strisunshinewentingwang/pytorch

Pull Requests Commits

Fix erase_number_types test (#23181)

smessmer committed 5 years ago

19be7ece

Fix onnx export (#23180)

smessmer committed 5 years ago

e56f11b7

DictConstruct sets correct types (#23171)

smessmer committed 5 years ago

60afcabc

Exclude unused onnx targets (#23195)

bddppq committed 5 years ago

67aede98

ListConstruct sets correct element type (#23189)

smessmer committed 5 years ago

9d03133c

Use concrete types in jit test for generic lists (#23192)

smessmer committed 5 years ago

2073cc73

Remove trailing semicolon from TORCH_CHECK macros.

ezyang committed 5 years ago

21f52ce0

Switch from KaTeX to imgmath for documentation rendering.

ezyang committed 5 years ago

174f7a58

Fix typos in comments

ssnl committed 5 years ago

792d5277

Let CMake handle NCCL detection instead of our handcrafted Python script. (#22930)

xuhdev committed 5 years ago

60c46dd4

Fix typo in dataloader.py

ssnl committed 5 years ago

e4b75c65

Add document of function torch.as_strided (#22842)

kianasun committed 5 years ago

45d3f495

Update nccl to 2.4.8-1 (#23186)

soumith committed 5 years ago

c9e62f69

Updating submodules

svcscm committed 5 years ago

9a6ae5c0

quantized conv module (#23178)

jerryzh168 committed 5 years ago

d7448c78

ConvReLU2d module (#23008)

jerryzh168 committed 5 years ago

f3a37278

Export torch.arange to ONNX (#22601)

BowenBao committed 5 years ago

eb5137a5

Revert D16368004: [pytorch][PR] Fix error message for a wrong fork CUDA

pjh5 committed 5 years ago

06d11f04

Verify flatten works for quantized Tensor (#23121)

jerryzh168 committed 5 years ago

38615206

Fix broken indexing when using None and ellipses indexing together (#22905)

Chillee committed 5 years ago

a24f6c13

Fix load op to return the shape info as before when loading multiple blobs (#23182)

pradeepdfb committed 5 years ago

648f10be

nn_quantized test (#23169)

jerryzh168 committed 5 years ago

1c574458

Turning on fbgemm for nightlies (#22784)

pjh5 committed 5 years ago

e08f8f45

Fix error message for a wrong fork CUDA (#23030)

Guanheng Zhang committed 5 years ago

a6e45a69

Add get_accessed_features function to ModelLayer class (#23036)

kevinwilfong committed 5 years ago

3ca7c0ff

Pin numba to 0.44.0 to fix Windows CI.

ezyang committed 5 years ago

ff23a02a

Remove empty THCThreadLocal{.h/.cpp} (#23157)

vishwakftw committed 5 years ago

b6d06d54

Invert ownership between PyFunction and THPFunction.

ezyang committed 5 years ago

fdfc676e

Support converting Python number to IValue in pybind_utils.h (#22817)

mrshenli committed 5 years ago

ae5b5208

BlackBoxPredictor OSS part 4: Open-source other transforms (#23099)

salexspb committed 5 years ago

2becbd3f

Newer Older