通过 Token Pooling 以最小的性能影响减少多向量检索的足迹
Reducing the Footprint of Multi-Vector Retrieval with Minimal Performance Impact via Token Pooling
AI总结:
本文提出一种基于聚类的 token pooling 方法,通过在索引期间合并向量,大幅减少 ColBERT 等多向量检索模型的存储和内存占用,且对检索性能影响极小,无需修改模型架构。
摘要:
在过去的几年中,以 ColBERT 为首的多向量检索方法已成为神经信息检索(Neural IR)中日益流行的方法。通过在 token 级别而非文档级别存储表示,这些方法展现出了非常强的检索性能,特别是在域外设置中。然而,存储大量相关向量所需的存储和内存要求仍然是一个重要的缺点,阻碍了其实际应用。在本文中,我们引入了一种简单的基于聚类的 token pooling 方法,以大幅减少需要存储的向量数量。该方法可以将 ColBERT 索引的空间和内存占用减少 50%,而几乎没有检索性能下降。该方法还允许进一步减少,将向量数量减少 66% 至 75%,在绝大多数数据集上性能下降保持在 5% 以下。重要的是,这种方法不需要架构更改或查询时处理,可以在索引期间作为简单的即插即用组件与任何类似 ColBERT 的模型一起使用。
英文摘要:
Over the last few years, multi-vector retrieval methods, spearheaded by ColBERT, have become an increasingly popular approach to Neural IR. By storing representations at the token level rather than at the document level, these methods have demonstrated very strong retrieval performance, especially in out-of-domain settings. However, the storage and memory requirements necessary to store the large number of associated vectors remain an important drawback, hindering practical adoption. In this paper, we introduce a simple clustering-based token pooling approach to aggressively reduce the number of vectors that need to be stored. This method can reduce the space & memory footprint of ColBERT indexes by 50% with virtually no retrieval performance degradation. This method also allows for further reductions, reducing the vector count by 66%-to-75% , with degradation remaining below 5% on a vast majority of datasets. Importantly, this approach requires no architectural change nor query-time processing, and can be used as a simple drop-in during indexation with any ColBERT-like model.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。