Fetching the paper…

Training Deeper Neural Machine Translation Models with Transparent Attention · Around