Commit Graph

11 Commits (25d44d40acfca5ed92dbc57fbaa2b01367a66f99)

Author SHA1 Message Date
Tao Luo 20120d9c97
Merge pull request #14608 from jczaja/prv-conv2d-transpose-mkldnn
6 years ago
qingqing01 731d45a39a
Enable BatchNorm to use global mean and variane during training (#14630)
6 years ago
Jacek Czaja fb24690a58 - conv2d transpose MKL-DNN
6 years ago
Krzysztof Binias fb4b4f8d57 Refactor code
7 years ago
Krzysztof Binias 50d3e6e96b Reusing primitives for forward Batch Norm operator
7 years ago
mozga-intel b8a04c2fa1 Duplicated code was moved to common function
7 years ago
mozga-intel 3b128337a1 The mkldnn batch norm supports other data format
7 years ago
pzelazko-intel 9a15c92317 bnorm+relu fuse for mkldnn (inference) (#11434)
7 years ago
Luo Tao 5aac910b89 add url of cuda9.0_cudnn7_avx_mkl library
7 years ago
mozga-intel 7d5643562f MKLDNN layout: Support for batch norm operator
7 years ago
Tomasz Patejko 4a497b826d MKLDNN implementation of batch normalization (#9904)
7 years ago