strisunshinewentingwang/pytorch

Pull Requests Commits

Remove insert_observers pass (#24999)

jerryzh168 committed 5 years ago

fbb88f5d

Remove deprecated graph mode quantization tests (#24998)

jerryzh168 committed 5 years ago

0b60f5c0

add import for test_quantizer.py (#25222)

jerryzh168 committed 5 years ago

ab022938

Initial commit for android torchvision utils (#25185)

IvanKobzarev committed 5 years ago

9e27cf61

add to Tensor symmetric methods getDataAsIntArray, getDataAsByteArray (#25183)

IvanKobzarev committed 5 years ago

c0334015

skip tests if fbgemm is not supported for test_quantizer.py (#25209)

jerryzh168 committed 5 years ago

c2e03839

Add interface declarations to JIT (#21972)

zdevito committed 5 years ago

4b22cf6b

Simplify NamedType

zdevito committed 5 years ago

6e42580d

Revert D16852280: Work around for bias quantization for conv and linear operators

raghuramank100 committed 5 years ago

26a438d4

Revert D16879133: Handle empty qconfig for functional Modules

raghuramank100 committed 5 years ago

17f69eff

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

Newer Older