strisunshinewentingwang/pytorch

Pull Requests Commits

Polish torch::Dict (#23344)

smessmer committed 5 years ago

b9202d45

Align String str() with the format in FunctionSchema (#23201)

houseroad committed 5 years ago

722f80a0

Remove superfluous check (#23370)

pietern committed 5 years ago

7c383ba4

Fix a tiny bug and refactor (#22808)

jianyuh committed 5 years ago

39de49c7

pjh5 committed 5 years ago

39fd2647

Align Dict str() with the format in FunctionSchema (#23200)

houseroad committed 5 years ago

ed316c0c

Add type checks in _intrinsics.modules.fused (#23361)

jerryzh168 committed 5 years ago

f477cab2

Support parsing variadic return schema (#23199)

houseroad committed 5 years ago

25e06618

Disable fusion of grad_sum_to_size (#23372)

t-vi committed 5 years ago

cf50249b

Specify build dir as a global variable in BUILD_DIR in the build system.

xuhdev committed 5 years ago

82545ecc

Let users pass CMake-specific options starting with CMAKE_ to CMake. (#22776)

xuhdev committed 5 years ago

915261c8

Do not explicitly set USE_FBGEMM in tools/setup_helpers/cmake.py (#23314)

xuhdev committed 5 years ago

f91b19c2

Add document of functions nn.init.ones_/zeros_ (#23145)

kianasun committed 5 years ago

ba6f65cf

jjsjann123 committed 5 years ago

25271026

Improve documentation for torch.enable_grad , torch.no_grad and torch.set_grad_enabled (#23310)

prasunanand committed 5 years ago

0c79753c

Fix lint failure introduced in #23346

pietern committed 5 years ago

2938299d

Fix upload issue with linux libtorch nightlies (#23368)

rgommers committed 5 years ago

0842624d

Enhance interpretation of GLOO_SOCKET_IFNAME (#22978)

pietern committed 5 years ago

95e82262

Improve FindMKLDNN.cmake to avoid binary compatibility issue in MKL-DNN (#23292)

gujinghui committed 5 years ago

1dd4d555

Remove usage of FOLDER constant in test_distributed.py (#23223)

pietern committed 5 years ago

e8ad1672

Make optimize a thread_local flag

suo committed 5 years ago

711be829

Replace uint8 with int8 in Linear and LSTM quantization path (#23347)

mingzhe09088 committed 5 years ago

b3980f46

Try kill -9ing apt-get (#23354)

ezyang committed 5 years ago

fba325be

Build binaries with local version string specifying CUDA version (#23325)

ezyang committed 5 years ago

ff3cc795

Enabled cumsum and cumprod for bool tensors (#23346)

izdeby committed 5 years ago

cf0f3556

Open source 3D depthwise conv (#23164)

Du Tran committed 5 years ago

c9312e1a

Specifying libtorch variant in libtorch builds (#23348)

pjh5 committed 5 years ago

73dee44e

Switch keys to be sequential and stable in pickle serialization

zdevito committed 5 years ago

3297d8e2

Fix pickler bug where it would not load if no tensors were saved

zdevito committed 5 years ago

93da1030

Memoize storages in pickler

zdevito committed 5 years ago

7922b505

Newer Older