strisunshinewentingwang/pytorch

Pull Requests Commits

Don't graph break on patched module methods (#93115)

williamwen42 committed 3 years ago

5bae5805

[ FL-gradient quantization] Adding QNN unpack feature (#92714)

kwanghoon-meta committed 3 years ago

a2e0f8e5

Fix BC-breaking change introduced by #91499 (#93091)

malfet committed 3 years ago

661800a2

fixing flag to skip nvfuser_tests build (#93080)

jjsjann123 committed 3 years ago

7fade4f7

[vision hash update] update the pinned vision hash (#93114)

pytorchmergebot committed 3 years ago

e2739372

Install Torchvision in all Linux shards (#93108)

huydhn committed 3 years ago

074f5ce0

Get rid of dedicated inductor dynamic_shapes config (#93076)

ezyang committed 3 years ago

025ef99d

[dtensor][7/N] remove backend in with_comms (#93040)

wanchaol committed 3 years ago

f3fcc806

[DTensor] implement dist_cat as a sharding prop rule (#92677)

XilunWu committed 3 years ago

8b3e01cd

[ONNX] Export 'aten::index_put(self, mask, v)' when rank(mask) < rank(self) (#92862)

BowenBao committed 3 years ago

24172eeb

Add kwargs support to torch.export() API (#92013)

Thiago Crepaldi committed 3 years ago

95dfad9d

[ci] Move sm86 from trunk to pull (#93085)

Catherine Lee committed 3 years ago

ae171cf6

Fix topk IMA (#93095)

zasdfgbnm committed 3 years ago

d1807dc1

Make __deepcopy__ of GraphModule able to handle circular reference. (#93038)

qihqi committed 3 years ago

8d7f9e2f

[CI] Move parallel native builds to 3.8 (#93103)

malfet committed 3 years ago

ceb44350

[Reland] aot autograd explicitly errors on double backward (#92893)

soulitzer committed 3 years ago

f6f46ba3

Revert "Disable torch_jit_fuser_te for dynamo CI (#92945)"

pytorchmergebot committed 3 years ago

913cf290

Torchinductor randn_like lowering (#93005)

eellison committed 3 years ago

340811bf

Properly compute device for elementwise operations with CPU scalar tensor (#93073)

ezyang committed 3 years ago

1b5bfe9d

Update flatbuffer test models to match pkl models (#93022)

qihqi committed 3 years ago

1f352f7c

[MacOS] Explicitly use cmake from cloned conda environment (#92737)

huydhn committed 3 years ago

68a49322

Remove try catch in test_torchinductor (#93004)

eellison committed 3 years ago

15c46eb8

Make meshgrid support symbolic shapes (#93075)

ezyang committed 3 years ago

17803fb3

Don't copy name_to_input in OutputGraph (#93034)

mlazos committed 3 years ago

5de19dd3

Update XLA docker image to v0.8 (#93041)

huydhn committed 3 years ago

f30787e5

Update RELEASE.md with pinning xla and builder PRs (#93079)

atalman committed 3 years ago

d9f0d148

Add sparse COO tensor support to torch.sum(dim=..., keepdim=...) (#92979)

pearu committed 3 years ago

0e92bbe5

[BE][CI] Move more builds from 3.7 to 3.8 (#92928)

malfet committed 3 years ago

ca2a23c2

Setup shebang and set -x on generated runner script (#93007)

ezyang committed 3 years ago

729f1a8e

Revert "Improve `bsr @ strided` performance in `baddmm` for `bfloat16/half` with Triton kernels. (#88078)"

pytorchmergebot committed 3 years ago

7012d985

Newer Older