2018/08/26 by Chunqi Wang, Ji Zhang, Wang, Chunqi +3 · 12 citations
Computer Science · Mathematics · #Algorithm #Artificial intelligence #Autoregressive model #Computation and Language (cs.CL) #Computer science #Data compression #Decoding methods #Econometrics #FOS: Computer and information sciences #Lossless compression #Machine translation #Mathematics #Multimodal Machine Learning Applications #Natural Language Processing Techniques #Parallel computing #Speech recognition #Speedup #Topic Modeling #Transformer #Translation (biology) #Voltage #cs.CL
paper · pdf · doi:10.48550/arxiv.1808.08583
published in arXiv (Cornell University) (Cornell University) · EMNLP 2018
openalex publication_date 2018/08/26 · arxiv created 2018/10/27 · arxiv updated 2018/10/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/08
Existing approaches to neural machine translation are typically autoregressive models. While these models attain state-of-the-art translation quality, they are suffering from low parallelizability and thus slow at decoding long sequences. In this paper, we propose a novel model for fast sequence generation --- the semi-autoregressive Transformer (SAT). The SAT keeps the autoregressive property in global but relieves in local and thus is able to produce multiple successive words in parallel at each time step. Experiments conducted on English-German and Chinese-English translation tasks show that the SAT achieves a good balance between translation quality and decoding speed. On WMT'14 English-German translation, the SAT achieves 5.58× speedup while maintains 88% translation quality, significantly better than the previous non-autoregressive methods. When produces two words at each time step, the SAT is almost lossless (only 1% degeneration in BLEU score).