[
https://issues.apache.org/jira/browse/HAMA-961?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14612806#comment-14612806
]
Edward J. Yoon edited comment on HAMA-961 at 7/3/15 4:08 AM:
-------------------------------------------------------------
Also, I propose we create new module as a ANN framework on top of Hama like
below:
{code}
+-------------------------+-------------------------+----------------------------+
| Core package | Graph package | ANN package
|
| (pure BSP) | (Google's Pregel) | (Google's DistBelief)
|
+-------------------------+-------------------------+----------------------------+
|
|
| Apache Hama
|
|
|
|
+-------------------------------------------------------+
| | YARN
|
+------------------------+-------------------------------------------------------+
| Hadoop HDFS
|
+--------------------------------------------------------------------------------+
{code}
was (Author: udanax):
Also, I propose we create new module as a ANN framework on top of Hama.
> Parameter Server for large scale MLP
> ------------------------------------
>
> Key: HAMA-961
> URL: https://issues.apache.org/jira/browse/HAMA-961
> Project: Hama
> Issue Type: Improvement
> Components: machine learning
> Affects Versions: 0.7.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.8.0
>
>
> I've recently started to review the MLP source codes closely, and I'm
> thinking about some improvement and API refactoring e.g., APIs for
> user-defined neuron and synapse models, data structure, ..., etc.
> This issue is one of them, and related to train large models. I'm considering
> distributed parameter server (http://parameterserver.org) for managing
> parameters.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
