strisunshinewentingwang/pytorch

Pull Requests Commits

Change arg dtype from float to double in LPPool and nn/utils/clip_grad.h (#29584)

Will Feng committed 4 years ago

433baf1b

Use c10::variant-based enums for SmoothL1Loss module and functional

Will Feng committed 4 years ago

65bfcde0

Use c10::variant-based enums for F::grid_sample

Will Feng committed 4 years ago

57eab22c

Make all non-input arguments to functionals part of its options (#29404)

Will Feng committed 4 years ago

9f879ef5

Design doc for distributed autograd. (#29175)

pritamdamania committed 4 years ago

c3b2c2e3

Consolidate the places that find pybind11 include dirs (#29659)

bddppq committed 4 years ago

b0c245d5

Remove observer module after insert_quant_dequant (#29622)

jerryzh168 committed 4 years ago

fd8f74e6

Cleanup special handling of Containers, allowing custom forwards (#28988)

Elias Ellison committed 4 years ago

fbe90b65

Make nn.Sequential iterable (#28987)

Elias Ellison committed 4 years ago

3175f554

updated name_inference doc for cumsum and cumprod (#29453)

anjali411 committed 4 years ago

eeb7199c

Fixing data type in quantized pool benchmarking

z-a-f committed 4 years ago

9bb0e283

Skip copy_same_type_transpose_ for quantized tensor (#29609)

jerryzh168 committed 4 years ago

82913a26

Benchmarking per channel quantization

z-a-f committed 4 years ago

3b43cfde

Quantized interpolation benchmarks

z-a-f committed 4 years ago

5db361bd

Separate out pytorch_jni into pytorch_jni_jit and pytorch_jni_common. (#29617)

xcheng16 committed 4 years ago

9c9c361f

Benchmarking quantized methods (#29625)

z-a-f committed 4 years ago

f95e8ea1

Suppress implicit int-float conversion warning in ROCm build (#29604)

bddppq committed 4 years ago

f111f1b1

Fix jit tracing namedtuple (#29477)

bddppq committed 4 years ago

949d6ae1

Complex support on GPU for dynamic casting (#29612)

zasdfgbnm committed 4 years ago

450949c7

Enable test_distributed for ROCm but only with nccl backend

jithunnair-amd committed 4 years ago

7073ee20

quantized topk benchmarking

z-a-f committed 4 years ago

3b452ca4

Quantized unary ops benchmarking (mostly template)

z-a-f committed 4 years ago

a0d4d506

Updating submodules

svcscm committed 4 years ago

e651494d

change `drop_on_export` warning category (#29610)

suo committed 4 years ago

2fb40596

Convert conv_prepack to conv2d_prepack and conv_unpack to conv2d_unpack (#29529)

jianyuh committed 4 years ago

bbff06ee

Add Support for ONNX scripting Interpolate with missing shape (#29489)

lara-hdr committed 4 years ago

2acca09e

Updating submodules

svcscm committed 4 years ago

8db06732

Apply the latest master docker images(jni.h in every image) (#29588)

IvanKobzarev committed 4 years ago

0c9e6727

Add ONNX Export Support for torch.scalar_tensor (#28713)

lara-hdr committed 4 years ago

8b53515b

Disable android gradle jobs (#29606)

IvanKobzarev committed 4 years ago

5249c43d

Newer Older