ignite-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (IGNITE-6783) Create common mechanism for group training.
Date Sat, 30 Dec 2017 22:03:00 GMT

    [ https://issues.apache.org/jira/browse/IGNITE-6783?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16306968#comment-16306968

ASF GitHub Bot commented on IGNITE-6783:

Github user asfgit closed the pull request at:


> Create common mechanism for group training.
> -------------------------------------------
>                 Key: IGNITE-6783
>                 URL: https://issues.apache.org/jira/browse/IGNITE-6783
>             Project: Ignite
>          Issue Type: Task
>          Components: ml
>            Reporter: Artem Malykh
>            Assignee: Artem Malykh
> In distributed ML it is a common task to train several models in parallel with ability
to communicate with each other during training. Simple example of this case is training of
neural network with SGD on different chunks of data located on several nodes. In such training
we do the following in a loop: on each node we do one or several SGD steps then send gradient
on central node which averages gradients from each of worker nodes and send back the averaged
gradient. There is a pattern in this procedure which can be applied to other ML algos and
it could be useful to extract this pattern.

This message was sent by Atlassian JIRA

View raw message