strisunshinewentingwang/pytorch

Pull Requests Commits

GHA: add back runner type for distributed tests (#67336)

janeyx99 committed 2 years ago

7d1c0992

[LTC] Code-gen exp, elu*, and log_sigmoid*

ramiro050 committed 2 years ago

a25e8855

Add support for vmap+fwdAD for basic out-of-place op (#66291)

soulitzer committed 2 years ago

f2f7b02b

Add barrier to ProcessGroup trampoline (#67236)

mrshenli committed 2 years ago

a3aa9df5

[tensorexpr][ir][quant] Adding qscale and qzero to tensorexpr IR Buf (#66675)

IvanKobzarev committed 2 years ago

e52d0e77

Enable c10d trampoline tests on MacOS (#67205)

mrshenli committed 2 years ago

632719c2

[hpc][inference] enable cuda graph in engine holder (#66738)

Bangsheng Tang committed 2 years ago

c88da701

Convert lazy_tensors:: types to c++, std, or c10

wconstab committed 2 years ago

e0fb8320

[LTC] Refactor gen_lazy_nativefunc_definition to a functor

Jiewen Tan committed 2 years ago

27f3ad4f

[LTC] Refactor how shape functions are declared and called

Jiewen Tan committed 2 years ago

949889bc

[LTC] Code-gen relu/relu_

Jiewen Tan committed 2 years ago

57a60c5a

[LTC] Add support for non-structured in-place operator variants

Jiewen Tan committed 2 years ago

01df3ca8

[Vulkan] Add vulkan_perf_test with google benchmark (#67230)

beback4u committed 2 years ago

28570664

[Vulkan] Optimize cat operator for channel dimension (#67207)

beback4u committed 2 years ago

cdc9b262

Revert D31937065: [pytorch][PR] fix binding to the wrong python module

malfet committed 2 years ago

d691bc12

[Pytorch][Bootcamp] Add fix and testing for non-vectorized Adadelta optimizer to handle complex numbers (#66587)

Christopher Gray Howard committed 2 years ago

dfa7225a

Revert D31935958: Add register_frozenpython.cpp to the torch::deploy interpreter library in the OSS build

seemethere committed 2 years ago

fcefed95

.github: Switch 8xlarge to 4xlarge instance_type (#67299)

seemethere committed 2 years ago

1541bb82

Convert lazy_tensors::PrimitiveType to c10::ScalarType

wconstab committed 2 years ago

2254e6b4

Replace Cat op with new codegen-like hand-coded variant

wconstab committed 2 years ago

feeec74f

fix binding to the wrong python module (#67246)

Krovatkin committed 2 years ago

7ac8ed74

[Pytorch Delegated Backend] Add macro to define sentinel value of debug handle. (#66584)

kimishpatel committed 2 years ago

0e8bd0c8

[LTC] Code-gen nll_loss_forward/nll_loss_backward (#67238)

alanwaketan committed 2 years ago

Verified 7f363df9

Add register_frozenpython.cpp to the torch::deploy interpreter library in the OSS build (#67280)

shunting314 committed 2 years ago

00b0d4ee

[jit][edge] Export maybe-used interface methods from modules. (#65966)

zhxchen17 committed 2 years ago

f510193e

disallow requires_grad=True in make_tensor for integral inputs (#67149)

ngimel committed 2 years ago

a72a6365

.github: Use 4xlarge instances for linux gpu (#67264)

seemethere committed 2 years ago

81d18810

Port triangular_solve to structured kernel (#61857)

IvanYashchuk committed 2 years ago

fdc74e23

[PyTorch][Static Runtime] Cleanup: add valueVecFromFastSet (#66996)

swolchok committed 2 years ago

6ce14e7b

[PyTorch][Static Runtime][easy] Fix ValueGroup comment (#66965)

swolchok committed 2 years ago

066a980e

Newer Older