strisunshinewentingwang/pytorch

Pull Requests Commits

ONNX export: propagate node metadata across passes (#45256)

tbennun committed 2 years ago

9d454929

[ci] try setting MAX_JOBS on windows builds to reduce OOMs (#66986)

suo committed 2 years ago

a33f341c

[SR] Fix bug in FuseListUnpackV2 (#67021)

Mike Iovine committed 2 years ago

53cf7e84

Splitter: Transformer_encoder (#66952)

FusionPower committed 2 years ago

a7ec4b53

Unsqueeze bug fix (#66889)

FusionPower committed 2 years ago

d73b88b4

Fix bug [#66780]: wrong input to torch.is_floating_point (#66783)

pearu committed 2 years ago

23321ba7

[skip ci] Set test owner for test_dispatch.py (#66840)

janeyx99 committed 2 years ago

13b85998

[qnnpack] Remove conv_utils.h (#66605)

John Shen committed 2 years ago

8cbdf49d

[skip ci] Set test owner for named tensor tests (#66849)

janeyx99 committed 2 years ago

960e3216

[skip ci] Set test owner for cpp-extensions tests (#66837)

janeyx99 committed 2 years ago

f5c5ab28

[Rocm]Reduce severity of detected possible memory leak from assertion to warning (#65973)

arindamroy-eng committed 2 years ago

32e79099

[ROCm] update Magma for 4.3 release (#65203)

jaglinux committed 2 years ago

70a5113e

Add torch.nn.init.uniform_ operator to ShardedTensor. (#63997)

bowangbj committed 2 years ago

b6df043f

[nnc] Use a descriptive name for fused kernels when profiling (#66990)

bertmaher committed 2 years ago

bdb889ac

[PyTorchEdge] Make aten function common to aten and torch_common (#66663)

pavithranrao committed 2 years ago

8beabffa

[quant][graphmode][fx] Add support for single linear and conv2d (#66950)

jerryzh168 committed 2 years ago

f8f04d54

[quant][fx][graphmode] Adding a new convert function that produces reference pattern by default (#66925)

jerryzh168 committed 2 years ago

a89851a0

[SmartCompose][OnDevice]fix function name bug in mobile export & Script to convert mobile model (#66915)

gardenia22 committed 2 years ago

db416589

[Static Runtime] Add FuseListUnpackV2 (#66509)

Mike Iovine committed 2 years ago

ab1e4eac

Add support for cat in output stitching (#66098)

Elias Ellison committed 2 years ago

17889ad2

Add support for multi output nodes in partial eval graph stitching (#66097)

Elias Ellison committed 2 years ago

2dd23ebf

Add Handling of Cat in Shape Analysis (#65575)

Elias Ellison committed 2 years ago

0196b984

Add x + 0 optimization (#65574)

Elias Ellison committed 2 years ago

eaba976d

Fix bug preventing optimization from firing (#65573)

Elias Ellison committed 2 years ago

b059f035

[JIT] Add partial evaluation graph stitching logic (#65377)

Elias Ellison committed 2 years ago

63b41e1f

[JIT][Easy] Shape cleanups (#65148)

Elias Ellison committed 2 years ago

4ad6c144

Avoid inlining error reporting in checked_convert (#66721)

andrewor committed 2 years ago

e046386b

[Static Runtime] Fix a bug in aten::index (#66940)

d1jang committed 2 years ago

18bbc4c2

[PyTorch][1/N] Basic implementation of ShardedEmbedding using ShardedTensor. (#66604)

fduwjj committed 2 years ago

08cb31a0

Fix attr_to_scope's key in `torch/utils/tensorboard/_pytorch_graph.py` (#65692)

liulixinkerry committed 2 years ago

25723997

Newer Older