strisunshinewentingwang/pytorch

Pull Requests Commits

Use intrinsics for trigonometric functions on CPU (#26431)

xuhdev committed 4 years ago

6f92aa2f

Switch internal CUDA build to C++14 (#26757)

smessmer committed 4 years ago

5c67b014

Fix the Bernoulli distribution sampler (#26864)

EthanSteinberg committed 4 years ago

bf1d957d

Choose num_threads in parallel_for based on GRAIN_SIZE (#26886)

peterbell10 committed 4 years ago

e425bdb8

Get rid of -u (expansion of undefined variable) setting (#26907)

ezyang committed 4 years ago

9f0deb47

fix typo in job name: nigthly->nightly

kostmo committed 4 years ago

b2f671a3

Fix nuclear norm with requires_grad=True (#26303)

vishwakftw committed 4 years ago

43b07ff2

Fix circular deps in loading (#26758)

zdevito committed 4 years ago

0e3389dc

Refactor dispatch structure so fallback code lives inline. (#26367)

ezyang committed 4 years ago

78a52549

Change calling convention of ATenDispatch from getOp to callUnboxed. (#26857)

ezyang committed 4 years ago

272d7c02

Updating submodules

svcscm committed 4 years ago

0bfe12d0

batch size 0 tests for Quantize/Dequantize DNNLOWP ops (#26873)

jspark1105 committed 4 years ago

d3cab657

batch size 0 support in FC DNNLOWP operators (#26872)

jspark1105 committed 4 years ago

78b0c58a

batch size 0 tests for element-wise DNNLOWP ops (#26870)

jspark1105 committed 4 years ago

41c1cc2f

batch size 0 support in Conv DNNLOWP ops (#26871)

jspark1105 committed 4 years ago

1aaf4810

Add bitwise distributed reduction ops (#26824)

pietern committed 4 years ago

2991bfdb

Add some missing constructors to IValue.

ezyang committed 4 years ago

dec0b6b7

Make resize_as_ generic, so XLA works. (#26809)

ezyang committed 4 years ago

60b57d96

batch size 0 support in ChannelShuffle DNNLOWP op (#26858)

jspark1105 committed 4 years ago

8fb756d3

Add more inplace arguments to quantization top level API (#26782)

Dmytro Dzhulgakov committed 4 years ago

0a8a779a

Enable batch_size = 0 support in DNNLOWP Concat operator (#26849)

jspark1105 committed 4 years ago

5231699d

remove backward functions from jit-op-registry for mobile build (#26851)

ljk53 committed 4 years ago

60372dc7

add mobile friendly at:parallel_for backend

ljk53 committed 4 years ago

ed260748

Improvements to GuardElimination and InsertBailouts

Krovatkin committed 4 years ago

14d7d571

Updating submodules

svcscm committed 4 years ago

20ed6ba0

Remove unnecessary functions and cleanup code in quantization.cpp.

Mikhail Zolotukhin committed 4 years ago

058ba0e7

Fix building with PARALLEL_BACKEND=NATIVE_TBB (#26742)

Arnaud Bonnet committed 4 years ago

8f359a48

Remove three unused declaration. (#26699)

xuhdev committed 4 years ago

c25c507f

Try to disable annoying hypothesis warnings again (#26853)

James Reed committed 4 years ago

f37aa2de

Re-write of tensor-scalar quantized add

James Reed committed 4 years ago

20ebd13f

Newer Older