strisunshinewentingwang/pytorch

Pull Requests Commits

clee2000 committed 2 years ago

16995ce0

some unique ones

clee2000 committed 2 years ago

955ed10c

[inductor] fix typos in test_torchinductor.py (#96233)

Aidyn-A committed 2 years ago

f5c39b7b

[ONNX] Merge 'initializers' into 'TorchScriptGraph' (#95676)

BowenBao committed 2 years ago

0f4652f4

[EASY] Add complex dtypes to partitioner (#96297)

ezyang committed 2 years ago

e9e6b3b6

--subprocess for pytest (#96210)

clee2000 committed 2 years ago

a7fe11de

Set ref for linux_job checkout in lint (#96317)

clee2000 committed 2 years ago

8921b222

Add basic Module serialization BC test (#96238)

albanD committed 2 years ago

c8216e55

Fix usages of contextmanager without finally (#96170)

Chillee committed 2 years ago

5bbec680

Remove unimported expecttest deps and usage (#96314)

kirklandsign committed 2 years ago

34d18c8b

[TorchScript] Fix torch.cuda._exchange_device (#95306)

Aidyn-A committed 2 years ago

0f6d6d61

[reland][inductor] Add an AOT compilation mode for Inductor CPP backend (#95985)

desertfire committed 2 years ago

deaf9e5e

Ignore shape inference exception from Caffe2 ATen fallback (#90408)

Thiago Crepaldi committed 2 years ago

b9c25f18

[EASY] Update inductor training dynamic skips (#96298)

ezyang committed 2 years ago

c988de10

[CI] Add a workflow for quick perf comparison (#96166)

desertfire committed 2 years ago

b3a07981

Move MacOS x86_64 build and test jobs to periodic (#96279)

huydhn committed 2 years ago

4a1b9717

Revert "Error when jit.trace/script is used with torch.compile (#91681)"

pytorchmergebot committed 2 years ago

9137f53e

Notify on outdated lintrunner (#96241)

ZainRizvi committed 2 years ago

7362e22f

[SDPA] Add an optional scale kwarg (#95259)

drisspg committed 2 years ago

11aab72d

Revert "Ignore shape inference exception from Caffe2 ATen fallback (#90408)"

pytorchmergebot committed 2 years ago

3f840cc6

[BE] Delete `pre-cxx-11-abi MacOS libtorch builds (#96301)

malfet committed 2 years ago

9c5a24b9

[Quant][test] Add test for mixed dtypes in the same model (#96104)

andrewor14 committed 2 years ago

e7dd9b11

Ignore shape inference exception from Caffe2 ATen fallback (#90408)

Thiago Crepaldi committed 2 years ago

1d4e8723

[aot autograd] merge all outputs of funtionalization analysis into single metadata (#95991)

bdhirsh committed 2 years ago

98ece750

aot autograd: handle detach() and no_grad() mutations on input (#95980)

bdhirsh committed 2 years ago

29b216ac

[inductor] do not handle `int` in `placeholder` (#96230)

nkaretnikov committed 2 years ago

bb650b34

aot autograd: dont allow symint outputs to get tangents in the bw graph (#96219)

bdhirsh committed 2 years ago

f96bd528

Revert "Fix hooks handling for unpickled nnmodule (#96224)"

pytorchmergebot committed 2 years ago

6bbae862

Hooking backward for QNNPACK (#94432)

kwanghoon-meta committed 2 years ago

a1d7014c

[FSDP][optim_state_dict] Fix a memory leakage in optim_state_dict (#96263)

fegin committed 2 years ago

92edac72

Older