strisunshinewentingwang/pytorch

Pull Requests Commits

fix searchsorted output type (#42933)

albanD committed 4 years ago

1f6d0985

Clip Binomial results for different endpoints in curand_uniform (#42702)

mattip committed 4 years ago

059aa34b

Add back Tensor.nonzero type annotation (#43053)

rgommers committed 4 years ago

71bbd5f1

Remove `itruediv` because it's already defined in torch/tensor.py (#42962)

KKawamura1 committed 4 years ago

75dfa5a4

Add complex tensor dtypes for the __cuda_array_interface__ spec (#42918)

awthomp committed 4 years ago

1c616c5a

Test the type promotion between every two dtypes thoroughly (#42585)

xuhdev committed 4 years ago

c3fb1522

Add inplace option for torch.nn.Hardsigmoid and torch.nn.Hardswish layers (#42346)

miraliahmadli committed 4 years ago

ff6a2b0b

Build test_e2e_tensorpipe only if Gloo is enabled (#43041)

malfet committed 4 years ago

2f9fd8ad

Trim trailing whitespace

dreiss committed 4 years ago

31788ae1

Make Mish support large inputs. (#43037)

YesterdayDX committed 4 years ago

a2b86d95

Fix typo in collect_env.py (#43050)

vfdev-5 committed 4 years ago

c7d2774d

Automated submodule update: FBGEMM (#42834)

facebook-github-bot committed 4 years ago

d60d6d0d

Guard TensorPipe agent by USE_TENSORPIPE (#42682)

lw committed 4 years ago

ed242cbe

Get, save, and load module information for each operator (#42133)

taivu committed 4 years ago

ccd9f324

Fix illegal memory acess issue for CUDA versionn of SplitByLengths operator.

Ren Chen committed 4 years ago

e182ec97

Implement torch.nextafter (#42580)

muthuArivoli committed 4 years ago

b8102b15

torch.complex and torch.polar (#39617)

wjgan7 committed 4 years ago

e4373083

[TensorExpr] Cleanup logic in the TensorExpr fuser pass. (#42938)

Mikhail Zolotukhin committed 4 years ago

b9a105bc

[TensorExpr] Remove redundant checks from canHandle in TE fuser. (#42937)

Mikhail Zolotukhin committed 4 years ago

fc304bec

[TensorExpr] Wrap fuser in a class. (#42936)

Mikhail Zolotukhin committed 4 years ago

48c183af

Reconstruct scopes (#41615)

taivu committed 4 years ago

02c8ad70

Add more verbose error message about PackedSequence lengths argument (#42891)

malfet committed 4 years ago

3dc84531

Clearer Semantics and Naming for Customized Quantization Range Initialization in Observer (#42602)

Paul Shao committed 4 years ago

b992a927

[reland][quant][fix] Remove activation_post_process in qat modules (#42343) (#43015)

jerryzh168 committed 4 years ago

a55b7e2a

Back out "change pt_defs.bzl to python file"

ljk53 committed 4 years ago

8cf01c5c

Python/C++ API Parity: TransformerDecoderLayer (#42717)

VinodSKumar committed 4 years ago

830423b8

[quant][doc] Print more info for fake quantize module (#43031)

jerryzh168 committed 4 years ago

85752b98

[jit][static runtime] Simplify the graph and add operator whitelist (#43024)

bwasti committed 4 years ago

523b2ce9

Allow RPC to be initialized again after shutdown. (#42723)

pritamdamania committed 4 years ago

89b0b3bc

Nightly checkout tool (#42635)

scopatz committed 4 years ago

21823aa6

Newer Older