strisunshinewentingwang/pytorch

Pull Requests Commits

Compile THCTensorTopK per dtype. (#36074)

gchanan committed 4 years ago

2173746f

Fixing Potential TSAN issue with joining RPC helper threads (#36094)

osalpekar committed 4 years ago

7d1f0646

[WIP] Enable NNC's LLVM dependency in CI (#35564)

Owen Anderson committed 4 years ago

b8383b3d

[rpc] call threadPool.waitWorkComplete after listenerThread.join() to fix (#35394)

rohan-varma committed 4 years ago

2ef1ace8

qactivation_benchmarks: small bug fix (#35731)

vkuzo committed 4 years ago

cc789147

add more quantized activation benchmarks and input sizes (#35729)

vkuzo committed 4 years ago

6405f26a

add benchmark for quantized batchnorm (#35389)

vkuzo committed 4 years ago

b68c3827

[dt][caffe2] enable using smart exceptions in async nets (#34753)

d4l3k committed 4 years ago

8ef82fc2

[JIT] Fix fake_range() (#36083)

James Reed committed 4 years ago

3228939f

[ROCm] Enable BFloat16 type for add_out_sparse (#35978)

rohithkrn committed 4 years ago

3e402a59

Pin Sphinx to 2.4.4 (take 2), fix docs CIs (#36072)

zou3519 committed 4 years ago

cb385cb6

[JIT] optimize mutableType calls (#35474)

Elias Ellison committed 4 years ago

0475d7b0

[ROCm] Hotfix: Black list tensorexpr test set that has failures on ROCm (#36049)

iotamudelta committed 4 years ago

45fc881f

Strip newline when ingesting `version.txt` (#36002)

malfet committed 4 years ago

59ed0c5f

add type hints on recently added ops to make them scriptable (#35885)

vkuzo committed 4 years ago

4ef383d5

[ONNX] Added support for constant folding onnx::Add and onnx::Sub (#35869)

tianyou.gty committed 4 years ago

8dba98da

[TensorExpr] add more detail to malformed_input exceptions (#35891)

nickgg committed 4 years ago

d568c7d9

disable the test to stop breaking the builds (#36053)

lly-zero-one committed 4 years ago

82d58ed4

[ONNX] fix size for opset 11 (#35984)

neginraoof committed 4 years ago

e56ba848

[pytorch] Fix the extra_repr print message for float16 dynamic quantization (#36044)

jianyuh committed 4 years ago

8224398c

Disable tracing for Pytorch Mobile client (#36007)

iseeyuan committed 4 years ago

81c8ca1e

Temporary methods for real and imag values of complex tensors (#35879)

anjali411 committed 4 years ago

66d50060

Fix torch complex exp CPU implementation (#35532) (#35715)

aayn committed 4 years ago

b3cdec88

Rename boxing/unboxing files and utilities (#35411)

smessmer committed 4 years ago

7ee88d61

[caffe2] tune prefetch distance

jspark1105 committed 4 years ago

8a6173ed

Revert D20802884: [Shape Inference] Set new shape according to precedence of dimType over previous value

smessmer committed 4 years ago

7b2b17f7

Add DICT_CONSTRUCT and NAMED_TUPLE_CONSTRUCT to lite interpreter (#36015)

iseeyuan committed 4 years ago

82087ee7

Use `std::abs` instead of `abs` in lbfgs.cpp (#35974)

Will Feng (FAIAR) committed 4 years ago

5fab1bf3

[Shape Inference] Set new shape according to precedence of dimType over previous value (#35910)

ChunliF committed 4 years ago

e3e2dd77

Support for XNNPACK max pooling operator. (#35354)

Ashkan Aliabadi committed 4 years ago

b7f4b6a6

Newer Older