strisunshinewentingwang/pytorch

Pull Requests Commits

Fix wrong lintrunner version (#87295)

huydhn committed 2 years ago

888e1540

Ensure that symbolic variables incorporate fresh constraints before they're used (#87254)

Chillee committed 2 years ago

bd757b36

run torch::deploy test using pip install (#86507)

PaliC committed 2 years ago

bcde7542

[rpc] Wrap exception creation with try/catch (#87224)

rohan-varma committed 2 years ago

07bd053a

[discussion] fix for aot autograd outputs that dont require grad (#86838)

ezyang committed 2 years ago

c97ffcff

Fix line numbers bug (#87247)

mlazos committed 2 years ago

c9b61844

`cuda._is_in_bad_fork`->`_C._cuda_isInBadFork` (#87317)

malfet committed 2 years ago

c8889f4e

[Dynamo] Support optimizing over any Tensor with requires_grad = True (#87141)

yanboliang committed 2 years ago

56b150ac

Symintify pad ops (#87046)

albanD committed 2 years ago

12b2f70a

Advance nightly docker to 11.6 (#86941)

atalman committed 2 years ago

c5de535b

OpInfo: Sample input cleanup (4/n) (#86324)

peterbell10 committed 2 years ago

6eeeb881

Fix compilation warning and spurious print (#87297)

albanD committed 2 years ago

c141f28b

Tweak several test serialization to store models state_dict (#87143)

malfet committed 2 years ago

4a533f12

[maskedtensor] add docs (#84887)

george-qi committed 2 years ago

cf2be34f

Revert "[primTorch] Add refs for `softmax`, `softmin`, `log_softmax` (#84956)"

pytorchmergebot committed 2 years ago

cd216135

[FSDP] Add set_state_dict_type API to setup state_dict_type without using context manager (#86243)

fegin committed 2 years ago

c08c7997

Revert "Dynamo FX graph stack traceback fix (#87136)"

pytorchmergebot committed 2 years ago

f3cc588d

[primTorch] Add refs for `softmax`, `softmin`, `log_softmax` (#84956)

nkaretnikov committed 2 years ago

c09ca93e

[allocator] disable tests that don't work for cudaMallocAsyncAllocator (#87250)

zdevito committed 2 years ago

00c91f44

[functorch] Get rid of defunct functorch/setup.py (#87235)

zou3519 committed 2 years ago

15ca6852

[functorch] add test for torch.manual_seed inside grad transform (#87233)

zou3519 committed 2 years ago

ac80da22

[allocator] Move getFreeMutex (#87237)

zdevito committed 2 years ago

f56ce8db

Dynamo FX graph stack traceback fix (#87136)

williamwen42 committed 2 years ago

89e6078b

Reenable aot tests on windows for cuda 11.7 and up (#87193)

atalman committed 2 years ago

40d0fa53

Pin ios conda dependencies (#87229)

huydhn committed 2 years ago

86a58192

[MPS] Do not dispatch empty job in `bitwise_not` (#87286)

malfet committed 2 years ago

a79e034d

fix 0d cpu tensor handling when it's the first arg (#87273)

Natalia Gimelshein committed 2 years ago

6775c3e1

add an API for external backends to register custom device names (#86992)

bdhirsh committed 2 years ago

fb6826bf

Clean Inductor complication cache during dynamo dashboard run (#87246)

williamwen42 committed 2 years ago

cc64863d

functionalization: skip meta reference compute for aot autograd (#87108)

bdhirsh committed 2 years ago

b3071e2e

Newer Older