UDAPDR: Unsupervised Domain Adaptation via LLM Prompting and Distillation of Rerankers

Jon Saad-Falcon, Omar Khattab, Keshav Santhanam, Radu Florian, Martin Franz, Salim Roukos, Avirup Sil, Md Arafat Sultan, Christopher Potts

Submission Type: Regular Long Paper

Submission Track: Information Retrieval and Text Mining

Submission Track 2: Efficient Methods for NLP

Keywords: Natural Language Processing, Information Retrieval, Domain Adaptation

TL;DR: We present UDAPDR, a self-supervised approach employs large language models to adapt multi-vector retrieval models to new domains.

Abstract: Many information retrieval tasks require large labeled datasets for fine-tuning. However, such datasets are often unavailable, and their utility for real-world applications can diminish quickly due to domain shifts. To address this challenge, we develop and motivate a method for using large language models (LLMs) to generate large numbers of synthetic queries cheaply. The method begins by generating a small number of synthetic queries using an expensive LLM. After that, a much less expensive one is used to create large numbers of synthetic queries, which are used to fine-tune a family of reranker models. These rerankers are then distilled into a single efficient retriever for use in the target domain. We show that this technique boosts zero-shot accuracy in long-tail domains and achieves substantially lower latency than standard reranking methods.

Submission Number: 297