Helping the Weak Makes You Strong: Simple Multi-Task Learning Improves Non-Autoregressive Translators
This addresses the accuracy gap for NAR translation models, which is crucial for efficient machine translation applications, though it is an incremental improvement over existing methods.
The paper tackles the problem of non-autoregressive (NAR) neural machine translation models having lower accuracy due to conditional independence assumptions, by proposing a simple multi-task learning framework that introduces weak autoregressive decoders to provide more informative learning signals. Experiments on WMT and IWSLT datasets show consistent accuracy improvements across multiple NAR baselines without extra decoding overhead.
Recently, non-autoregressive (NAR) neural machine translation models have received increasing attention due to their efficient parallel decoding. However, the probabilistic framework of NAR models necessitates conditional independence assumption on target sequences, falling short of characterizing human language data. This drawback results in less informative learning signals for NAR models under conventional MLE training, thereby yielding unsatisfactory accuracy compared to their autoregressive (AR) counterparts. In this paper, we propose a simple and model-agnostic multi-task learning framework to provide more informative learning signals. During training stage, we introduce a set of sufficiently weak AR decoders that solely rely on the information provided by NAR decoder to make prediction, forcing the NAR decoder to become stronger or else it will be unable to support its weak AR partners. Experiments on WMT and IWSLT datasets show that our approach can consistently improve accuracy of multiple NAR baselines without adding any additional decoding overhead.