An Empirical Study of Adequate Vision Span for Attention-Based Neural Machine Translation

Open in new window