Motivated by the fact that most of the information relevant to the prediction of target tokens is drawn from the source sentence $S=s_1, \ldots, s_S$, we propose truncating the target-side window used for computing self-attention by making an $N$-gram assumption. Experiments on WMT EnDe and EnFr data sets show that the $N$-gram masked self-attention model loses very little in BLEU score for $N$ values in the range $4, \ldots, 8$, depending on the task.

中文翻译：

---

更快的 Transformer 解码：N-gram Masked Self-Attention

由于与目标标记预测相关的大部分信息都来自源语句 $S=s_1, \ldots, s_S$，我们建议截断用于计算自注意力的目标侧窗口$N$-gram 假设。在 WMT EnDe 和 EnFr 数据集上的实验表明，$N$-gram 掩蔽自注意力模型在 $N$ 值范围内 $4, \ldots, 8$ 的 BLEU 分数中损失很小，具体取决于任务。