FedNano: Toward Lightweight Federated Tuning for Pretrained Multimodal Large Language Models

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zhang, Yao, Gao, Hewei, Chen, Haokun, Li, Weiguo, Ma, Yunpu, Tresp, Volker
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909651764772864
author Zhang, Yao
Gao, Hewei
Chen, Haokun
Li, Weiguo
Ma, Yunpu
Tresp, Volker
author_facet Zhang, Yao
Gao, Hewei
Chen, Haokun
Li, Weiguo
Ma, Yunpu
Tresp, Volker
contents Multimodal Large Language Models (MLLMs) excel in tasks like multimodal reasoning and cross-modal retrieval but face deployment challenges in real-world scenarios due to distributed multimodal data and strict privacy requirements. Federated Learning (FL) offers a solution by enabling collaborative model training without centralizing data. However, realizing FL for MLLMs presents significant challenges, including high computational demands, limited client capacity, substantial communication costs, and heterogeneous client data. Existing FL methods assume client-side deployment of full models, an assumption that breaks down for large-scale MLLMs due to their massive size and communication demands. To address these limitations, we propose FedNano, the first FL framework that centralizes the LLM on the server while introducing NanoEdge, a lightweight module for client-specific adaptation. NanoEdge employs modality-specific encoders, connectors, and trainable NanoAdapters with low-rank adaptation. This design eliminates the need to deploy LLM on clients, reducing client-side storage by 95%, and limiting communication overhead to only 0.01% of the model parameters. By transmitting only compact NanoAdapter updates, FedNano handles heterogeneous client data and resource constraints while preserving privacy. Experiments demonstrate that FedNano outperforms prior FL baselines, bridging the gap between MLLM scale and FL feasibility, and enabling scalable, decentralized multimodal AI systems.
format Preprint
id arxiv_https___arxiv_org_abs_2506_14824
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FedNano: Toward Lightweight Federated Tuning for Pretrained Multimodal Large Language Models
Zhang, Yao
Gao, Hewei
Chen, Haokun
Li, Weiguo
Ma, Yunpu
Tresp, Volker
Machine Learning
Artificial Intelligence
Multimedia
Multimodal Large Language Models (MLLMs) excel in tasks like multimodal reasoning and cross-modal retrieval but face deployment challenges in real-world scenarios due to distributed multimodal data and strict privacy requirements. Federated Learning (FL) offers a solution by enabling collaborative model training without centralizing data. However, realizing FL for MLLMs presents significant challenges, including high computational demands, limited client capacity, substantial communication costs, and heterogeneous client data. Existing FL methods assume client-side deployment of full models, an assumption that breaks down for large-scale MLLMs due to their massive size and communication demands. To address these limitations, we propose FedNano, the first FL framework that centralizes the LLM on the server while introducing NanoEdge, a lightweight module for client-specific adaptation. NanoEdge employs modality-specific encoders, connectors, and trainable NanoAdapters with low-rank adaptation. This design eliminates the need to deploy LLM on clients, reducing client-side storage by 95%, and limiting communication overhead to only 0.01% of the model parameters. By transmitting only compact NanoAdapter updates, FedNano handles heterogeneous client data and resource constraints while preserving privacy. Experiments demonstrate that FedNano outperforms prior FL baselines, bridging the gap between MLLM scale and FL feasibility, and enabling scalable, decentralized multimodal AI systems.
title FedNano: Toward Lightweight Federated Tuning for Pretrained Multimodal Large Language Models
topic Machine Learning
Artificial Intelligence
Multimedia
url https://arxiv.org/abs/2506.14824