Non-autoregressive Translation with Layer-Wise Prediction and Deep Supervision

How do we perform efficient inference while retaining high translation quality? Existing neural machine translation models, such as Transformer, achieve high performance, but they decode words one by one, which is inefficient. Recent non-autoregressive translation models speed up the inference, but their quality is still inferior. In this work, we propose DSLP, a highly efficient and high-performance model for machine translation. The key insight is to train a nonautoregressive Transformer with Deep Supervision and feed additional Layer-wise Predictions. We conducted extensive experiments on four translation tasks (both directions of . Results show that our approach consistently improves the BLEU scores compared with respective base models. Specifically, our best variant outperforms the autoregressive model on three translation tasks, while being 14.8 times more efficient in inference. 1 * Work partially done during an internship at ByteDance AI Lab. 1 Our code, training/evaluation scripts, and output are available at https://github.com/chenyangh/DSLP .