论文标题
基于块的最近的邻居翻译
Chunk-based Nearest Neighbor Machine Translation
论文作者
论文摘要
通过检索增强生成的半参数模型，由于能够从示例数据存储中检索细粒度的信息，因此导致了语言建模和机器翻译的令人印象深刻的结果。 $ k $ nn-mt是最突出的方法之一，它通过从域特异性datastores \ citep {khandelwalwal2020 neareast}中检索令牌来表现出强大的域适应能力。但是，$ k $ nn-mt需要为每个生成的代币进行昂贵的检索操作，从而导致非常低的解码速度（比参数模型慢8倍）。在本文中，我们介绍了一个\ textit {基于块} $ k $ nn-mt模型，该模型从数据存储中检索了几块令牌，而不是单个令牌。我们提出了几种将检索到的块纳入生成过程的策略，并选择模型需要在数据存储中搜索邻居的步骤。在两种设置中进行机器翻译的实验，即静态和``即时''域的适应性，表明基于块的$ K $ NN-MT型号可导致显着的加速（最高4次），而翻译质量的下降只有很小的下降。
Semi-parametric models, which augment generation with retrieval, have led to impressive results in language modeling and machine translation, due to their ability to retrieve fine-grained information from a datastore of examples. One of the most prominent approaches, $k$NN-MT, exhibits strong domain adaptation capabilities by retrieving tokens from domain-specific datastores \citep{khandelwal2020nearest}. However, $k$NN-MT requires an expensive retrieval operation for every single generated token, leading to a very low decoding speed (around 8 times slower than a parametric model). In this paper, we introduce a \textit{chunk-based} $k$NN-MT model which retrieves chunks of tokens from the datastore, instead of a single token. We propose several strategies for incorporating the retrieved chunks into the generation process, and for selecting the steps at which the model needs to search for neighbors in the datastore. Experiments on machine translation in two settings, static and ``on-the-fly'' domain adaptation, show that the chunk-based $k$NN-MT model leads to significant speed-ups (up to 4 times) with only a small drop in translation quality.
