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

Posted to dev@madlib.apache.org by GitBox <gi...@apache.org> on 2019/09/23 23:55:18 UTC

[GitHub] [madlib] khannaekta opened a new pull request #444: DL: Improve performance for madlib_keras_predict()

khannaekta opened a new pull request #444: DL: Improve performance for madlib_keras_predict()
URL: https://github.com/apache/madlib/pull/444
 
 
   JIRA: MADLIB-1345
   
   Passing huge model weights as a param to `internal_keras_predict()` for
   each row in the table slows the performance of overall `madlib_keras_predict()`.
   With this PR, the model weights will be passed in as a param only for the
   very first row(min(ctid)) fetched on each segment and NULL for the rest.
   With this change, we see `~4x` performance boost in the execution time
   of the existing `madlib_kerase_predict()`.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services