strisunshinewentingwang/pytorch

Pull Requests Commits

Automated submodule update: FBGEMM (#46395)

facebook-github-bot committed 3 years ago

e69f2f82

Added support for complex torch.pinverse (#45819)

IvanYashchuk committed 3 years ago

c1141b6f

BFloat16 support for torch.sign (#45244)

zasdfgbnm committed 3 years ago

5ce46fbb

Add fence. (#45148)

Ashkan Aliabadi committed 3 years ago

8c26111a

Add image sampler. (#45037)

Ashkan Aliabadi committed 3 years ago

e3eef0cd

Redo Vulkan command and descriptor pools. (#44496)

Ashkan Aliabadi committed 3 years ago

50f83324

Fix error message for scatter reduction (#46397)

albanD committed 3 years ago

1e654a4b

Updated derivatives for complex mm, mv, ger, bmm, triangular_solve (#45737)

IvanYashchuk committed 3 years ago

528158af

Allow Undefined to get kernel from Math/DefaultBackend. (#46352)

Ailing Zhang committed 3 years ago

7f458e16

[JIT] Revert Freezing shared type PR (#46285)

Elias Ellison committed 3 years ago

908c2357

Add windows JNI support (#44257)

Elijah Rippeth committed 3 years ago

b5479737

Fix formatting issues in torch.tensor_split documentation (#46328)

kurtamohler committed 3 years ago

bd449334

Stop running clang-tidy on torch/csrc/generic/*.cpp. (#46335)

gchanan committed 3 years ago

75809626

Fix incorrect signatures in get_testing_overrides, and add test for incorrect signatures (#45983)

hameerabbasi committed 3 years ago

e366591d

Rationalize inlining of kernels into the unboxing wrapper (#42845)

smessmer committed 3 years ago

2d6fd22e

Update COMPILE_TIME_MAX_DEVICE_TYPES to 12 (#46327)

xta0 committed 3 years ago

053c252c

[shape inference] add shape inference support

Zeliang Chen committed 3 years ago

38c97fb6

Fix perfornance issue of GroupNorm on CUDA when feature map is small. (#46170)

xiaomengy committed 3 years ago

a87a1c11

[JIT] Improve class type annotation inference (#45940)

Meghan Lele committed 3 years ago

75bf5f2b

[JIT] Make InsertInstruction overflow check a warning instead of fatal (#46369)

gmagogsfm committed 3 years ago

86abc8cd

Add guideline about which dispatch keyword to use in native_functions.yaml. (#46126)

Ailing Zhang committed 3 years ago

5393588e

Bug fixes in profiling allocator (#45993)

kimishpatel committed 3 years ago

4aaad887

[Reland] Update native_functions.yaml to add DefaultBackend. (#46236)

Ailing Zhang committed 3 years ago

419dafe7

[pytorch] activation checkpointing: enable mixing tensor without requires_grad (#45934)

Weiyi Zheng committed 3 years ago

22f4a58a

Bazel build has warnings (#46233)

guol-fnst committed 3 years ago

103b100d

[PyTorch] Stringize kernel tag names consistently during macro expansion and require all tag names to be a compile time character array (#46074)

dhruvbird committed 3 years ago

af8c75e2

Fix possible padding length overflow in DistributedSampler (#45329)

agemor committed 3 years ago

a6991086

Revert D24290811: [quant][eagermode] Move custom_module registration to prepare/convert_custom_config_dict

Mike Ruberry committed 3 years ago

ff0af724

Make setup.py python 2 friendly (#46317)

malfet committed 3 years ago

a38eeeff

Add warning on ProcessGroup and ProcessGroup::Work APIs (#46220)

agolynski committed 3 years ago

e7e919fc

Newer Older