projects / platform / upstream / pytorch.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: c02e2ff) | patch
Implementation convolutionTranspose operator for mkl-dnn (#12866)
authorCheng,Penghui <penghui.cheng@intel.com>
Thu, 21 Feb 2019 00:54:51 +0000 (16:54 -0800)
committerFacebook Github Bot <facebook-github-bot@users.noreply.github.com>
Thu, 21 Feb 2019 01:26:10 +0000 (17:26 -0800)
commit376bb403793b19e74959267a5c8147d2aa054bc1
treefb7e4b75dabbe9b72af0f1264570b0a75b4b26d9tree | snapshot
parentc02e2ff0b0a67e5ee39081b82da1316369520f12commit | diff
Implementation convolutionTranspose operator for mkl-dnn (#12866)

Summary:
the speed-up of a single operation is up to 2-3X on BDW.
This PR depend on https://github.com/pytorch/pytorch/pull/14308
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12866

Differential Revision: D13936110

Pulled By: ezyang

fbshipit-source-id: 34e3c2ca982a41e8bf556e2aa0477c999fc939d3
caffe2/ideep/operators/conv_transpose_op.cc [new file with mode: 0644] blob
caffe2/ideep/operators/conv_transpose_unpool_base_op.h [new file with mode: 0644] blob
caffe2/ideep/operators/operator_fallback_ideep.cc diff | blob | history
caffe2/python/ideep/conv_transpose_test.py [new file with mode: 0644] blob
Domain: Machine Learning / ML Framework;