strisunshinewentingwang/pytorch

Pull Requests Commits

Bumps supported file format in anticipate of torch.div changes (#39529)

Mike Ruberry committed 4 years ago

7d56ef27

Added Operator_Schema's for missing FakeFP16 Operators (#39363)

venkatacrc committed 4 years ago

17aebe90

[quant][graphmode] Test for another type of ops in insert_observer for if (#39380)

jerryzh168 committed 4 years ago

f94a171e

Remove useless copy on zip file load (#36362)

davidriazati committed 4 years ago

da8191a9

misc updates to fake fp16 tests (#39405)

venkatacrc committed 4 years ago

ed12df64

Do not raise decorator (#39532)

malfet committed 4 years ago

2a513a6a

Reduce time spent per guard by comparing TensorType with Tensor (#39098)

Ailing Zhang committed 4 years ago

b861daf0

Add/fix typing annotations to some functions (#39075)

malfet committed 4 years ago

8811e4d0

deepcopy() of Objects should call __g/setstate__ (#39500)

James Reed committed 4 years ago

da2f8c9f

[ONNX] Fix type casting for reduce ops (#38829)

neginraoof committed 4 years ago

4e5af8d1

Upgrade lint. (#39483)

ezyang committed 4 years ago

da2004e1

Fix overflow issues when unpacking large numbers (#39140)

RockingJavaBean committed 4 years ago

fe684679

[JIT] fix broadcasting lists of ints (#39481)

Elias Ellison committed 4 years ago

49b69b2a

.circleci: Move binary builds into their own workflow (#39379)

seemethere committed 4 years ago

7676aa79

Selective enabling of xnnpack based max_pool2d in ceil_mode. (#39447)

kimishpatel committed 4 years ago

eb5e0376

Move some of the definitions in LegacyNNDefinitions.cpp closer to sites (#37531)

ezyang committed 4 years ago

76803581

Add arcosh, arcsinh and arctanh to unary ops (#38388)

krshrimali committed 4 years ago

335e4a1e

Restore docs coverage test via sphinx (#39331)

mattip committed 4 years ago

ada2652c

Fix lint (#39527)

ezyang committed 4 years ago

b4aceb38

Remove cuda init patch (#39222)

jithunnair-amd committed 4 years ago

af91df68

fix build table for ppc64le (#39475)

dncliss committed 4 years ago

ac252677

Add SymbolicShape and replace all uses of VaryingShape<ShapeSymbol> with it (#38544)

gmagogsfm committed 4 years ago

002b19da

Clean up thrust::complex from rsqrt (#39294)

zasdfgbnm committed 4 years ago

11a60b99

Fix lint (#39517)

Mike Ruberry committed 4 years ago

92c67767

Implement timeout support for RRefs (#38590)

rohan-varma committed 4 years ago

8b2bb02e

[pytorch] move tracing logic to a separate dispatch backend (#38467)

ljk53 committed 4 years ago

72b0447f

Optimize GroupNorm on CPU (#28203)

xiaomengy committed 4 years ago

03eca384

Revert D21652452: [pytorch][PR] Fix for num_threads==1 in OpenMP "parallel for"

oulgen committed 4 years ago

4a0a38c1

Add rpc.async_function decorator for TorchScript functions (#39267)

mrshenli committed 4 years ago

67cea74d

Implement rad2deg, deg2rad (#38852)

aayn committed 4 years ago

0829cadc

Newer Older