strisunshinewentingwang/pytorch

Pull Requests Commits

Revert "[opinfo] empty_strided (#100890)"

pytorchmergebot committed 1 year ago

dfac4364

match sdpa patterns from HF (#100609)

Natalia Gimelshein committed 1 year ago

f33725b8

[quant][pt2] Handle maxpool + conv + bn case in prepare QAT (#100941)

andrewor14 committed 1 year ago

8e51521c

Adding serialization ID to inline container (#100994)

atannous committed 1 year ago

3ed1569e

Support map autograd and pytree in/out. (#101633)

ydwu4 committed 1 year ago

326a4cc8

Handle multi-user case in split-cat simplification (#101473)

devashishshankar committed 1 year ago

38e537db

Fix determenistic typos (#101631)

kit1980 committed 1 year ago

e17d9f2c

[PT2][Quant] Move to module partitioner for linear pattern quantization (#101122)

kimishpatel committed 1 year ago

07e759ec

[transformer benchmark] sort by cuda time (#101349)

jspark1105 committed 1 year ago

ebae77e8

Fix benchmark model names printouts with tqdm (#101627)

jansel committed 1 year ago

403ce1a1

[PT] Update module partitioner to return parameter node (#101121)

kimishpatel committed 1 year ago

bec655f8

inductor(CPU): fix issue when padding/stride/dilation size is one for cpu weight packing pass(reland) (#101353)

XiaobingSuper committed 1 year ago

75375b41

[PT2][Quant] Remove None annotations (#101120)

kimishpatel committed 1 year ago

2c807a4a

[functorch] fix UB in interpreter stack (#101568)

zou3519 committed 1 year ago

783a46ad

s390x SIMD: Propagate NaN in minimum and maximum operations (#99716)

AlekseiNikiforovIBM committed 1 year ago

cb6fa890

s390x zvector: implement expm1 for complex vectorized types (#99872)

AlekseiNikiforovIBM committed 1 year ago

a85f6aa4

Implement CSE for dynamo guards. (#98488)

ysiraichi committed 1 year ago

f72f0119

[dynamo] Change dimension constraint summary to log.info (#101584)

angelayi committed 1 year ago

f994d0b6

Switch AOT Inductor test to export, add dynamic, fix invocation bug (#101585)

voznesenskym committed 1 year ago

39f52c02

Implement adding bias vector into structured sparse linear operator (#100881)

alexsamardzic committed 1 year ago

c3a893c6

Enables barrier to support the specified device (#99589)

shaoyf42 committed 1 year ago

97180aca

[inductor][cpp] support non contiguous vectorization codegen (#99966)

Valentine233 committed 1 year ago

6261aa5c

Actually functionalize torch.export (#101433)

tugsbayasgalan committed 1 year ago

47f43ed8

Extend storage create for custom storageImpl (#100237)

fakeYan committed 1 year ago

0c470b17

Fix Buck OSS build after flatbuffers update in #100716 (#101626)

huydhn committed 1 year ago

d1a472a3

work around precision error in constraint solver (#101607)

avikchaudhuri committed 1 year ago

41d668c9

[vision hash update] update the pinned vision hash (#101635)

pytorchmergebot committed 1 year ago

3c4f97c2

Enable `backward` on `_foreach_zero_` (#101149)

crcrpar committed 1 year ago

ba2bc7df

[inductor] Refactor RNG operators (#100064)

jansel committed 1 year ago

3bbf0683

[MPS] Add histogram ops (#96652)

qqaatw committed 1 year ago

bb355896

Newer Older