Tensorflow implement of paper: Sequence to Sequence: Video to Text - View it on GitHub