strisunshinewentingwang/pytorch

Pull Requests Commits

Fix lint (#47095)

James Reed committed 3 years ago

2e2dc587

Enable python code coverage on windows (#44548)

guyang3532 committed 3 years ago

f5477e37

Fix median bug on discontigous tensors (#46917)

heitorschueroff committed 3 years ago

ddeacf15

[WIP] Move torch.fx into its own target (#46658)

James Reed committed 3 years ago

9bc8f071

[Transposed Conv]add ConvTranspose3d with FBGEMM as backend (#46608)

YazhiGao committed 3 years ago

71901554

Pin destination memory for cuda_tensor.to("cpu", non_blocking=True) (#46878)

mcarilli committed 3 years ago

3c643d11

fix calculation of number of elements to not overflow (#46997)

ngimel committed 3 years ago

e17b8dea

Replace -f with -x for pytest tests. (#46967)

pritamdamania committed 3 years ago

78de12f5

[ONNX] bump CI ort to 1.5.2 rel for stability (#46595)

BowenBao committed 3 years ago

a4caa3f5

Delete TypeDefault.h and TypeDerived.h codegen entirely. (#47002)

ezyang committed 3 years ago

843cab3f

Delete TypeDefault call code generation logic in VariableType (#47000)

ezyang committed 3 years ago

c689b4d4

Delete SchemaRegister.cpp, make flag operate on TypeDefault.cpp (#46991)

ezyang committed 3 years ago

41f8641f

Desugar missing dispatch field into singleton Math entry (#46970)

ezyang committed 3 years ago

54d83296

Some miscellaneous cleanup in codegen (#46940)

ezyang committed 3 years ago

87e86fa8

Delete Vulkan from code generator. (#46938)

ezyang committed 3 years ago

dc6f723c

view_as_real doesn't work for all backends since it relies on strides. (#47018)

Ailing Zhang committed 3 years ago

156c08b0

enable PE everywhere but mobile (#47001)

Krovatkin committed 3 years ago

71c0133e

reland fast TypeMeta/ScalarType conversion (#45544)

Basil Hosmer committed 3 years ago

377a09c8

[ONNX] Enable NoneType inputs to export API (#45792)

shubhambhokare1 committed 3 years ago

1ea14e30

fix_combine_two_partition_size (#47053)

scottxu0730 committed 3 years ago

c556d455

[ONNX] Support nd mask index in opset >= 11 (#45252)

BowenBao committed 3 years ago

129b4122

[fix] torch.nn.functional.embedding -> padding_idx behavior (#46714)

kshitij12345 committed 3 years ago

1d233d7d

Bump up NCCL to v2.8 (#46742)

mingzhe09088 committed 3 years ago

3e499e49

Fix DDP issue where parameters share same grad_accumulator (#46755)

rohan-varma committed 3 years ago

d850b5c9

[RFC] Decouple fast pass functions (#46469)

iseeyuan committed 3 years ago

68057153

implement NumPy-like functionality column_stack, row_stack (#46313)

RockingJavaBean committed 3 years ago

74d730c0

Correctly mark unannotated NamedTuple field to be inferred TensorType (#46969)

tmanlaibaatar committed 3 years ago

fee585b5

Show Flake8 errors in GitHub CI again (#46990)

samestep committed 3 years ago

1e275bc1

Implement torch.igamma (#46183)

mfkasim1 committed 3 years ago

6eaa324c

[caffe2/FC DNNLOWP] Shrink Y_int32_ vector capacity when appropriate

blezsan committed 3 years ago

dd95bf65

Newer Older