strisunshinewentingwang/pytorch

Pull Requests Commits

[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

[ONNX] Make einsum_helper in opset12 private (#82402)

msebindavis committed 2 years ago

561f1568

[fsdp] add ability to iterate through dataclasses in fsdp.utils (#82638)

edward-io committed 2 years ago

e7ff9d44

[ONNX] Fix `argmin` and `argmax` test cases (#79503)

qqaatw committed 2 years ago

6bdf89b0

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

voznesenskym committed 2 years ago

6d76b2c1

[ao] Changed ratio of channels needed for input-weight rec (#82795)

vspenubarthi committed 2 years ago

5ca098fe

[ao] Fix punctuation issue with Dynamic Static Report (#82794)

vspenubarthi committed 2 years ago

95c7fc39

[CI] Disable Libtorch Linux Bionic Cuda for Infra Flakiness (#82862)

zengk95 committed 2 years ago

7520f875

Skip TestNNAPI tests if QNNPACK is not supported (#82882)

Flamefire committed 2 years ago

cafdc52c

Make M1 tests green (#82213)

albanD committed 2 years ago

2255911f

Fix leak when create_graph and full backward hook registered (#82788)

soulitzer committed 2 years ago

1cafb102

Move OpInfo subclasses into opinfo.core (#82830)

peterbell10 committed 2 years ago

d8ae83ba

Move OpInfo class into new opinfo folder (#82540)

peterbell10 committed 2 years ago

4d405517

[Quant][fx] Remove dequant-quant around getitem (#82675)

andrewor14 committed 2 years ago

8f38f677

[vulkan] Pass correct output_padding_arg when creating conv2d contexts (#82861)

SS-JIA committed 2 years ago

18858765

Handle complex optimization in AdamW by treating complex numbers as 2D real numbers (#80280)

zaxtax committed 2 years ago

dcbe9ce2

[xla hash update] update the pinned xla hash (#82877)

pytorchmergebot committed 2 years ago

0f32d71b

[torchdynamo hash update] update the pinned torchdynamo hash (#82876)

pytorchmergebot committed 2 years ago

7b486667

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

chengjunlu committed 2 years ago

de0e0300

Re-land sym_numel (#82374) (#82726) (#82731) (#82855)

Krovatkin committed 2 years ago

bfebf254

Fix profiling with record_shapes=True and nested tensor (#82854)

ngimel committed 2 years ago

bdb0abb2

fix matching against MemcpyAsync (#82782)

davidchencsl committed 2 years ago

e47637aa

Migrate remaining pytorch code to use new flatbuffer_loader.h APIs (#82620)

dbort committed 2 years ago

6e712823

Newer Older