strisunshinewentingwang/pytorch

Pull Requests Commits

Fix autograd issue with identity conversions

alexsamardzic committed 1 year ago

94a5bb57

[ONNX] Support aten::bit_wise_not in fx-onnx exporter (#94919)

titaiwangms committed 1 year ago

28e69954

[complex] nansum & nanmean (#93199)

khushi-411 committed 1 year ago

a0389681

Inductor: fix incorrect result of inplace unsqueeze (#94797)

blzheng committed 1 year ago

6ae06e49

Revert "Re-enable a FX-to-ONNX kwargs Test (#94763)"

pytorchmergebot committed 1 year ago

aa9e481e

Revert "Change test_torchinductor_opinfo.py to mark skips/xfails in a better way (#94813)"

pytorchmergebot committed 1 year ago

a049bbb1

[vision hash update] update the pinned vision hash (#94866)

pytorchmergebot committed 1 year ago

e7515538

Enable half type support for unique cpu (#91666)

zyan0 committed 1 year ago

753c33bf

Re-enable a FX-to-ONNX kwargs Test (#94763)

wschin committed 1 year ago

04b4704a

[dynamo] keep submodule's name for nn.Sequential when unroolling (#94913)

ydwu4 committed 1 year ago

4b2d1bec

[inductor] enable `test_lowmem_dropout1_dynamic_shapes` (#94884)

nkaretnikov committed 1 year ago

8c44ae2f

fix: make sure `sorter` indices are inbound in `searchsorted` (#94863)

ganler committed 1 year ago

5e1de315

Hide failing merge rule's name in the internal debugging section (#94932)

ZainRizvi committed 1 year ago

a863d5e3

[dynamo] support custom __getattr__ on torch.nn.Modules (#94658)

davidberard98 committed 1 year ago

a4085ab8

Change test_torchinductor_opinfo.py to mark skips/xfails in a better way (#94813)

Fabio Rocha committed 1 year ago

bfc0d5e2

[ONNX] Enable skipped gpt2 test (#94930)

BowenBao committed 1 year ago

3d40a86a

[BE][1/N] Add deprecate msg to Sharded Partial and Replicate Tensor (#94928)

fduwjj committed 1 year ago

b4c81867

[SDPA] Update dispatch logic to check for sm86 and head_size == 128 for flash attention (#94921)

drisspg committed 1 year ago

07bc6b95

[executorch] Add RuntimeContext to generated C++ API Signature (#94570)

larryliu0820 committed 1 year ago

41865bd8

Add check that embedding_bag's weight is 2D (#94931)

soulitzer committed 1 year ago

e5c2a35d

[mta] implement `_foreach_pow` (#92303)

crcrpar committed 1 year ago

3e9df622

Enable persistent reductions (#94847)

jansel committed 1 year ago

e28ba681

add backwards for layer norm nested (#94781)

drisspg committed 1 year ago

0d7913c9

Add some simple sanity tests to ValueRanges (#94905)

ezyang committed 1 year ago

904d549c

[MPS] Move max_pool2d to mps dispatch key (#90772)

qqaatw committed 1 year ago

e8dc34ea

[SPMD] Pull the minimal working distribute API and SPMD module to PyTorch (#94802)

fegin committed 1 year ago

250c054b

[MPS] Fix bilinear backward pass (#94892)

DenisVieriu97 committed 1 year ago

bc361fdf

[pt2][inductor] update choice caller hashes (#94853)

nmacchioni committed 1 year ago

dd7e2b7c

Revert "Fix XNNPACK OSS Buck build (#94935)"

pytorchmergebot committed 1 year ago

0698af67

Add float to list of allowed ops (#94910)

malfet committed 1 year ago

c01f5118

Older