strisunshinewentingwang/pytorch

Pull Requests Commits

Test binary jobs for Windows magma upgrade

peterjc123 committed 3 years ago

2c4b2234

Revert D25172740: [pytorch][PR] [CUDA graphs] Make CUDAGeneratorImpl capturable

Mike Ruberry committed 3 years ago

98fddc1f

Add CUDA kernel checks to fbcode/caffe2/caffe2/sgd (#48347)

Kurtis David committed 3 years ago

0066b941

Change the type hints of "pooling.py". (#48412)

Xianguang-Zhou committed 3 years ago

736e8965

Revert D25222215: [quant][fix] Add bias once in conv_fused

malfet committed 3 years ago

c81f2d9a

Fix incorrect warnings in ParameterList/Dict (#48315)

CedricPicron committed 3 years ago

dc7ab46d

Add LazyConvXd and LazyConvTransposeXd (#47350)

asi1024 committed 3 years ago

492683bd

[vulkan] convolution old prepacking via cpu-shader (#48330)

IvanKobzarev committed 3 years ago

ccd20e99

[ONNX] Cast Gather index to Long if needed (#47653)

jiafatom committed 3 years ago

55fc0e9e

[ONNX] Support nonzero(*, as_tuple=True) export (#47421)

BowenBao committed 3 years ago

02e58aab

[caffe2][torch] Clean up unused variable 'device' (#48600)

jdonald committed 3 years ago

acd4fca3

Testing: Improve interaction between dtypes and ops decorators (#48426)

peterbell10 committed 3 years ago

9500e8a0

[quant][fix] Add bias once in conv_fused (#48593)

jerryzh168 committed 3 years ago

d2e42986

add aot_based_partition (#48336)

scottxu0730 committed 3 years ago

7a59a1b5

[Gradient Compression] Add a random generator to PowerSGD state for initializing low-rank matrix Q (#48507)

Yi Wang committed 3 years ago

ddb65949

[PyTorch][JIT] Parameter passing & std::map API usage pass on ProfilingRecord::instrumentGraph (#47960)

swolchok committed 3 years ago

61936cb1

caffe2: refactor context to allow being typed (#48340)

d4l3k committed 3 years ago

dc7d8a88

[te] Fix comparison ops on booleans (#48384)

bertmaher committed 3 years ago

adb4fd3f

[TensorExpr] Add a envvar to disable LLVM backend and use IR Eval instead. (#48355)

Mikhail Zolotukhin committed 3 years ago

d9f5ac08

[TensorExpr] IREval: fix div for Half dtype. (#48354)

Mikhail Zolotukhin committed 3 years ago

a6f0c3c4

Disable fast sigmoid since it causes divergence (#48623)

bertmaher committed 3 years ago

671a9592

Fused8BitRowwiseQuantizedToFloat operator support (#48407)

Frank Seide committed 3 years ago

29f0e1e2

remove unused params in scalar_tensor_static (#48550)

Basil Hosmer committed 3 years ago

c3bb3827

[vulkan] Fix Addmm prepacking to persist after GPU flush (#48313)

SS-JIA committed 3 years ago

ea0ffbb6

Update Windows CI to CUDA 11.1, cuDNN 8.0.5 (#48469)

peterjc123 committed 3 years ago

5b6b1495

[ROCm] update debug flags (#46717)

jeffdaily committed 3 years ago

7f869dca

Fix multiple spelling and grammar mistakes (#48592)

ProGamerGov committed 3 years ago

d6ddd78e

[CUDA graphs] Make CUDAGeneratorImpl capturable (#47989)

mcarilli committed 3 years ago

2200e722

[caffe2] Register BlackBoxPredictor AllocationArenaPool as CPUCachingAllocator (#48161)

Hao Lu committed 3 years ago

4976208e

[dper] supress excessive msg (#48404)

xw285cornell committed 3 years ago

d386d332

Older