Learning to Inference Adaptively for Multimodal Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xu, Zhuoyan, Nguyen, Khoi Duc, Mukherjee, Preeti, Bagchi, Saurabh, Chaterji, Somali, Liang, Yingyu, Li, Yin
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918115576643584
author Xu, Zhuoyan
Nguyen, Khoi Duc
Mukherjee, Preeti
Bagchi, Saurabh
Chaterji, Somali
Liang, Yingyu
Li, Yin
author_facet Xu, Zhuoyan
Nguyen, Khoi Duc
Mukherjee, Preeti
Bagchi, Saurabh
Chaterji, Somali
Liang, Yingyu
Li, Yin
contents Multimodal Large Language Models (MLLMs) have shown impressive capabilities in visual reasoning, yet come with substantial computational cost, limiting their deployment in resource-constrained settings. Despite recent effort on improving the efficiency of MLLMs, prior solutions fall short in responding to varying runtime conditions, in particular changing resource availability (e.g., contention due to the execution of other programs on the device). To bridge this gap, we introduce AdaLLaVA, an adaptive inference framework that learns to dynamically reconfigure operations in an MLLM during inference, accounting for the input data and a latency budget. We conduct extensive experiments across benchmarks involving question-answering, reasoning, and hallucination. Our results show that AdaLLaVA effectively adheres to input latency budget, achieving varying accuracy and latency tradeoffs at runtime. Further, we demonstrate that AdaLLaVA adapts to both input latency and content, can be integrated with token selection for enhanced efficiency, and generalizes across MLLMs. Our project webpage with code release is at https://zhuoyan-xu.github.io/ada-llava/.
format Preprint
id arxiv_https___arxiv_org_abs_2503_10905
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Learning to Inference Adaptively for Multimodal Large Language Models
Xu, Zhuoyan
Nguyen, Khoi Duc
Mukherjee, Preeti
Bagchi, Saurabh
Chaterji, Somali
Liang, Yingyu
Li, Yin
Artificial Intelligence
Computer Vision and Pattern Recognition
Machine Learning
Multimodal Large Language Models (MLLMs) have shown impressive capabilities in visual reasoning, yet come with substantial computational cost, limiting their deployment in resource-constrained settings. Despite recent effort on improving the efficiency of MLLMs, prior solutions fall short in responding to varying runtime conditions, in particular changing resource availability (e.g., contention due to the execution of other programs on the device). To bridge this gap, we introduce AdaLLaVA, an adaptive inference framework that learns to dynamically reconfigure operations in an MLLM during inference, accounting for the input data and a latency budget. We conduct extensive experiments across benchmarks involving question-answering, reasoning, and hallucination. Our results show that AdaLLaVA effectively adheres to input latency budget, achieving varying accuracy and latency tradeoffs at runtime. Further, we demonstrate that AdaLLaVA adapts to both input latency and content, can be integrated with token selection for enhanced efficiency, and generalizes across MLLMs. Our project webpage with code release is at https://zhuoyan-xu.github.io/ada-llava/.
title Learning to Inference Adaptively for Multimodal Large Language Models
topic Artificial Intelligence
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2503.10905