FOLDER: Accelerating Multi-modal Large Language Models with Enhanced Performance

TL;DR

FOLDER is a plug-and-play module that reduces visual token sequence length in multi-modal large language models, significantly accelerating inference and training while maintaining or improving performance by removing up to 70% of visual tokens.

Contribution

The paper introduces FOLDER, a novel token reduction method that preserves key information and accelerates multi-modal large language models without sacrificing accuracy.

Findings

FOLDER reduces visual tokens by up to 70%.

Models with FOLDER achieve comparable or better performance.

FOLDER accelerates inference and training processes.

Abstract

Recently, Multi-modal Large Language Models (MLLMs) have shown remarkable effectiveness for multi-modal tasks due to their abilities to generate and understand cross-modal data. However, processing long sequences of visual tokens extracted from visual backbones poses a challenge for deployment in real-time applications. To address this issue, we introduce FOLDER, a simple yet effective plug-and-play module designed to reduce the length of the visual token sequence, mitigating both computational and memory demands during training and inference. Through a comprehensive analysis of the token reduction process, we analyze the information loss introduced by different reduction strategies and develop FOLDER to preserve key information while removing visual redundancy. We showcase the effectiveness of FOLDER by integrating it into the visual backbone of several MLLMs, significantly accelerating the inference phase. Furthermore, we evaluate its utility as a training accelerator or even performance booster for MLLMs. In both contexts, FOLDER achieves comparable or even better performance than the original models, while dramatically reducing complexity by removing up to 70% of visual tokens.