Enhancing Fitness Intelligence through Domain-Specific LLM Post-Training
作者: Xingtao Zhao, Tian Yang, Han Jiang
分类: cs.AI
发布日期: 2026-07-05
💡 一句话要点
提出FitOne以解决科学健身教练领域的知识不足问题
🎯 匹配领域: 支柱二:RL算法与架构 (RL & Architecture) 支柱九:具身大模型 (Embodied Foundation Models)
关键词: 大型语言模型 科学健身教练 领域特定知识 后训练 强化学习 知识工程 模型评估
📋 核心要点
- 现有的通用大型语言模型在科学健身教练领域缺乏足够的专业知识,导致性能不足。
- 本文提出FitOne,通过三阶段后训练流程,结合领域特定数据集,提升SFC应用的可靠性和专业性。
- FitOne-8B/32B在专业健身认证考试中表现优异,相较于基线模型有显著提升,验证了方法的有效性。
📝 摘要(中文)
科学健身教练(SFC)通常由人类专业人士提供,成本高且难以普及。尽管大型语言模型(LLMs)在健身教练中展现出潜力,但直接应用现有的通用LLMs在SFC中存在显著局限性,主要是缺乏足够的领域特定知识。本文提出FitOne,一系列专为SFC设计的健身LLMs(8B和32B参数),通过持续预训练、监督微调和强化学习的三阶段后训练流程,结合大规模高质量数据集,显著提升了模型的可靠性和领域专业性。实验结果表明,FitOne-8B/32B在ACSM-EP和NSCA-CSCS考试中分别提升了10.09%/9.29%和12.73%/7.01%。
🔬 方法详解
问题定义:本文旨在解决现有通用大型语言模型在科学健身教练领域知识不足的问题,这导致其在复杂场景中的表现不佳。
核心思路:提出FitOne系列模型,通过后训练流程增强模型的领域特定知识,同时保持其通用能力,以适应科学健身教练的需求。
技术框架:FitOne的开发分为三个主要阶段:持续预训练、监督微调和强化学习,利用高质量的数据集进行训练,确保模型在健身领域的专业性。
关键创新:FitOne的创新在于其后训练流程的设计,特别是结合了领域特定知识与通用能力的平衡,显著提升了模型在SFC中的应用效果。
关键设计:模型参数设置为8B和32B,采用了适合健身领域的损失函数和网络结构,确保在训练过程中有效捕捉领域知识。通过大规模知识工程构建高质量数据集,增强模型的学习效果。
🖼️ 关键图片
📊 实验亮点
FitOne-8B/32B在ACSM-EP和NSCA-CSCS考试中分别提升了10.09%和9.29%,以及12.73%和7.01%的平均成绩,相较于Qwen3基线模型表现出显著的性能提升,验证了其在科学健身教练领域的有效性。
🎯 应用场景
FitOne模型的潜在应用场景包括健身教练、健康咨询和个性化训练计划制定等领域。其高效的领域知识整合能力将使得健身指导更加普及和可及,降低专业健身服务的成本,推动健身行业的智能化发展。
📄 摘要(原文)
Scientific Fitness Coaching (SFC) is typically delivered by human professionals, making it costly and inaccessible to many. While recent advances in Large Language Models (LLMs) show considerable promise for more inclusive fitness coaching, directly deploying prevailing general-purpose LLMs in SFC reveals critical limitations. These models often lack sufficient domain-specific knowledge integration, leading to weak performance on complex SFC scenarios. In this paper, we introduce FitOne, a series of fitness LLMs (with 8B and 32B parameters) designed to improve reliability and domain specialization for SFC applications. Built upon the Qwen3 foundation models, FitOne is developed through a three-stage post-training pipeline consisting of continual pre-training, supervised fine-tuning, and reinforcement learning, using large-scale, high-quality datasets derived from rigorous knowledge engineering. We conduct comprehensive evaluations of FitOne on professional fitness certification exams, including ACSM-EP and NSCA-CSCS, as well as general capabilities such as knowledge reasoning and instruction following. Experimental results show that, while retaining strong general capabilities, FitOne-8B/32B achieves average improvements of up to 10.09%/9.29% and 12.73%/7.01% on the ACSM-EP and NSCA-CSCS exams, respectively, compared with the Qwen3 base models. Furthermore, in-depth ablation studies confirm the necessity of each training stage, highlighting the pipeline's effectiveness in balancing domain expertise enhancement with general ability retention. We believe this research advances LLM systems toward more reliable fitness intelligence and will inspire future research on developing domain-specific LLMs.