Guiding Non-Autoregressive Neural Machine Translation Decoding with Reordering Information
This addresses the problem of slow inference in neural machine translation for users needing faster translations, but it is incremental as it builds on existing NAT approaches.
The paper tackled the translation quality gap in non-autoregressive neural machine translation (NAT) by proposing ReorderNAT, a framework that models reordering information to narrow the decoding space, achieving better performance than existing NAT models and comparable quality to autoregressive models with significant speedup.
Non-autoregressive neural machine translation (NAT) generates each target word in parallel and has achieved promising inference acceleration. However, existing NAT models still have a big gap in translation quality compared to autoregressive neural machine translation models due to the enormous decoding space. To address this problem, we propose a novel NAT framework named ReorderNAT which explicitly models the reordering information in the decoding procedure. We further introduce deterministic and non-deterministic decoding strategies that utilize reordering information to narrow the decoding search space in our proposed ReorderNAT. Experimental results on various widely-used datasets show that our proposed model achieves better performance compared to existing NAT models, and even achieves comparable translation quality as autoregressive translation models with a significant speedup.