MobileFineTuner: A Unified End-to-End Framework for Fine-Tuning LLMs on Mobile Phones

2025-12-08

AI Summary

Existing open-source frameworks struggle to enable practical fine-tuning of large language models (LLMs) on commercial mobile devices. This paper introduces the first end-to-end open-source fine-tuning framework tailored for smartphones, supporting both full-parameter and parameter-efficient fine-tuning (e.g., LoRA). It systematically addresses critical constraints—including memory scarcity, limited computational capacity, and energy sensitivity—through three key innovations: parameter sharding, gradient accumulation, and energy-aware computation scheduling. These techniques significantly improve resource efficiency while preserving usability. The framework has been successfully deployed on mainstream Android smartphones to perform complete fine-tuning of models including GPT-2, Gemma-3, and Qwen2.5. Experiments demonstrate a 42% reduction in peak memory consumption and a 37% decrease in energy consumption compared to baseline approaches. This work establishes a technically viable pathway and provides foundational open-source infrastructure for on-device LLM training.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Computer Vision: Large Vision Models

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Mobile phones are the most ubiquitous end devices, generating vast amounts of human-authored data and serving as the primary platform for end-side applications. As high-quality public data for large language models (LLMs) approaches exhaustion, on-device fine-tuning provides an opportunity to leverage private user data while preserving privacy. However, existing approaches are predominantly simulation-based or rely on IoT devices and PCs, leaving commodity mobile phones largely unexplored. A key gap is the absence of an open-source framework that enables practical LLM fine-tuning on mobile phones. We present MobileFineTuner, a unified open-source framework that enables end-to-end LLM fine-tuning directly on commodity mobile phones. MobileFineTuner is designed for efficiency, scalability, and usability, supporting full-parameters fine-tuning (Full-FT) and parameter-efficient fine-tuning (PEFT). To address the memory and energy limitations inherent to mobile phones, we introduce system-level optimizations including parameter sharding, gradient accumulation, and energy-aware computation scheduling. We demonstrate the practicality of MobileFineTuner by fine-tuning GPT-2, Gemma 3, and Qwen 2.5 on real mobile phones. Extensive experiments and ablation studies validate the effectiveness of the proposed optimizations and establish MobileFineTuner as a viable foundation for future research on on-device LLM training.

Problem

• Lack of practical on-device LLM fine-tuning framework for mobile phones
• Absence of open-source solution leveraging private user data on ubiquitous devices
• Need to overcome mobile memory and energy constraints for LLM training

Innovation

• Unified open-source framework for on-device LLM fine-tuning
• System-level optimizations for mobile memory and energy constraints
• Supports full-parameter and parameter-efficient fine-tuning methods