strisunshinewentingwang/pytorch

Pull Requests Commits

[torch.distributions] Implement positive-semidefinite constraint (#71375)

nonconvexopt committed 2 years ago

89c844db

Refactor convolution_backward's cudnn cases (#71491)

zou3519 committed 2 years ago

640bfa7e

Refactor convolution_backward's CudaDepthwise3d case (#71490)

zou3519 committed 2 years ago

06f14c2d

Refactor convolution_backward's CudaDepthwise2d case (#71489)

zou3519 committed 2 years ago

17d2a516

[BE] Improve gitutils

malfet committed 2 years ago

42f7afc4

Fix composite compliance problems for linalg.{matrix_power, inv, cholesky} (#69437)

zou3519 committed 2 years ago

a9f44b22

[DataPipe] improving DataPipe unit tests (#70215)

NivekT committed 2 years ago

011fd1d9

[AutoAccept][Codemod][FBSourceClangFormatLinter] Daily `arc lint --take CLANGFORMAT`

CodemodService FBSourceClangFormatLinterBot committed 2 years ago

9f0c8085

fix: do_constant_folding arg when exporting ONNX (#71348)

hoangtnm committed 2 years ago

06838ce8

add flatbuffer_loader and flatbuffer_serializer as BUCK target (#71463)

qihqi committed 2 years ago

21b697b6

Add silu and hardsigmoid converter (#71453)

Shirong Wu committed 2 years ago

99df96d8

Enable Python bindings for UntypedStorage (#68945)

cbalioglu committed 2 years ago

80b19c4c

Additional unit test for sharded linear. (#70476)

pritamdamania committed 2 years ago

f5b19ba6

Add GitHub merge rules (#71514)

malfet committed 2 years ago

a5d5b112

[PyTorch] Fix a bunch of structured kernel refcounting (#71140)

swolchok committed 2 years ago

c59942ac

[flatbuffer] Fix forward flatbuffer type handling with dynamic type. (#71500)

zhxchen17 committed 2 years ago

b98e955b

[Pytorch] Speed up LayerNorm 4-5% (#71423)

swolchok committed 2 years ago

565f78f5

[PyTorch][Static Runtime] Fix extra refcount bumps in layer_norm (#71237)

swolchok committed 2 years ago

958f9cf5

Fix trivial typo at the doc of `torch.lobpcg` (#71464)

nonconvexopt committed 2 years ago

811af259

Update min python version to 3.7 in setup.py and mypy configs (#71494)

malfet committed 2 years ago

dc5cda0c

[acc_ops] Remove usage of kwarg expansion via **locals() for jit scripting support (#71425)

jfix71 committed 2 years ago

06bc6748

[distributed] Make rref_proxy._invoke_rpc trully async when needed. (#70206)

Rodrigo Kumpera committed 2 years ago

ef4bc3fa

[nnc] Update block and thread extents in cuda_codegen to use int64_t (#71428)

navahgar committed 2 years ago

70c9146c

[fx2trt] Issue warnings instead of error if there's possible const folding opportunities (#71031)

842974287 committed 2 years ago

2dbbb1a9

Add trymerge workflow (#71488)

malfet committed 2 years ago

61713acb

Consolidate the overloads of TensorImpl::shallow_copy_and_detach (#68953)

cbalioglu committed 2 years ago

f45e217c

test //c10/... without Google libraries in OSS (#70853)

Michael Dagitses committed 2 years ago

805b7575

port //c10/macros to common build structure (#70852)

Michael Dagitses committed 2 years ago

78e1f9db

use c10/macros/cmake_macros.h in fbcode build (#70851)

Michael Dagitses committed 2 years ago

661d10aa

[fx] add documentation to AccOpProperties (#71450)

alexbeloi committed 2 years ago

bdeec0c7

Newer Older