[
https://issues.apache.org/jira/browse/MADLIB-1330?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16854970#comment-16854970
]
Ekta Khanna commented on MADLIB-1330:
-------------------------------------
Merged as part of PR: [https://github.com/apache/madlib/pull/377]
> DL: Do not compile model for predict and add GPU support
> --------------------------------------------------------
>
> Key: MADLIB-1330
> URL: https://issues.apache.org/jira/browse/MADLIB-1330
> Project: Apache MADlib
> Issue Type: Improvement
> Components: Deep Learning
> Reporter: Nandish Jayaram
> Assignee: Nandish Jayaram
> Priority: Major
> Fix For: v1.16
>
>
> In internal_keras_predict, we compile the model for every row. Starting from
> keras 1.0.3, keras predict does not need a compiled model. This will make our
> current predict code(which does not use SD) easier and hopefully slightly
> faster even with CPU. See more details here
> https://github.com/keras-team/keras/issues/3074
> Acceptance:
> 1. Validate that removing the compile code does not impact the accuracy of
> the results
> 2. Add new optional param named use_gpu in predict interface.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
