pytorch/caffe2/python/ideep
Cheng,Penghui 1717ea1da0 Implementation of ChannelShuffle Op for MKLDNN (#15106)
Summary:
the speed-up of a single operation is up to 3X .
Pull Request resolved: https://github.com/pytorch/pytorch/pull/15106

Differential Revision: D13429596

Pulled By: bddppq

fbshipit-source-id: f8d987cafeac9bef9c3daf7e43ede8c6a4ee2ce5
2018-12-12 20:25:12 -08:00
..
__init__.py Add __init__.py so files get picked up on install (#14898) 2018-12-07 13:40:23 -08:00
channel_shuffle_op_test.py Implementation of ChannelShuffle Op for MKLDNN (#15106) 2018-12-12 20:25:12 -08:00
concat_split_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
conv_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
convfusion_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
copy_op_test.py build with mkl-dnn by default (#13303) 2018-11-08 11:18:27 -08:00
dropout_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
elementwise_sum_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
fc_op_test.py Implementation of WeightedSum op for mkl-dnn and fix FC op output shape issue. 2018-12-07 12:35:19 -08:00
LRN_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
moment_sgd_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
operator_fallback_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
pool_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
relu_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
sigmoid_op_test.py Add sigmoid op based on MKL-DNN 2018-11-19 22:56:35 -08:00
softmax_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
spatial_bn_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
squeeze_op_test.py seperate mkl, mklml, and mkldnn (#12170) 2018-10-29 10:52:55 -07:00
test_ideep_net.py [Caffe2] Fix of the performance issue of IDEEP (#7503) 2018-05-11 13:43:41 -07:00
transform_ideep_net.py [Caffe2] Ideep net optmization passes (#7514) 2018-05-11 23:50:18 -07:00
weightedsum_op_test.py Implementation of WeightedSum op for mkl-dnn and fix FC op output shape issue. 2018-12-07 12:35:19 -08:00