spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Joseph K. Bradley (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SPARK-12685) word2vec trainWordsCount gets overflow
Date Fri, 08 Jan 2016 19:44:40 GMT

    [ https://issues.apache.org/jira/browse/SPARK-12685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15089814#comment-15089814
] 

Joseph K. Bradley commented on SPARK-12685:
-------------------------------------------

This affects earlier versions too.  Would you mind checking to see which ones, and also how
hard it'd be to backport to them?  Thanks!

> word2vec trainWordsCount gets overflow
> --------------------------------------
>
>                 Key: SPARK-12685
>                 URL: https://issues.apache.org/jira/browse/SPARK-12685
>             Project: Spark
>          Issue Type: Bug
>          Components: MLlib
>    Affects Versions: 1.6.0
>            Reporter: yuhao yang
>            Assignee: yuhao yang
>            Priority: Minor
>
> the log of word2vec  reports 
> trainWordsCount = -785727483
> during computation over a large dataset.
> I'll also add vocabsize to the log.
> Update the priority as it will affects the computation process.
> alpha =
> learningRate * (1 - numPartitions * wordCount.toDouble / (trainWordsCount + 1))



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

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


Mime
View raw message