Octavius: Mitigating Task Interference in MLLMs via LoRA-MoE

TL;DR

This paper introduces Octavius, a novel framework combining LoRA and MoE techniques to mitigate task interference in Multimodal Large Language Models, significantly improving performance across diverse multimodal tasks.

Contribution

It pioneers the integration of MoE into MLLMs with LoRA to address task interference, demonstrating notable performance gains.

Findings

Approximately 20% performance improvement on downstream tasks

Effective mitigation of task interference in multimodal learning

Versatile application across 2D and 3D tasks

Abstract

Recent studies have demonstrated Large Language Models (LLMs) can extend their zero-shot generalization capabilities to multimodal learning through instruction tuning. As more modalities and downstream tasks are introduced, negative conflicts and interference may have a worse impact on performance. While this phenomenon has been overlooked in previous work, we propose a novel and extensible framework, called Octavius, for comprehensive studies and experimentation on multimodal learning with Multimodal Large Language Models (MLLMs). Specifically, we combine the well-known Mixture-of-Experts (MoE) and one of the representative PEFT techniques, i.e., LoRA, designing a novel LLM-based decoder, called LoRA-MoE, for multimodal learning. To the best of our knowledge, we are one of the pioneering efforts to introduce MoE into MLLMs to address this problem. The experimental results (about 20% improvement) have shown the effectiveness and versatility of our design in various 2D and 3D downstream tasks. Code and datasets are available at https://openlamm.github.io/tutorial/.