eblearn / News: Recent posts

network templatization

The neural nets are now templatizable when they were fixed with double precision before. e.g.:
linear_module l;
is now declared:
linear_module<float> l;

Posted by Pierre Sermanet 2010-01-06