strisunshinewentingwang/pytorch

Pull Requests Commits

Port set_ to ATen. (#34403)

gchanan committed 4 years ago

ca1e2cda

[JIT] add id function (#34975)

Elias Ellison committed 4 years ago

62f11f0a

Add TensorExpr Fuser tests (resubmit). (#35085)

Mikhail Zolotukhin committed 4 years ago

12f0052e

Add guard elimination cases for operators encountered on an RL workload. (#34967)

resistor committed 4 years ago

3c409fc6

Revert D20254663: [pytorch][PR] Vectorize in-place comparison operators

ezyang committed 4 years ago

faa853fe

[android] Maven publishing license fix (#32474)

IvanKobzarev committed 4 years ago

ea41bf31

Add tensorexpr benchmarks. (#35064)

Mikhail Zolotukhin committed 4 years ago

8998a1b3

fix missing comma in activation benchmarks (#35104)

vkuzo committed 4 years ago

bf41a762

randn cuda kernel complex dtype (#35056)

anjali411 committed 4 years ago

7d5a8998

Define +, -, *, / between complex numbers and integers (#34506)

xuhdev committed 4 years ago

451e4d57

Vectorize in-place comparison operators (#33252)

xuhdev committed 4 years ago

91d39de1

Adds truncated normal initializer (#32397)

Enealor committed 4 years ago

8bcedf7d

use new cuda kernel launch code in nvprof parsing (#35016)

xwang233 committed 4 years ago

a5b5ea98

[caffe2] SWA operator (#34394)

pcpLiu committed 4 years ago

e3272559

[C++ API Parity] Add xor_convergence test for lbfgs (#35001)

anjali411 committed 4 years ago

781f590f

`Engine::~Engine()` should wait for non-reentrant threads to shutdown (#34529)

malfet committed 4 years ago

1c958f8e

Enforce rref python pickling to be in the scope of RPC call (#34755)

zhaojuanmao committed 4 years ago

ec9f6809

[nomnigraph] Copy device option when customize the op conversion (#34976)

Yinghai Lu committed 4 years ago

6000dca5

Revert D20541921: [pytorch][PR] [RELAND] Eager autocasting, out-of-place ops only (with MSVC 2017 fix)

Mike Ruberry committed 4 years ago

fe276d54

Revert e7fc55e (#35080)

lly-zero-one committed 4 years ago

0ccdec6b

torch.cat: disallow inputs on different devices (#35053)

Baranowski committed 4 years ago

eb78f7ea

Fix torch.mm export to ONNX (#34661)

KsenijaS committed 4 years ago

89110fbe

Also vectorize complex types in fill. (#34973)

xuhdev committed 4 years ago

e65ac7af

repr and _*state_dict for qRNN (#31540)

z-a-f committed 4 years ago

463f7920

[RELAND] Eager autocasting, out-of-place ops only (with MSVC 2017 fix) (#35011)

definitelynotmcarilli committed 4 years ago

991b9727

[ROCm] Enable BFloat16 type for TopK operator on ROCm. (#34849)

rohithkrn committed 4 years ago

edb794fb

Reduce the number of iterations in test_autograd_context (#35037)

mrshenli committed 4 years ago

bb63710c

Revert D20540599: Add TensorExpr Fuser tests.

Natalia Gimelshein committed 4 years ago

3c90a907

Revert D20464855: [pytorch][PR] Add the fusion of quantized batchnorm and relu

Mike Ruberry committed 4 years ago

e7fc55ef

enforce rref JIT pickling to be in the scope of rpc calls (#34689)

zhaojuanmao committed 4 years ago

a4afac60

Newer Older