wangwei created SINGA-300:
-----------------------------

             Summary: Add residual networks for imagenet classification
                 Key: SINGA-300
                 URL: https://issues.apache.org/jira/browse/SINGA-300
             Project: Singa
          Issue Type: New Feature
            Reporter: wangwei


This ticket would add a couple of residual networks trained over ImageNet. 
Pre-trained parameters from torch are converted into python dictionary and 
dumped via cPickle.

In particular, the following residual networks are implemented
1. Original residual network https://arxiv.org/abs/1512.03385
2. Pre-activation residual network https://arxiv.org/abs/1603.05027
3. Wide residual network http://arxiv.org/abs/1605.07146

reference: 
https://github.com/facebook/fb.resnet.torch
https://github.com/szagoruyko/wide-residual-networks




--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Reply via email to