Non-autoregressive translation models (NAT) have achieved impressive
inference speedup. A potential issue of the existing NAT algorithms, however,
is that the decoding is conducted in parallel, without directly considering
previous context. In this paper, we propose an imitation learning framework for
non-autoregressive machine translation, which still enjoys the fast translation
speed but gives comparable translation performance compared to its
auto-regressive counterpart. We conduct experiments on the IWSLT16, WMT14 and
WMT16 datasets. Our proposed model achieves a significant speedup over the
autoregressive models, while keeping the translation quality comparable to the
autoregressive models. By sampling sentence length in parallel at inference
time, we achieve the performance of 31.85 BLEU on WMT16 Ro$\rightarrow$En and
30.68 BLEU on IWSLT16 En$\rightarrow$De.

本文提出了基于仿真学习的非自回归机器翻译框架，该框架仍然具有快速翻译速度但与其自回归对应方法相比具有可比较的翻译性能。通过对 IWSLT16、WMT14 和 WMT16 数据集进行实验，发现我们提出的模型在保持翻译质量可比的同时，比自回归模型有更快的翻译速度。在推断时并行采样句子长度，WMT16 Ro→En 的 BLEU 值达到 31.85，而 IWSLT16 En→De 的 BLEU 值为 30.68。