MultiWay-Adapater: Adapting large-scale multi-modal models for scalable image-text retrieval

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Long, Zijun, Killick, George, McCreadie, Richard, Camarasa, Gerardo Aragon
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929234577981440
author Long, Zijun
Killick, George
McCreadie, Richard
Camarasa, Gerardo Aragon
author_facet Long, Zijun
Killick, George
McCreadie, Richard
Camarasa, Gerardo Aragon
contents As Multimodal Large Language Models (MLLMs) grow in size, adapting them to specialized tasks becomes increasingly challenging due to high computational and memory demands. Indeed, traditional fine-tuning methods are costly, due to the need for extensive, task-specific training. While efficient adaptation methods exist that aim to reduce these costs, in practice they suffer from shallow inter-modal alignment, which severely hurts model effectiveness. To tackle these computational challenges and improve inter-modal alignment, we introduce the MultiWay-Adapter (MWA), a novel framework featuring an 'Alignment Enhancer'. This enhancer deepens inter-modal alignment, enabling high transferability with minimal tuning effort. Our experiments show that unlike prior efficient tuning approaches, MWA maintains model effectiveness, while reducing training time by up-to 57%. MWA is also lightweight, increasing model size by only 2-3% (in terms of parameters) for state-of-the-art foundation models like BEiT-3 Large. These results demonstrate that MWA provides an efficient and effective adaptation method for MLLMs, significantly broadening their applicability.
format Preprint
id arxiv_https___arxiv_org_abs_2309_01516
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle MultiWay-Adapater: Adapting large-scale multi-modal models for scalable image-text retrieval
Long, Zijun
Killick, George
McCreadie, Richard
Camarasa, Gerardo Aragon
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Multimedia
As Multimodal Large Language Models (MLLMs) grow in size, adapting them to specialized tasks becomes increasingly challenging due to high computational and memory demands. Indeed, traditional fine-tuning methods are costly, due to the need for extensive, task-specific training. While efficient adaptation methods exist that aim to reduce these costs, in practice they suffer from shallow inter-modal alignment, which severely hurts model effectiveness. To tackle these computational challenges and improve inter-modal alignment, we introduce the MultiWay-Adapter (MWA), a novel framework featuring an 'Alignment Enhancer'. This enhancer deepens inter-modal alignment, enabling high transferability with minimal tuning effort. Our experiments show that unlike prior efficient tuning approaches, MWA maintains model effectiveness, while reducing training time by up-to 57%. MWA is also lightweight, increasing model size by only 2-3% (in terms of parameters) for state-of-the-art foundation models like BEiT-3 Large. These results demonstrate that MWA provides an efficient and effective adaptation method for MLLMs, significantly broadening their applicability.
title MultiWay-Adapater: Adapting large-scale multi-modal models for scalable image-text retrieval
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Multimedia
url https://arxiv.org/abs/2309.01516