You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hama.apache.org by "Edward J. Yoon (JIRA)" <ji...@apache.org> on 2015/06/23 11:19:00 UTC
[jira] [Resolved] (HAMA-675) Deep Learning Computation Model
[ https://issues.apache.org/jira/browse/HAMA-675?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Edward J. Yoon resolved HAMA-675.
---------------------------------
Resolution: Duplicate
Assignee: Edward J. Yoon
I'm close this issue as a Duplicate. This will be addressed in HAMA-961.
I'm thinking use multi-thread or LocalBSPJobRunner for mini-batch, and http://parameterserver.org/. See https://docs.google.com/drawings/d/1cjz50sGbpnFp2oab30cZ5MNYsaD3PtaBRVsUWuLiglI/edit?usp=sharing
Regarding interface, gradient() computing method and fetch/push methods that communicates with PM server will be needed. Interface designing is not a big deal I think.
> Deep Learning Computation Model
> -------------------------------
>
> Key: HAMA-675
> URL: https://issues.apache.org/jira/browse/HAMA-675
> Project: Hama
> Issue Type: New Feature
> Components: machine learning
> Reporter: Thomas Jungblut
> Assignee: Edward J. Yoon
>
> Jeff Dean mentioned a computational model in this video: http://techtalks.tv/talks/57639/
> There they are using the same idea of the Pregel system, they are defining a upstream and a downstream computation function for a neuron (for cost and its gradient). Then you can roughly tell about how the framework should partition the neurons.
> All the messaging will be handled by the underlying messaging framework.
> Can we implement something equally?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)