strisunshinewentingwang/pytorch

Pull Requests Commits

Add ONNX export support for torch.log1p. (#25808)

Spandan Tiwari committed 5 years ago

fc93d1ae

add torch.jit.is_scripting api (#25955)

Elias Ellison committed 5 years ago

1897440e

Remove some more BUILD_NAMEDTENSOR flags

zou3519 committed 5 years ago

5d3267cd

Disable flaky test_invalid_names in test_rpc.py (#25916)

mrshenli committed 5 years ago

2655b271

Add names= argument to torch.tensor ctor (#25424)

zou3519 committed 5 years ago

42312875

make python rpc handler to be singleton class (#25742)

zhaojuanmao committed 5 years ago

2856fd6c

update build_android.sh to not build host protoc for libtorch (#25896)

ljk53 committed 5 years ago

16c19078

remove pthreadpool dependency in aten/CMake (#25894)

ljk53 committed 5 years ago

4bd9ddb0

Fix int32 overflow in SummaryOps.cu getBin #25747 (#25748)

kaczorrrro committed 5 years ago

ec8e75ea

add set_grad_enabled to TorchScript and fix data attribute

wanchaol committed 5 years ago

a7eaec6c

Add ONNX Export Support to rsqrt

lara-hdr committed 5 years ago

387d5a44

Make persistent softmax WARP_SIZE aware. (#25937)

iotamudelta committed 5 years ago

d377556f

Migrate pow from TH to Aten (CUDA) (#25517)

pbelevich committed 5 years ago

a14e8845

Only create a new clone of observer when we actually insert it.

Mikhail Zolotukhin committed 5 years ago

55219d55

Make lookup table warp size aware

iotamudelta committed 5 years ago

618804f2

C++ Fold nn module

ShahriarSS committed 5 years ago

3680cef4

Make spatial depthwise convolution warp size aware (#25922)

iotamudelta committed 5 years ago

2ab0f221

Make arguments of Module::dump easier to remember. (#25740)

Mikhail Zolotukhin committed 5 years ago

c749be9e

fix scatter CPU kernel when (input size, src size) > index size (#25839)

Ailing Zhang committed 5 years ago

26f67e7a

make sparse coalesce warp size aware (#25918)

iotamudelta committed 5 years ago

5dfef472

Add Dropout to blacklist (#25881)

Haixin Liu committed 5 years ago

9c10f729

Enable libflame as a LAPACK choice (#25795)

iotamudelta committed 5 years ago

26675b50

Tensor type set (#25308)

ezyang committed 5 years ago

aa49aa85

add NO_EXPORT macro to unset __visibility__ attribute (#25816)

ljk53 committed 5 years ago

6630c3f3

introduce INTERN_DISABLE_AUTOGRAD flag to create inference only library for mobile

ljk53 committed 5 years ago

84857101

gate static aten registerer with USE_STATIC_DISPATCH (#25815)

ljk53 committed 5 years ago

41cf5564

Rename packed tensor accessor (#25654)

peterbell10 committed 5 years ago

76ee02f1

Fix cpp_extensions test failures with GCC 9.1 from ArrayRef(initializer_list) (#25384)

rgommers committed 5 years ago

e8cc1fdd

Store bias in PackedConvWeight in fbgemm (#25626)

supriyar committed 5 years ago

c60dddbb

In the CUDA implementation of erfinv, erfinv() should be used for double (#25337)

xuhdev committed 5 years ago

57b23c61

Newer Older