New port: graphics/caffe

Caffe is a deep learning framework made with expression, speed, and
modularity in mind. It is developed by the Berkeley Vision and Learning
Center (BVLC) and by community contributors.

WWW: http://caffe.berkeleyvision.org/

PR: 217428
Submitted by: eric@camachat.org (maintainer)
Reviewed by: mat
Approved by: swills (mentor, implicit)
Differential Revision: https://reviews.freebsd.org/D10053