LLaCA: Multimodal Large Language Continual Assistant

Jingyang Qiao, zhizhong zhang, Xin Tan, Yanyun Qu, Shouhong Ding, Yuan Xie

Keywords: Multimodal Continual Instruction Tuning, Anti-Forgetting, Exponential Movement Average, LoRA, LLaVA

TL;DR: A method to resist forgetting in multimodal continual instruction tuning based on Multimodal Large Language Modals

Abstract: Instruction tuning guides the Multimodal Large Language Models (MLLMs) in aligning different modalities by designing text instructions, which seems to be an essential technique to enhance the capabilities and controllability of foundation models. In this framework, Multimodal Continual Instruction Tuning (MCIT) is adopted to continually instruct MLLMs to follow human intent in sequential datasets. We observe existing gradient update would heavily destroy the tuning performance on previous datasets and the zero-shot ability during continual instruction tuning. Exponential Moving Average (EMA) update policy owns the ability to trace previous parameters, which can aid in decreasing forgetting. However, its stable balance weight cannot deal with the ever-changing datasets, leading to the out-of-balance between plasticity and stability of MLLMs. In this paper, we propose a method called Multimodal Large Language Continual Assistant (LLaCA) to address the challenge. Starting from the trade-off prerequisite and EMA update, we propose the plasticity and stability ideal condition. Based on Taylor expansion in the loss function, we find the optimal balance weight is basically according to the gradient information and previous parameters. We automatically determine the balance weight and significantly improve the performance. Through comprehensive experiments on LLaVA-1.5 in a continual visual-question-answering benchmark, compared with baseline, our approach not only highly improves anti-forgetting ability (with reducing forgetting from 22.67 to 2.68), but also significantly promotes continual tuning performance (with increasing average accuracy from 41.31 to 61.89). Our code will be published soon.

Primary Area: transfer learning, meta learning, and lifelong learning

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 3461