strisunshinewentingwang/pytorch

Pull Requests Commits

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

Make RRef leak detection always print a warning log (#31922)

xush6528 committed 4 years ago

a730920a

Revert D18838848: disable __torch_function__ overides for operators in torch.functional

Karl Ostmo committed 4 years ago

227d1a43

Run a non-quiet submodule update to prevent timeouts on Circle CI (#31900)

Bram Wasti committed 4 years ago

8a0503b3

For torch::from_blob() add clue when memory is non-owned. (#31222)

jjlilley committed 4 years ago

114562cf

disable __torch_function__ overides for operators in torch.functional (#30839)

ngoldbaum committed 4 years ago

ca72df06

named tensor max pooling support

lixinyu committed 4 years ago

bb279c5c

Fix tracing for modules with List[Tensor] as output (#31343)

volkhin committed 4 years ago

3a2757c6

Raise an error if torch.cat is given `out` as one of the input tensors (#30577)

peterbell10 committed 4 years ago

74d69e29

Restructure docs organization and naming (#31849)

Jessica Lin committed 4 years ago

c888473b

Integrate async mode for autograd engine with distributed autograd. (#31508)

pritamdamania committed 4 years ago

bf8e1c07

Newer Older