[
https://issues.apache.org/jira/browse/SINGA-238?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15420273#comment-15420273
]
ASF subversion and git services commented on SINGA-238:
-------------------------------------------------------
Commit 5b332a4086ff32b0c3a298169c0befef78f003ca in incubator-singa's branch
refs/heads/dev from [~wangwei.cs]
[ https://git-wip-us.apache.org/repos/asf?p=incubator-singa.git;h=5b332a4 ]
SINGA-238 RBM on MNIST
Enable the training on GPU.
Fixed a bug from KernelSum() by removing it and implemented the
Sum(Tensor)->float function using Dot() (from blas).
> RBM on mnist
> -------------
>
> Key: SINGA-238
> URL: https://issues.apache.org/jira/browse/SINGA-238
> Project: Singa
> Issue Type: New Feature
> Reporter: ZHAOJING
>
> Implement RBM python version on mnist data set
> 1. The model is following: http://www.cs.toronto.edu/~hinton/science.pdf
> 2. This model is implemented using python tensors
> 3. Users should first download mnist.pkl.gz
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
