[ https://issues.apache.org/jira/browse/SPARK-16365?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15376522#comment-15376522 ]
Sean Owen commented on SPARK-16365: ----------------------------------- This might be better on user@ or dev@. I don't know if this ends up in a concrete change. There are already a bunch of ways to deploy models. I'd rather Spark not invent another one; at least reuse one if that adds value. I think we've had several half-finished efforts around supporting technologies like PMML supports, a custom serialization, this sort-of separation of half the MLlib code. I'm concerned about the focus here vis-a-vis maturity of solutions outside Spark. Mostly thinking of JPMML + its Spark support here. > Ideas for moving "mllib-local" forward > -------------------------------------- > > Key: SPARK-16365 > URL: https://issues.apache.org/jira/browse/SPARK-16365 > Project: Spark > Issue Type: Brainstorming > Components: ML > Reporter: Nick Pentreath > > Since SPARK-13944 is all done, we should all think about what the "next > steps" might be for {{mllib-local}}. E.g., it could be "improve Spark's > linear algebra", or "investigate how we will implement local models/pipelines > in Spark", etc. > This ticket is for comments, ideas, brainstormings and PoCs. The separation > of linalg into a standalone project turned out to be significantly more > complex than originally expected. So I vote we devote sufficient discussion > and time to planning out the next move :) -- This message was sent by Atlassian JIRA (v6.3.4#6332) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org