strisunshinewentingwang/pytorch

Pull Requests Commits

[Distribution] Implement kl divergence for Cauchy distribution (#36477)

xidulu committed 4 years ago

4e365b9c

fix explicitly defaulted constexpr assignment operator fails to compile error for gcc 5.3.0 (#36561)

XiaobingSuper committed 4 years ago

4d2502a0

[quant][graphmode] tanh pattern and test (#36880)

z-a-f committed 4 years ago

e0e70589

[profiler] Allow record_function ctx manager to profile futures (#35055)

rohan-varma committed 4 years ago

752d3c28

Report error for lt, le, gt, ge in complex Vec256 (consistent with <, <=, >, >=) (#36646)

xuhdev committed 4 years ago

1e054bfb

ROCm: don't warn about CUDA compute capabilities (#35949)

t-vi committed 4 years ago

dc4d8881

add a fast path for EmbeddingBag calling FBGEMM (#36679)

xing-liu committed 4 years ago

2fa17ded

[rpc] Remove redundant call to createExceptionResponse (#36857)

rohan-varma committed 4 years ago

68f847c4

Add at::aten::hardsigmoid symbol (#36851)

JackCaoG committed 4 years ago

399f494d

ai-pep: match the qlinear benchmark to linear (#36674)

vkuzo committed 4 years ago

13391ceb

ai-pep: align qconv benchmark to conv (#36673)

vkuzo committed 4 years ago

25649684

Bmm sparse dense (#33430)

kurtamohler committed 4 years ago

c7cf4c1b

Revert D21078446: [pytorch] Route default warning sync to LOG(WARNING)

Dmytro Dzhulgakov committed 4 years ago

30e7055e

Makes CUDA -float->uint8 cast consistent with CPU (#36832)

mruberry committed 4 years ago

0f0d6900

[pytorch] Route default warning sync to LOG(WARNING) (#36768)

Dmytro Dzhulgakov committed 4 years ago

9d5dda7c

Logging for ATen op subtype

Dmytro Dzhulgakov committed 4 years ago

49457a7b

Backward-compatible workaround for ATenOp index with dtype=uint8 (#36667)

Dmytro Dzhulgakov committed 4 years ago

246e9abf

Remove CUDA9Workarounds.cuh (#36840)

zasdfgbnm committed 4 years ago

60c30606

Update persons_of_interest.rst

jspisak committed 4 years ago

Verified 3c55b5a8

Fix MaxPool3d CUDA backward incorrect results for non-square output (#36820)

linziyi96 committed 4 years ago

1341ea48

[WIP] reenable bfloat16 masked_select (#36859)

ngimel committed 4 years ago

1b3741aa

Minor tweak of FakeLowp CMakefile (#36861)

Yinghai Lu committed 4 years ago

be9748f2

Revert D20896697: [pytorch][PR] QuantizedCUDA implementation

albanD committed 4 years ago

49b10c58

QuantizedCUDA implementation (#35463)

Aleksandr Fedorov committed 4 years ago

f6daa622

Use both absolute and relative tolerance in testing (#34258)

nairbv committed 4 years ago

54ed6fd3

[AIDemos] Add missing operators for AIDemos (#36756)

xta0 committed 4 years ago

3aec9f79

[nnpi glow unit test] SLS tests shape sweep with hypothesis testing (#36833)

amylittleyang committed 4 years ago

b0b9e704

C++ make constructor NamedAnyModule(name,any) public (#36869)

meganset committed 4 years ago

8b685a8a

Vectorize reduction when reducing on fastest striding dimension [resubmit] (#36873)

zasdfgbnm committed 4 years ago

6ba734ba

Enhance error message for MPI unavailability. (#36781)

pritamdamania committed 4 years ago

136d84dd

Newer Older