Commit Graph

752 Commits (bee95fc8917e09f61ba46586a94d2b9003cddf13)

Author SHA1 Message Date
Luo Tao c53d214205 fix compiler warning from MKLDNNLayer and so on
9 years ago
tensor-tang 672c968176 remove gpu code when backward mkldnn_softmax
9 years ago
tensor-tang 2c6ac6297d refine code and remove min clip
9 years ago
tensor-tang 7483087c8c enable mkldnn_softmax
9 years ago
tensor-tang bea39f6314 add compare simple net of mkldnn and cpu
9 years ago
tensor-tang ed27a3be78 add some log info
9 years ago
Tao Luo 4400284685 Merge pull request #4224 from tensor-tang/act
9 years ago
tensor-tang 4ec3a77a43 should run resetBwd before bwdAct
9 years ago
tensor-tang fa722385de refine test_MKLDNN and skip memory copy for relu
9 years ago
Yiqun Liu 8deda323cd Merge pull request #4197 from luotao1/refine_layer
9 years ago
Luo Tao 69c613fb5a refine Layer.cpp for some CostLayer
9 years ago
Yiqun Liu a9202e89cf Merge pull request #3764 from Xreki/build_ios
9 years ago
Tao Luo c15ac20257 Merge pull request #4199 from tensor-tang/mkldnn_act
9 years ago
tensor-tang 9d692e3bcc add gtest for MKLDNN activation and pass them
9 years ago
Luo Tao 70141a73c4 remove unused XXXLayer.h in Layer.cpp
9 years ago
tensor-tang 24f13b1a59 add MKLDNN activation files
9 years ago
Yuan Gao 17bc593ca2 Update DetectionOutputLayer.cpp
9 years ago
Yuan Gao d3deb6faba Update DetectionOutputLayer.cpp
9 years ago
gaoyuan 71b3fbb18a Fix a ssd bug
9 years ago
tensor-tang 49d4b39f28 fix typo and remove some unused code
9 years ago
tensor-tang 37268dbd33 update using exclusive padding for avg-pool
9 years ago
tensor-tang cc28fb4bb3 Merge remote-tracking branch 'upstream/develop' into mkldnn_pool
9 years ago
Luo Tao cc7a7557a2 set CudnnPooling don't include padded values in default
9 years ago
Yu Yang b012013a58 Revert "Do not invoke GPU method when use_gpu=false"
9 years ago
tensor-tang 9d108a21a7 add unit test for mkldnn_pool and pass them
9 years ago
Yu Yang 17e1aa8d43 Do not invoke GPU method when use_gpu=false
9 years ago
tensor-tang 2a98cba2d2 enable mkldnn_pool forward and backward
9 years ago
hedaoyuan b1bca06667 Refine the ExpandConvLayer.
9 years ago
hedaoyuan e4c340e4e7 Bug fix.
9 years ago
hedaoyuan 88b31422b3 Delete ExpandConvBaseLayer.cpp and ExpandConvBaseLayer.h
9 years ago
hedaoyuan 3616ef7739 Remove ExpandConvBaseLayer.
9 years ago
hedaoyuan b3a50d53c9 Remove useless code.
9 years ago
hedaoyuan 2a110af309 Use Matrix::addBias.
9 years ago
tensor-tang d3a86c673b init mkldnn_pool files
9 years ago
tensor-tang b0d9b68a5f unify functions of mkldnn_fc and refine comments
9 years ago
Tao Luo 654344b94c Merge pull request #4032 from tensor-tang/mkldnn-conv
9 years ago
tensor-tang f2317b67f7 separate resetFwd and resetBwd to some sub functions
9 years ago
Liu Yiqun 9e524fce1b Remove the generation of executable, such as paddle_trainer, when setting WITH_C_API.
9 years ago
tensor-tang 66fdbd0cee add some comment and simplify some code
9 years ago
chengduoZH e76eb8534d fix SwitchOrderLayer forward
9 years ago
tensor-tang c39b771a53 add test cases for mkldnn_conv
9 years ago
tensor-tang f3a23b6840 add MKLDNNConvLayer
9 years ago
Tao Luo 171fee2c9e Merge pull request #4011 from tensor-tang/refine
9 years ago
tensor-tang 7f7fa32528 use CPU_DEVICE instead of magic number
9 years ago
tensor-tang 4f0869bef3 add some necessary params for reset functions
9 years ago
chengduo d131152b28 Merge pull request #4021 from chengduoZH/fix_DeConv3D_switch_imgSize_and_output_
9 years ago
chengduoZH e9a1f2ad8b fix DeConv3D switch(imgSize*_, output*_)
9 years ago
tensor-tang f3bb7b99dd refine MKLDNNTester add UpdateCallback for test
9 years ago
tensor-tang 94ea8ee0e5 refine MKLDNNLayer logical: move forward and backward to MKLDNNLayer and remove copyOutputInfoToOtherDevice
9 years ago
tensor-tang f40d5f580d remove syncOutputGrad, rename syncInputValue to updateInputData
9 years ago