PrimeQA: The Prime Repository for State-of-the-Art Multilingual Question Answering Research and Development

TL;DR

PrimeQA is an open-source, comprehensive toolkit designed to democratize and facilitate research, development, and replication of state-of-the-art multilingual question answering systems.

Contribution

It introduces PRIMEQA, a unified, end-to-end QA repository supporting core and auxiliary functionalities to advance QA research and application development.

Findings

Supports multiple QA functionalities including retrieval and reading comprehension.

Enables easy replication of SOTA QA methods on public benchmarks.

Provides a flexible toolkit for building QA applications and expanding existing methods.

Abstract

The field of Question Answering (QA) has made remarkable progress in recent years, thanks to the advent of large pre-trained language models, newer realistic benchmark datasets with leaderboards, and novel algorithms for key components such as retrievers and readers. In this paper, we introduce PRIMEQA: a one-stop and open-source QA repository with an aim to democratize QA re-search and facilitate easy replication of state-of-the-art (SOTA) QA methods. PRIMEQA supports core QA functionalities like retrieval and reading comprehension as well as auxiliary capabilities such as question generation.It has been designed as an end-to-end toolkit for various use cases: building front-end applications, replicating SOTA methods on pub-lic benchmarks, and expanding pre-existing methods. PRIMEQA is available at : https://github.com/primeqa.