You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@mxnet.apache.org by "Bojian Zheng (JIRA)" <ji...@apache.org> on 2019/02/20 04:17:00 UTC

[jira] [Created] (MXNET-1326) Performance Improvement of SequenceMask and SequenceReverse Operator

Bojian Zheng created MXNET-1326:
-----------------------------------

             Summary: Performance Improvement of SequenceMask and SequenceReverse Operator
                 Key: MXNET-1326
                 URL: https://issues.apache.org/jira/browse/MXNET-1326
             Project: Apache MXNet
          Issue Type: Improvement
          Components: Apache MXNet Backend
            Reporter: Bojian Zheng


`SequenceReverse` and `SequenceMask` operators are very frequently used in sequence processing tasks. Current implementations of `SequenceReverse` and `SequenceMask`, however, are far from efficient when running on the GPU. This patch contains the fix the resolve the inefficiency of `SequenceReverse` and `SequenceMask`.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@mxnet.apache.org
For additional commands, e-mail: issues-help@mxnet.apache.org