strisunshinewentingwang/pytorch

Pull Requests Commits

Properly compare floats in gemm-block-sparse-microkernel-tester (#82947)

kit1980 committed 2 years ago

4d3f7df3

[Refactoring] making the code more Pythonic (#82929)

se4ml committed 2 years ago

51d12548

[MPS] And native bitwise_[and|or|xor] (#82307)

malfet committed 2 years ago

0377615b

Fix possible out-of-range error of 'as_nd' util function (#82393)

Xia-Weiwen committed 2 years ago

18db7f36

Remove flatbuffer types/headers from flatbuffer_loader.h (#82893)

dbort committed 2 years ago

1d56ea5e

Revert "Add DLPack support for XPU backend by mapping to kDLOneAPI in DLPack (#82867)"

pytorchmergebot committed 2 years ago

0c7ca2d9

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

pytorchmergebot committed 2 years ago

b170a52a

Fix FSDP device_id when CPU offloading (#82892)

rohan-varma committed 2 years ago

ab3c0399

[torchdynamo hash update] update the pinned torchdynamo hash (#82943)

pytorchmergebot committed 2 years ago

73889130

Add EnableTorchFunction (#82647)

ezyang committed 2 years ago

a61c9649

Improved legalize_graph pass in FX (#82874)

Chillee committed 2 years ago

51bbf632

Remove manual bindings for arange (#81380)

peterbell10 committed 2 years ago

4f255dbf

Remove manual bindings for linspace, logspace and full (#81378)

peterbell10 committed 2 years ago

adc5e7d3

Make python TensorOption signatures consistent with JIT schemas (#82241)

peterbell10 committed 2 years ago

2c2278a9

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

Newer Older