[
https://issues.apache.org/jira/browse/SINGA-37?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14635162#comment-14635162
]
ASF subversion and git services commented on SINGA-37:
------------------------------------------------------
Commit 5bf1c92808c97dcc42df46504f3615c746ab368b in incubator-singa's branch
refs/heads/master from wang sheng
[ https://git-wip-us.apache.org/repos/asf?p=incubator-singa.git;h=5bf1c92 ]
SINGA-37 Enable users to set parameter sharing in model configuration
merged pull request #35
> Enable users to set parameter sharing in model configuration
> ------------------------------------------------------------
>
> Key: SINGA-37
> URL: https://issues.apache.org/jira/browse/SINGA-37
> Project: Singa
> Issue Type: New Feature
> Reporter: wangwei
>
> For some models (e.g., RBM, RNN), the parameters of some layers are shared
> (tied). To enable this feature, SINGA needs to provide a configuration field
> which indicates from whom to share the parameter.
> In specific, we can add one field in the ParamProto:
> {code}
> // the name of a Param who owns the Param and shares its values with this
> Param.
> optional string share_from;
> {code}
> The NeuralNet class with share the Params based on the share_from field.
> Params only share values, i.e., data field. Everyone has its own gradient,
> history, etc fields. SINGA will automatically average the gradients for the
> shared Params before sending the update request.
>
> This ticket will not consider the case in which the original (owner of a
> shared) Param object is partitioned due to model partitioning, e.g., the
> weight matrix is partitioned into two sub-matrix if the inner-product layer
> is partitioned on dimension 1.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
