systemml-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mike Dusenberry (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SYSTEMML-1483) Add Deconvolution layer in nn library and Caffe2DML
Date Wed, 03 May 2017 23:56:04 GMT

    [ https://issues.apache.org/jira/browse/SYSTEMML-1483?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15995919#comment-15995919
] 

Mike Dusenberry commented on SYSTEMML-1483:
-------------------------------------------

Interesting, definitely want to figure out the root cause here.  I hadn't run into this problem
when I prototyped it before.  Is it possible that you have the filter and input swapped by
accident?  In that case, a {{conv2d_transpose}} would give a 3x3 output when given a 2x2 *filter*
and 3x3 *input*, which could explain the issue.

> Add Deconvolution layer in nn library and Caffe2DML
> ---------------------------------------------------
>
>                 Key: SYSTEMML-1483
>                 URL: https://issues.apache.org/jira/browse/SYSTEMML-1483
>             Project: SystemML
>          Issue Type: Sub-task
>            Reporter: Niketan Pansare
>
> http://caffe.berkeleyvision.org/tutorial/layers/deconvolution.html
> [~mwdusenb@us.ibm.com] [~prithvi_r_s] 



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message