Non-autoregressive Streaming Transformer for Simultaneous Translation
Zhengrui Ma, Shaolei Zhang, Shoutao Guo, Chenze Shao, Min Zhang, Yang Feng
Abstract
Simultaneous machine translation (SiMT) models are trained to strike a balance between latency and translation quality. However, training these models to achieve high quality while maintaining low latency often leads to a tendency for aggressive anticipation. We argue that such issue stems from the autoregressive architecture upon which most existing SiMT models are built. To address those issues, we propose non-autoregressive streaming Transformer (NAST) which comprises a unidirectional encoder and a non-autoregressive decoder with intra-chunk parallelism. We enable NAST to generate the blank token or repetitive tokens to adjust its READ/WRITE strategy flexibly, and train it to maximize the non-monotonic latent alignment with an alignment-based latency loss. Experiments on various SiMT benchmarks demonstrate that NAST outperforms previous strong autoregressive SiMT baselines.
BibTeX
@inproceedings{
ma2023nonautoregressive,
title={Non-autoregressive Streaming Transformer for Simultaneous Translation},
author={Zhengrui Ma and Shaolei Zhang and Shoutao Guo and Chenze Shao and Min Zhang and Yang Feng},
booktitle={The 2023 Conference on Empirical Methods in Natural Language Processing},
year={2023},
url={https://openreview.net/forum?id=gzRBs4gIbz}
}