12-18 16:44:31 -08:00
sparse Correct conv and pooling docstrings in nn module (#17052) 2019-02-15 06:58:02 -08:00
testing Lightweight String check Utility (#16858) 2019-02-19 12:31:57 -08:00
utils fix behavior of ConcatDataset w/ negative indices (#15756) 2019-02-14 13:02:54 -08:00
__init__.py Fix dll loading process in newer Python on Windows (#17191) 2019-02-19 17:16:41 -08:00
__init__.pyi.in Move argsort to C++ 2019-02-21 07:59:27 -08:00
_jit_internal.py Add module and name to func created with _jit_internal.boolean_dispatch (#16922) 2019-02-12 09:38:48 -08:00
_ops.py Use realpath for loaded libraries (#13936) 2018-11-15 11:23:20 -08:00
_six.py create type hint stub files for module torch (#12500) 2019-01-29 12:14:17 -08:00
_storage_docs.py Bool tensor. Part 0: Boolean storage implementation (#16810) 2019-02-19 08:22:13 -08:00
_tensor_docs.py Move argsort to C++ 2019-02-21 07:59:27 -08:00
_tensor_str.py Added scientific notation on set_printoptions (#16876) 2019-02-11 04:55:12 -08:00
_torch_docs.py Move argsort to C++ 2019-02-21 07:59:27 -08:00
_utils_internal.py Use fixed MASTER_PORT in test_distributed (#13109) 2018-10-25 08:51:34 -07:00
_utils.py create type hint stub files for module torch (#12500) 2019-01-29 12:14:17 -08:00
abi-check.cpp Fixes for Torch Script C++ API (#11682) 2018-09-17 09:54:50 -07:00
CMakeLists.txt Lightweight String check Utility (#16858) 2019-02-19 12:31:57 -08:00
extension.h Remove deprecated variable_tensor_functions (#15003) 2018-12-11 17:16:11 -08:00
functional.py Move argsort to C++ 2019-02-21 07:59:27 -08:00
hub.py Fix github branch prefix v (#15552) 2018-12-26 19:48:47 -08:00
random.py Improve the docstring of nn.random.fork_rng (#15960) 2019-01-14 02:41:18 -08:00
README.txt
script.h Use torch:: instead of at:: in all C++ APIs (#13523) 2018-11-06 14:32:25 -08:00
serialization.py Avoid unnecessary CPU-to-GPU copy of torch.load with CUDA (#17297) 2019-02-21 01:32:19 -08:00
storage.py Bool tensor. Part 0: Boolean storage implementation (#16810) 2019-02-19 08:22:13 -08:00
tensor.py Move argsort to C++ 2019-02-21 07:59:27 -08:00

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.