strisunshinewentingwang/pytorch

Pull Requests Commits

Serialization for nn.quantized.functional modules (#25220)

raghuramank100 committed 5 years ago

9d06a984

Add new qnnpack_add and qnnpack_maxpool op to C10 registry (#24103)

supriyar committed 5 years ago

5b4e0529

Fixing the enforcement of the zero_point

z-a-f committed 5 years ago

86a35d7b

disable deadline checking on test_adaptive_avg_pool2d

James Reed committed 5 years ago

3c3d95cf

Add USE_CUDNN check to AT_CUDNN_ENABLED definition (#25037)

pbelevich committed 5 years ago

2e224d62

Add libtorch android build with shared lib for 4 android abis (#25192)

IvanKobzarev committed 5 years ago

f82c4ce6

Implement a bunch of pickle serialization features that optimize for size. (#23759)

Owen Anderson committed 5 years ago

f8852c94

data -> data_ptr: upgrade the deprecated APIs (#25223)

jianyuh committed 5 years ago

3af758c0

Optimize LeftRight and either (#25133)

smessmer committed 5 years ago

a4fa1678

get rid of dynamic_cast in Quantizer (#25001)

ljk53 committed 5 years ago

9fd62436

generic overrideable convolution for backends (#23562)

Ailing Zhang committed 5 years ago

858493d1

Reducing the test size for adaptive avg pool (#25195)

z-a-f committed 5 years ago

ac862e6d

Handle empty qconfig for functional Modules (#25215)

raghuramank100 committed 5 years ago

f5a3d592

Implementation of cyclical learning rate (#23914)

Andrew Li committed 5 years ago

3779893d

Specify width for st.floats in hypothesis_utils.tensor (#25188)

jerryzh168 committed 5 years ago

c351a68f

Disable flaky test_adaptive_avg_pool2d test. (#25249)

ezyang committed 5 years ago

44a7879b

Fix scriptability for Observer (#25219)

raghuramank100 committed 5 years ago

c142dbf8

Move the detection of cuDNN to FindCUDNN.cmake (#24938)

xuhdev committed 5 years ago

92750acb

Implement name inference for torch.dot (#24474)

zou3519 committed 5 years ago

2f4f6c25

Revert D15901930: Add interface declarations to JIT

ezyang committed 5 years ago

9340b155

Add myself as a CODEOWNER for better discoverability (#25231)

apaszke committed 5 years ago

1f57b8b7

Update mapping dictionary to support functionalmodules and pooling operations (#25216)

raghuramank100 committed 5 years ago

f622ec80

Move test QAT tests to double precision to ensure numerics match (#25211)

raghuramank100 committed 5 years ago

4d2bf0b5

Remove InsertQuantDeQuantNode (#25000)

jerryzh168 committed 5 years ago

b15d9149

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

Newer Older