strisunshinewentingwang/pytorch

Pull Requests Commits

Disallow module forward input mutation in aot_export (#101834)

tugsbayasgalan committed 2 years ago

b5ee34e5

[inductor] Refactor RNG operators (#100064)

jansel committed 2 years ago

0c6f409c

[dynamo] Minor refactor to use is_allowed to decide inlining of NNModule methods (#101910)

anijain2305 committed 2 years ago

8b2a9f81

[vision hash update] update the pinned vision hash (#101917)

pytorchupdatebot committed 2 years ago

2886b3e6

inductor: fix name 'inf' is not defined issue when calling external_call function (#101865)

XiaobingSuper committed 2 years ago

bb62a373

inductor: fix bfloat16 store complier issue (#101856)

XiaobingSuper committed 2 years ago

350f0cd7

[accuracy minifier] cast copied model rather than update the original model (#101901)

shunting314 committed 2 years ago

029c6a99

[easy] refactor signature flattening transform (#101886)

avikchaudhuri committed 2 years ago

73e887b5

[dynamo] Bugfix for unspecialized nn module variable (#101859)

anijain2305 committed 2 years ago

7a17e9d0

[inductor] Test indirect indexing asserts with dimension of size 1 (#101811)

peterbell10 committed 2 years ago

48346a46

[inductor] Implement magic methods on IR values (#101076)

peterbell10 committed 2 years ago

89bd5d3d

[quant][pt2e] Introduce QuantizationAnnotation API (#101708)

jerryzh168 committed 2 years ago

15495f2d

[codemod][3.10][NamedTuple] Use typing_extensions to get NamedTuple Generics (#101830)

fried committed 2 years ago

03f50fcc

Fix finding existing Needs label comments (#101889)

kit1980 committed 2 years ago

b07e97c0

[pt2] Turn off lazy reinit when cuda graph is on (#101848)

xw285cornell committed 2 years ago

c8fd1cfa

[Profiler] Workaround CUPTI Lazy Reinit and CUDA Graphs crash in CUDA 11 (#101879)

aaronenyeshi committed 2 years ago

fa7ad77a

Dynamic Shape Doc (#101885)

msaroufim committed 2 years ago

3666ca9d

Fake Tensor Docs (#101882)

msaroufim committed 2 years ago

ff5b9428

Add Logging Doc to compile index (#101888)

msaroufim committed 2 years ago

581d13a0

Revert "separate out dynamo .requires_grad and .is_grad_enabled guards (#100570)"

pytorchmergebot committed 2 years ago

7f3fed12

Docs for torchcompile and functorch (#101881)

msaroufim committed 2 years ago

2dd33c71

Update BCEWithLogitsLoss pos_weight description in documentation (#101567)

cviviers committed 2 years ago

81c181dc

match sdpa patterns from HF (#100609)

Natalia Gimelshein committed 2 years ago

5ea7096e

Print restarting analysis at INFO level with a exception breadcrumb (#101573)

ezyang committed 2 years ago

96ee23e1

Update batch size for a couple models (#101837)

eellison committed 2 years ago

e5e451a9

Revert " fix missing-prototypes warnings in torch_cpu (Part 4) (#100849)"

pytorchmergebot committed 2 years ago

498c34e8

Revert "fix inference_mode with torch.compile (#101219)"

pytorchmergebot committed 2 years ago

083f304d

Revert " fix missing-prototypes warnings in torch_cpu (Part 5) (#101788)"

pytorchmergebot committed 2 years ago

e760a968

[export] Error when constraining on static values (#101655)

angelayi committed 2 years ago

4f9aa7cb

[BE]: Enable ruff rule TRY302 and apply fixes (#101874)

Skylion007 committed 2 years ago

3e2ea32d

Newer Older