strisunshinewentingwang/pytorch

Pull Requests Commits

Automated submodule update: kineto (#67445)

facebook-github-bot committed 2 years ago

a6a26165

Add complete type name in error message when fail to export model (#67750)

cccclai committed 2 years ago

a229c3e5

[SR] Fix aten::split schema (#68135)

Mike Iovine committed 2 years ago

1f07efd0

[SR] Add runtime check to correct bad schema alias info (#67825)

Hao Lu committed 2 years ago

47bc47f2

[PyTorch Edge] Better error message when training attribute is not found (#68103)

dhruvbird committed 2 years ago

ca7d0062

Make `torch.fx.experimental.fx2trt.passes` a package (#68139)

malfet committed 2 years ago

0e366b8e

add unpack_sequence and unpad_sequence functions (#66550)

dayyass committed 2 years ago

f171c78c

Fix lambda function broke torch.save

Shirong Wu committed 2 years ago

a510f413

Update unpack_dual to return named tuple (#68062)

soulitzer committed 2 years ago

22e73f61

[LT] Upstream backend interfaces (#67927)

wconstab committed 2 years ago

d6e6064e

Update rpc testing to include USE_TENSORPIPE directive (#68080)

H-Huang committed 2 years ago

c075f0f6

don't include label in ci: sev issue (#68093)

seemethere committed 2 years ago

a3bb95c1

[SR] Native implementation for aten::split (#67476)

Mike Iovine committed 2 years ago

ecd5b1a8

Logger integration format (#67962)

FusionPower committed 2 years ago

746a31b2

don't hardcode mask type in mha (#68077)

ngimel committed 2 years ago

8dfbc620

torch.allclose opinfo (#68023)

george-qi committed 2 years ago

ae586449

Factor backend routing logic out of convolution forward (#67790)

jbschlosser committed 2 years ago

9a2db6f0

Fixed deprection warnings with `.data<T>()` in SpectalOps.cpp (#67993)

vfdev-5 committed 2 years ago

147de824

[LTC] Upstream class BackendDevice (#68027)

Jiewen Tan committed 2 years ago

6011c35a

fix gradcheck to generate valid input for forward AD complex (#68001)

albanD committed 2 years ago

a6c0edff

Adds an optimizer instance variable to ChainedScheduler (#68010)

oliver committed 2 years ago

94b6fa6f

[PyTorch Edge] Don't use LeftRight in mobile (#66064)

dhruvbird committed 2 years ago

cb2a41e5

[PyTorch] Avoid reading file from stream for 0 byte Tensor storage (#67787)

dhruvbird committed 2 years ago

b0817e19

[PyTorch] Replace copy_ with data_ptr<float>() since input Tensor's dtype is guaranteed to be float (#67788)

dhruvbird committed 2 years ago

bf31d4b2

aliasing fixes (#66977)

Elias Ellison committed 2 years ago

6b44e75f

Fix ads dense arch regression (#68071)

Shirong Wu committed 2 years ago

3f1a3f7b

remove Generate* macro files (#67940)

ngimel committed 2 years ago

91af74c9

Add an option to disable reduced precision reductions for FP16 GEMM (#67946)

eqy committed 2 years ago

790763b0

[nnc][mobile] temporarily disable quantization external functions (#68029)

ljk53 committed 2 years ago

078c6559

Simplify example for nn.Flatten (#67472)

Dmitry committed 2 years ago

b1a42298

Newer Older