You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@madlib.apache.org by "Frank McQuillan (Jira)" <ji...@apache.org> on 2020/04/01 18:44:00 UTC

[jira] [Closed] (MADLIB-1414) DL: predict fails with OOM in gpdb6

     [ https://issues.apache.org/jira/browse/MADLIB-1414?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Frank McQuillan closed MADLIB-1414.
-----------------------------------
    Resolution: Fixed

https://github.com/apache/madlib/pull/489

> DL: predict fails with OOM in gpdb6
> -----------------------------------
>
>                 Key: MADLIB-1414
>                 URL: https://issues.apache.org/jira/browse/MADLIB-1414
>             Project: Apache MADlib
>          Issue Type: Bug
>          Components: Deep Learning
>            Reporter: Nikhil Kak
>            Priority: Major
>             Fix For: v1.17
>
>
> madlib keras predict fails with the following error if run with places10 (or any other model whose compiled weights are around 600MB or more in size)
> {code}
> ERROR: spiexceptions.InternalError: invalid memory alloc request size 1189758503 (plpy_elog.c:121)
> {code}
> The root cause is the same as as https://issues.apache.org/jira/projects/MADLIB/issues/MADLIB-1405



--
This message was sent by Atlassian Jira
(v8.3.4#803005)