strisunshinewentingwang/pytorch

Pull Requests Commits

[Pytorch Edge] Add selective macros to metal ops (#68134)

JacobSzwejbka committed 2 years ago

97a38680

Sparse CSR CPU: add `torch.add` with all inputs sparse (#64391)

IvanYashchuk committed 2 years ago

c2642b64

Fast cuda layer norm (#67977)

ngimel committed 2 years ago

84d3df80

Add host-side memory requirement for `test_softmax_64bit_indexing` (#67922)

eqy committed 2 years ago

a1ace029

OpInfo for `nn.functional.conv1d` (#67747)

krshrimali committed 2 years ago

9e7b3143

Implement Entropy methods for Binomial and Multinomial distributions (#67609)

Vinnam Kim committed 2 years ago

35f16170

[nnc] aotCompiler outputSpec support quantized outputs (#67711)

IvanKobzarev committed 2 years ago

864c6b37

[nnc] Lazy lowerings registration; custom classes network params (#67623)

IvanKobzarev committed 2 years ago

362c6069

Add feature: zeros_like() from a dense tensor to a sparse tensor (#68108)

Vinnam Kim committed 2 years ago

f89572f4

Ease constrain for fuse path in trt lower (#68148)

Shirong Wu committed 2 years ago

5efe5e24

OpInfos for new_blah functions and some _like functions (#67357)

zou3519 committed 2 years ago

d4ae7896

Working POC of define-by-run quantization (#64676)

vkuzo committed 2 years ago

4466ba8f

[Dist CI][BE] Run each test in its own process for test_distributed_spawn (#67901)

rohan-varma committed 2 years ago

f02efc74

skip test_jit_legacy (#68129)

Krovatkin committed 2 years ago

aea4e61e

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

Newer Older