Retrieval-augmented generation (RAG) has recently emerged as a promising solution for incorporating up-to-date or domain-specific knowledge into large language models (LLMs) and improving LLM factuality, but is predominantly studied in English-only settings. In this work, we consider RAG in the multilingual setting (mRAG), i.e. with user queries and the datastore in 13 languages, and investigate which components and with which adjustments are needed to build a well-performing mRAG pipeline, that can be used as a strong baseline in future works. Our findings highlight that despite the availability of high-quality off-the-shelf multilingual retrievers and generators, task-specific prompt engineering is needed to enable generation in user languages. Moreover, current evaluation metrics need adjustments for multilingual setting, to account for variations in spelling named entities. The main limitations to be addressed in future works include frequent code-switching in non-Latin alphabet languages, occasional fluency errors, wrong reading of the provided documents, or irrelevant retrieval. We release the code for the resulting mRAG baseline pipeline at https://github.com/naver/bergen, Documentation: https://github.com/naver/bergen/blob/main/documentations/multilingual.md.

Retrieval-Augmented Generation allows to enhance Large Language Models with external knowledge. In response to the recent popularity of generative LLMs, many RAG approaches have been proposed, which involve an intricate number of different configurations such as evaluation datasets, collections, metrics, retrievers, and LLMs. Inconsistent benchmarking poses a major challenge in comparing approaches and understanding the impact of each component in the pipeline. In this work, we study best practices that lay the groundwork for a systematic evaluation of RAG and present BERGEN, an end-to-end library for reproducible research standardizing RAG experiments. In an extensive study focusing on QA, we benchmark different state-of-the-art retrievers, rerankers, and LLMs. Additionally, we analyze existing RAG metrics and datasets.

La recherche conversationnelle est une tâche qui vise à retrouver des documents à partir de la questioncourante de l’utilisateur ainsi que l’historique complet de la conversation. La plupart des méthodesantérieures sont basées sur une approche multi-étapes reposant sur une reformulation de la question.Cette étape de reformulation est critique, car elle peut conduire à un classement sous-optimal des do-cuments. D’autres approches ont essayé d’ordonner directement les documents, mais s’appuient pourla plupart sur un jeu de données contenant des pseudo-labels. Dans ce travail, nous proposons une tech-nique d’apprentissage à la fois “légère” et innovante pour un modèle contextualisé d’ordonnancementbasé sur SPLADE. En s’appuyant sur les représentations parcimonieuses de SPLADE, nous montronsque notre modèle, lorsqu’il est combiné avec le modèle de ré-ordonnancement T5Mono, obtient desrésultats qui sont compétitifs avec ceux obtenus par les participants des campagnes d’évaluation TRECCAsT 2020 et 2021. Le code source est disponible sur https://github.com/anonymous.