strisunshinewentingwang/pytorch

Pull Requests Commits

Improve test_namedtensor.py with named tensor equality check (#24106)

zou3519 committed 5 years ago

90f3f9d9

Add name propagation for at::alias, add tensor.set_names (#24105)

zou3519 committed 5 years ago

1108fa1a

Optimizing out the division in the fusion

z-a-f committed 5 years ago

bb4f380f

Updating submodules

svcscm committed 5 years ago

b028cc75

Updating submodules

svcscm committed 5 years ago

a671609a

Make all at::Tensor in-place methods const (#23945)

pbelevich committed 5 years ago

b9a006f9

Move TensorOptions to ATen/core (#22020)

zou3519 committed 5 years ago

bde73860

Add interfaces in lr_scheduler.pyi (#23934)

tczhangzhi committed 5 years ago

a5f69761

serialize modules as classes

suo committed 5 years ago

77c08aa4

Revert D16552212: [jit] fix py-compat fbcode lint warnings

suo committed 5 years ago

5ec1c293

OpenCV 4 compatibility fix for caffe2/video (#24143)

leomao committed 5 years ago

6be24be9

send flake8 to stderr (#24100)

suo committed 5 years ago

365b3ff5

Replace Module::copy_into with Module::clone. (#24068)

Mikhail Zolotukhin committed 5 years ago

d3f6d588

is_quantized support in JIT

James Reed committed 5 years ago

98439938

JIT Serialization of nnq.Linear (#24048)

James Reed committed 5 years ago

a45dafc6

State dict serialization of nnq.Linear (#24047)

James Reed committed 5 years ago

ca2010cf

Simplified nnq.Linear class (#24046)

James Reed committed 5 years ago

442b3512

separate input shapes to reduce default execution time (#24136)

mingzhe09088 committed 5 years ago

b453fd99

support grad and data attribute for tensor in script

wanchaol committed 5 years ago

ca7e2a78

add initial support for sparse tensors

wanchaol committed 5 years ago

2790439b

Adding dequantize_val and requantize_val

z-a-f committed 5 years ago

83a594cf

Fix builtin function reference (#24056)

davidriazati committed 5 years ago

be5eb678

c10 dispatcher stores autograd kernels (#23666)

smessmer committed 5 years ago

211bafc2

Allow kernels that don't have a boxed version (#23665)

smessmer committed 5 years ago

296f218a

Unboxed kernels in c10 (#23447)

smessmer committed 5 years ago

9dbee5f8

Open up AliasAnalysisKind for any ops (#23834)

smessmer committed 5 years ago

352032c9

support dict augment assign in script

wanchaol committed 5 years ago

390bfd52

Added type annotations to unpooling layers (#24101)

Chillee committed 5 years ago

c79a07e3

Moves clamp from autodiff cpp to symbolic script (#23927)

Mike Ruberry committed 5 years ago

c21a7740

Port addcdiv operator from the TH code to Aten

ifedan committed 5 years ago

1d3d92e7

Newer Older