strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D16879132: Update mapping dictionary to support functionalmodules and pooling operations

raghuramank100 committed 5 years ago

a9fdc192

Revert D17053634: Move test QAT tests to double precision to ensure numerics match

raghuramank100 committed 5 years ago

978a964b

Integration tests for initial quantization graph mode (#24428)

jerryzh168 committed 5 years ago

a1bf4d7e

Revert D16923660: Support observer without any data calibration

raghuramank100 committed 5 years ago

77ee1f5f

Revert D16923651: Serialization for nn.quantized.functional modules

raghuramank100 committed 5 years ago

c3c36a5b

Revert D17059486: Fix scriptability for Observer

raghuramank100 committed 5 years ago

ff30201f

Fix scriptability for Observer (#25197)

raghuramank100 committed 5 years ago

85d1ebd2

Creates Torch-friendly Event class and adds Stream tracking to autograd (#25130)

Mike Ruberry committed 5 years ago

433fe47d

Implement name inference for addmv, addmv_, mv (#24471)

zou3519 committed 5 years ago

088201f9

Implement name inference for expand (#24469)

zou3519 committed 5 years ago

78fa8a8a

skip fstrings test if not py36 (#25184)

Elias Ellison committed 5 years ago

277cd748

Fix bugs in assignment to optionals (#25059)

zdevito committed 5 years ago

121839b2

Adding Scalar add/mul. (#24447)

z-a-f committed 5 years ago

3b3261ca

Adding return for the observer in the functional_modules.py

z-a-f committed 5 years ago

a3e6e82b

fix to loggin in AA

Krovatkin committed 5 years ago

c395f421

Implement name inference for mm, addmm (#24306)

zou3519 committed 5 years ago

0156d02b

Fix binary op name inference between unnamed and named tensors. (#24921)

zou3519 committed 5 years ago

6195aee2

Move test QAT tests to double precision to ensure numerics match (#25189)

raghuramank100 committed 5 years ago

5d6b3dfd

Serialization for nn.quantized.functional modules (#24924)

raghuramank100 committed 5 years ago

95a3ffc2

Support observer without any data calibration (#24923)

raghuramank100 committed 5 years ago

a5710e23

Update mapping dictionary to support functionalmodules and pooling operations (#24804)

raghuramank100 committed 5 years ago

794f63fe

Handle empty qconfig for functional Modules (#24803)

raghuramank100 committed 5 years ago

d7f6ac1d

Work around for bias quantization for conv and linear operators (#24789)

raghuramank100 committed 5 years ago

ea601d90

bind autograd.backward and tensor.backward in TorchScript (#23913)

wanchaol committed 5 years ago

969c918f

quant_fusion jit pass (#24427)

jerryzh168 committed 5 years ago

9a9ef21b

insert_quant_dequant jit pass (#24426)

jerryzh168 committed 5 years ago

105fbb9c

per channel quantization support (#25134)

dskhudia committed 5 years ago

d8062575

hyperparameter plugin (#23134)

lanpa committed 5 years ago

cd14518e

Add Python/C++ torch.nn API parity test harness (#23852)

Will Feng committed 5 years ago

1bf1970f

prevent generating caffe2::mkl for multiple times (#25167)

jasjuang committed 5 years ago

573b1cd2

Newer Older