strisunshinewentingwang/pytorch

Pull Requests Commits

Add doc previewing instructions (#31905)

davidriazati committed 4 years ago

8c59d482

remove THConv code which never be used (#31879)

XiaobingSuper committed 4 years ago

dedd16b4

Move cauchy to Aten(CPU) (#31824)

XiaobingSuper committed 4 years ago

9a3cb1e8

Add op bitwise_or (#31559)

XiaobingSuper committed 4 years ago

9ba6a768

Port softplus activation to Aten(CPU+CUDA) (#30504)

XiaobingSuper committed 4 years ago

4f9d2f74

Combine all the user inputs together and convert them to fp16 (#31898)

Yinghai Lu committed 4 years ago

d2fdf140

Use simd version for fp16 conversions (#31897)

Yinghai Lu committed 4 years ago

8b4feff0

Ensure the original grad_mode is restored during backward (#31884)

albanD committed 4 years ago

1314f7f4

temporary fix for jit test backward compatibility issues

albanD committed 4 years ago

c299cb05

docker hub image info (#31923)

Mingbo Wan committed 4 years ago

462bfc7f

Revert D19298735: Emit warning from deprecated torch function signatures

ezyang committed 4 years ago

5dfcfeeb

Quantized H Tangent function (#31031)

z-a-f committed 4 years ago

620060cb

Avoid reference invalidation in cuda SpectralOps' plan_caches (#31861)

peterbell10 committed 4 years ago

54777b1e

Revert D19290954: Implement backend-agnostic rpc._wait_all_workers() utility

mrshenli committed 4 years ago

7f723cbd

Add device debug info to CUDA build (#31929)

zasdfgbnm committed 4 years ago

c66ca74f

Remove C++11 compatibility from c10::optional (#30919)

smessmer committed 4 years ago

f0072b3a

Fix c10::util::get_fully_qualified_type_name for MSVC (#31313)

smessmer committed 4 years ago

f67851d6

Remove memory ordering from LeftRight (#31026)

smessmer committed 4 years ago

2a294aac

Fix -Wundef warning in conversions.h

jdonald committed 4 years ago

84dfa96f

Add more tests to the autograd wrt view and inplace (#31147)

albanD committed 4 years ago

ee817012

Implement backend-agnostic rpc._wait_all_workers() utility (#31888)

xush6528 committed 4 years ago

66647038

Rename TORCH_DCHECK to TORCH_INTERNAL_ASSERT_DEBUG_ONLY (#31917)

ezyang committed 4 years ago

9116f02b

Make c10::util::get_fully_qualified_type_name() backwards compatible with clang 4 (#31351)

smessmer committed 4 years ago

ab60cca4

constexpr typeid improvements (#31312)

smessmer committed 4 years ago

0dca9c30

Remove c++14-conditional constexpr (#30916)

smessmer committed 4 years ago

c21f8997

Fix IValue.isList

dreiss committed 4 years ago

4daa3ded

Properly return data from non-contiguous tensors in Java

dreiss committed 4 years ago

1b4d3d57

Support tensors with a storage offset in Java (#31584)

dreiss committed 4 years ago

2d6a2c89

Support tensors with empty shape in Java

dreiss committed 4 years ago

6d1fa829

Better error for `torch::jit::load`ing a eager file (#31709)

davidriazati committed 4 years ago

3c07eb33

Newer Older