Net Pruning

[mtngld]

Hi,

Now that neural networks perform pretty good, it seems the next effort for deep learning is not only to achieve better accuracy but to also achieve smaller and faster nets, 

pruning seems to be a hot method in this field,

Is there any implementation of pruning in caffe? 

Thanks,

Matan