strisunshinewentingwang/pytorch

Pull Requests Commits

[JIT] disable masked fill (#50147)

Elias Ellison committed 3 years ago

6eee2a0a

Back out "Revert D25757721: [pytorch][PR] Run mypy on more test files" (#50142)

ezyang committed 3 years ago

3ce53988

Clean up type annotations in torch/nn/quantized/modules (#49941)

r-barnes committed 3 years ago

63808695

Implementation of torch::cuda::synchronize (#50072)

jbschlosser committed 3 years ago

7d9eb6c6

[Needs Review] Convert some files to Python3 (#49351)

r-barnes committed 3 years ago

e606e603

Clean up some type annotations in torch/testing/_internal (#50078)

r-barnes committed 3 years ago

efe0533a

Fix mypy type hint for AdaptiveAvgPool2,3d, AdaptiveMaxPool2,3d (#49963)

VinhLoiIT committed 3 years ago

74c055b2

Push anonymous namespace into codegen, not template (#49498)

ezyang committed 3 years ago

68a6e463

[PyTorch] IValue::toTensor can now return const Tensor& (#48868)

swolchok committed 3 years ago

480a7561

[PyTorch] Store Tensor explicitly in IValue (#48824)

swolchok committed 3 years ago

1b31e135

[PyTorch] Additional IValue tests (#49718)

swolchok committed 3 years ago

688992c7

Unused variables in neural net classes and functions (#50100)

alexhenrie committed 3 years ago

5f2ec629

Add support for converting sparse bool tensors to dense (#50019)

NathanHowell committed 3 years ago

c517e15d

Fix cl.exe detection in cpu/fused_kernel.cpp (#50085)

skyline75489 committed 3 years ago

2ac180a5

Set USE_RCCL cmake option (dependent on USE_NCCL) [REDUX] (#34683)

jithunnair-amd committed 3 years ago

45ec3582

drop a unneeded comma from cmakelist.txt (#50091)

cyyever committed 3 years ago

0ad6f066

Revert D25239967: [fx] Add matrix multiplication fusion pass

ngimel committed 3 years ago

ad7d208b

[PyTorch] Reapply D25546409: Use .sizes() isntead of .size() in cat_serial_kernel_impl (#49762)

swolchok committed 3 years ago

282552dd

Fix for possible RNG offset calculation bug in cuda vectorized dropout with VEC=2 (#50110)

mcarilli committed 3 years ago

57d489e4

[reland][quant][graphmode][fx] Standalone module support {input/output}_quantized_idxs (#49754) (#50058)

jerryzh168 committed 3 years ago

f6f0fde8

Fix mypy typing check for test_dataset (#50108)

ejguan committed 3 years ago

05358332

Remove cpu half and dead code from multinomial (#50063)

t-vi committed 3 years ago

def8aa54

[fx] Add matrix multiplication fusion pass (#50120)

Meghan Lele committed 3 years ago

9b7f3fa1

Clean up type annotations in caffe2/torch/nn/modules (#49957)

r-barnes committed 3 years ago

d80d38cf

[PyTorch] Reapply D25545777: Use .sizes() instead of .size() in _cat_out_cpu (#49761)

swolchok committed 3 years ago

75028f28

[PyTorch] Reapply D25544731: Avoid extra Tensor refcounting in _cat_out_cpu (#49760)

swolchok committed 3 years ago

574a15b6

Fix doc for vmap levels (#50099)

ejguan committed 3 years ago

5f875965

Kill AT_SKIP_BFLOAT16_IF_NOT_ROCM (#48810)

zasdfgbnm committed 3 years ago

70734f12

Support out argument in torch.fft ops (#49335)

peterbell10 committed 3 years ago

26391143

torch.flip and torch.flip{lr, ud}: Half support for CPU and BFloat16 support for CPU & CUDA (#49895)

kshitij12345 committed 3 years ago

5d93e2b8

Newer Older