Multi-Task Learning with Shared Encoder for Non-Autoregressive Machine Translation
This work addresses the accuracy-speed trade-off in machine translation for researchers and practitioners, offering an incremental enhancement by combining multi-task learning with existing knowledge distillation methods.
The paper tackles the problem of inferior translation accuracy in Non-Autoregressive Machine Translation (NAT) models by proposing a multi-task learning approach that shares an encoder with Autoregressive Translation (AT) models, resulting in significant improvements over baseline NAT models on datasets like WMT14 English-German and WMT16 English-Romanian.
Non-Autoregressive machine Translation (NAT) models have demonstrated significant inference speedup but suffer from inferior translation accuracy. The common practice to tackle the problem is transferring the Autoregressive machine Translation (AT) knowledge to NAT models, e.g., with knowledge distillation. In this work, we hypothesize and empirically verify that AT and NAT encoders capture different linguistic properties of source sentences. Therefore, we propose to adopt Multi-Task learning to transfer the AT knowledge to NAT models through encoder sharing. Specifically, we take the AT model as an auxiliary task to enhance NAT model performance. Experimental results on WMT14 English-German and WMT16 English-Romanian datasets show that the proposed Multi-Task NAT achieves significant improvements over the baseline NAT models. Furthermore, the performance on large-scale WMT19 and WMT20 English-German datasets confirm the consistency of our proposed method. In addition, experimental results demonstrate that our Multi-Task NAT is complementary to knowledge distillation, the standard knowledge transfer method for NAT.