M-MiniGPT4: Multilingual VLLM Alignment via Translated Data

TL;DR

M-MiniGPT4 is a multilingual vision-language model that leverages translated data and parallel corpora to enhance understanding across 11 languages, outperforming existing models in its class.

Contribution

The paper introduces a multilingual alignment training stage and open-sources models and datasets, advancing multilingual VLU capabilities in the MiniGPT4 architecture.

Findings

Achieves 36% accuracy on the MMMU benchmark.

Outperforms state-of-the-art models in the same weight class.

Utilizes translated data and parallel corpora for training.

Abstract

This paper presents a Multilingual Vision Large Language Model, named M-MiniGPT4. Our model exhibits strong vision-language understanding (VLU) capabilities across 11 languages. We utilize a mixture of native multilingual and translated data to push the multilingual VLU performance of the MiniGPT4 architecture. In addition, we propose a multilingual alignment training stage that uses parallel text corpora to further enhance the multilingual capabilities of our model. M-MiniGPT4 achieves 36% accuracy on the multilingual MMMU benchmark, outperforming state-of-the-art models in the same weight class, including foundation models released after the majority of this work was completed. We open-source our models, code, and translated datasets to facilitate future research in low-resource and multilingual settings.