strisunshinewentingwang/pytorch

Pull Requests Commits

Update xla image tag (#94377)

yeounoh committed 1 year ago

5ea6f598

[Inductor] added aten.cauchy_ decomp (#92047)

min-jean-cho committed 1 year ago

66ae3aa0

Dynamo: Support min / max over iterables (#94350)

jbschlosser committed 1 year ago

0ce95c3a

Avoid guarding on zero-ness with meta tensors. (#94399)

ezyang committed 1 year ago

53a5c8c7

Track and record hint on SymNode and use when possible (#94201)

ezyang committed 1 year ago

dc70b00d

Dynamo: Fix graph break when iterating over tensor (#94326)

jbschlosser committed 1 year ago

b5ef37b9

Set torch.backends.cudnn.enabled to false when testing accuracy (#94363)

desertfire committed 1 year ago

7bfc5999

[ONNX] Use optional op to keep None in results for ONNX internal tests (#84789)

titaiwangms committed 1 year ago

04b06c96

[ONNX] Add full checker mode in torch.onnx.export (#83186)

titaiwangms committed 1 year ago

b27ac6dc

[dynamo 3.11] changes to python code object (#93985)

williamwen42 committed 1 year ago

4e984cb6

update aten op overload to not use `from` to avoid compile errors (#89797)

eellison committed 1 year ago

021d2676

Make triton debug util reusable (#94225)

wconstab committed 1 year ago

f2156ef4

[MPS] Add triangular solve op through MPSMatrixSolveTriangular (#94345)

DenisVieriu97 committed 1 year ago

22e1698c

[BE] Set PYTORCH_TEST_WITH_INDUCTOR only once (#94411)

malfet committed 1 year ago

82401c6a

fix: max_unpool3d buffer overflow (#94372)

Kristoff-starling committed 1 year ago

0bf78b57

Revert "[quant] Add quantize and dequantize operators to decomposition table (#93312)"

pytorchmergebot committed 1 year ago

3a5a7624

[CI] Add known ciflow labels to probot (#94368)

malfet committed 1 year ago

6ac0198c

[BE] Doc Update: Python 3.7 is past End of Life (#94314)

ZainRizvi committed 1 year ago

c0fe5fb9

[functorch][nn] Refactor NN stateless APIs by swapping module tensors (#92536)

XuehaiPan committed 1 year ago

b8de1cf0

[quant] Add quantize and dequantize operators to decomposition table (#93312)

jerryzh168 committed 1 year ago

3fd46a2f

[submodule] update libfmt to tag 9.1.0 (#93219)

cyyever committed 1 year ago

a405c699

[dynamo] fix general attr on tensor for user-provided attributes (#94332)

davidberard98 committed 1 year ago

8ba87fa5

Revert "sparse compressed tensor validation without syncs for low-(batch)dim tensors. (#94048)"

pytorchmergebot committed 1 year ago

f65a2064

[MPS] Fix the crash with hardswish_backward() (#94342)

razarmehr committed 1 year ago

e44cd942

Re-enable cudagraphs for benchmark scripts (#94192)

jansel committed 1 year ago

eb1aca16

[decompositions] GRU decompositon with and without packed sequence (#91466)

lezcano committed 1 year ago

fe0e28ab

[decompositions] LSTM with packed input (#91465)

lezcano committed 1 year ago

5a7c1b78

[decompositions] add decomposition for RNN with packed sequence (#91281)

lezcano committed 1 year ago

bef61225

[decompositions] add LSTM decomp (#91124)

lezcano committed 1 year ago

e5f6e1f6

[expanded weights] add RNN support via decomp (#91807)

lezcano committed 1 year ago

20d01d2d

Newer Older