A Unified Model and Document Representation for On-Device Retrieval-Augmented Generation

TL;DR

This paper introduces a unified model for on-device retrieval-augmented generation that compresses context and uses shared representations to reduce memory and storage needs while maintaining performance.

Contribution

It proposes the first unified model that combines retrieval and context compression with shared representations, optimizing on-device RAG systems.

Findings

Model uses 1/10 of the context size of traditional RAG.

Matches traditional RAG performance without increasing storage.

Reduces disk and memory usage through shared representations.

Abstract

Traditional Retrieval-Augmented Generation (RAG) approaches generally assume that retrieval and generation occur on powerful servers removed from the end user. While this reduces local hardware constraints, it introduces significant drawbacks: privacy concerns regarding data access, recurring maintenance and storage costs, increased latency, and the necessity of an internet connection. On-device RAG addresses these challenges by executing the entire pipeline locally, making it ideal for querying sensitive personal information such as financial documents, contact details, and medical history. However, on-device deployment necessitates a delicate balance between limited memory and disk space. Specifically, the context size provided to the generative model must be restricted to manage KV cache and attention memory usage, while the size of stored embeddings must be minimized to preserve disk space. In this work, we propose a unified model that compresses the RAG context and utilizes the same representations for retrieval. This approach minimizes disk utilization compared to using separate representations, while significantly reducing the context size required for generation. With an average of 1/10 of the context, our model matches the performance of a traditional RAG reader without increasing storage requirements compared to a multi-vector retrieval model. This approach represents the first model to unify retrieval and context compression using a shared model and representation. We believe this work will inspire further consolidation of distinct models to optimize on-device performance.