strisunshinewentingwang/pytorch

Pull Requests Commits

Add CPU-only (for now) philox randn impl to aten (#82383)

voznesenskym committed 2 years ago

6e86e9bc

Revert "Vectorize cpu tensor conversions (#80905)"

pytorchmergebot committed 2 years ago

dd3f6023

Revert "Allow ufunc OpInfos to have no reference (#82348)"

pytorchmergebot committed 2 years ago

814c19b2

Revert "Use UnaryUfuncInfo for type conversion functions (#82349)"

pytorchmergebot committed 2 years ago

785b13af

Use UnaryUfuncInfo for type conversion functions (#82349)

peterbell10 committed 2 years ago

13e78a2c

Allow ufunc OpInfos to have no reference (#82348)

peterbell10 committed 2 years ago

566d7343

Vectorize cpu tensor conversions (#80905)

peterbell10 committed 2 years ago

948cc542

Increase size limit on calling CublasLt in addmm by 32x (#82922)

erichan1 committed 2 years ago

5b51849b

Fix buick-build-and-test (#82935)

malfet committed 2 years ago

082e6506

[docs] fix inconsistent default `rcond` value (#82887)

raj-magesh committed 2 years ago

8d091ca4

Parameterize TestGenericProxyTensor on tracing_mode (#82746)

ezyang committed 2 years ago

d2472449

Reorganize test_proxy_tensor.py per tracing mode (#82739)

ezyang committed 2 years ago

b361f703

[xla hash update] update the pinned xla hash (#82931)

pytorchmergebot committed 2 years ago

9ea0a203

[torchdynamo hash update] update the pinned torchdynamo hash (#82930)

pytorchmergebot committed 2 years ago

e66da90f

pytest - print message re skip info (#82901)

clee2000 committed 2 years ago

8a6c104c

[ao] Updated ModelReportVisualizer per-channel line plot (#82918)

vspenubarthi committed 2 years ago

86437b86

[ao] Fix per-channel histogram visualization in ModelReportVisualizer (#82917)

vspenubarthi committed 2 years ago

da5272ef

Fix BSR->Dense Batched Bug (#82120)

amjames committed 2 years ago

eebcb911

Add support for `select` of batch dims for all sparse compressed formats. (#82119)

amjames committed 2 years ago

0e0dfaa0

Update CUDA compiler matrix (#82860)

malfet committed 2 years ago

737fa85d

Prevent ref cycle creation in inner hook (#82776)

albanD committed 2 years ago

7dd795cb

Better error message for torch.library (#82904)

SherlockNoMad committed 2 years ago

3d5e49e9

[ONNX] Type utility for converting among JIT, torch and ONNX data types (#81953)

justinchuby committed 2 years ago

6ddf4c6f

Handle complex optimization in Adamax by treating complex numbers as 2D real numbers (#80319)

zaxtax committed 2 years ago

c54d18db

[PyTorch] Refactor GlobalStateManager as a templated singleton class (#82152)

louisfeng committed 2 years ago

a4264da6

Revert "Add CPU-only (for now) philox randn impl to aten (#82383)"

pytorchmergebot committed 2 years ago

22e865a7

Revert "Implement `mul(dense, sparse), mul(sparse, dense)` for sparse COO tensors. (#81556)"

pytorchmergebot committed 2 years ago

45291c7e

Revert "Implement and extend `mul(sparse, sparse)` to work with 0-dim arguments on either side. (#82717)"

pytorchmergebot committed 2 years ago

796fba02

Add pre-compiled headers to one of the CI runners (#77351)

peterbell10 committed 2 years ago

efd8e083

[ao] Added ModelReportVisualizer info to README for ModelReport (#82796)

vspenubarthi committed 2 years ago

2e74b51a

Newer Older