Genyuan Wang,Hui Xie,Wenshuang Li

With the rapid advancement of next-generation artificial intelligence technologies, the open-source AI model ecosystem represented by large language models (LLMs)has become increasingly mature, presenting new opportunities and challenges for universities in cultivating interdisciplinary talents that meet industry demands. However, current AI education in universities often exhibits excessive reliance on cloud APIs, resulting in deficiencies in students' core engineering competencies such as localized model deployment and customized private knowledge base development. Guided by the industry-education integration strategy and closely aligned with the practical needs of the "Open-Source AI Model Localization Deployment" collaborative education project, this study systematically examines the current challenges, theoretical foundations, and practical pathways in AI talent cultivation. Through a review of recent academic literature, the paper proposes a training curriculum framework centered on a "theory-tool-practice" three-stage model, supported by lightweight fine-tuning and industry-education collaboration mechanisms. This model aims to achieve "low-code, high-efficiency" practical teaching by integrating enterprise-level open-source toolchains (e.g., Ollama/vLLM), enabling students to develop comprehensive engineering capabilities from model deployment to domain knowledge base construction. The research provides theoretical support and practical references for applied universities to deepen educational reforms under the context of new quality productivity.

industry-education integration; open-source AI model; localized deployment; practical training curriculum system

[1] Wang H.Research on the Optimization of Curriculum System for Cultivating Core AI Technical Competencies in Computer Science Majors at Guangxi Universities.Educational Research and Practice,2025,1(8). [2] Fu Rui,Fan Changying,Wu Zhenzhen,et al.Research on the Optimization of"4-3-2"Teaching and Education Decision-making for Applied Talent Cultivation Empowered by Large Language Models under the Perspective of Industry-Education Integration.Journal of Education,2025,3(6). [3] Zhang Sha-sha,Bai Yan-ru,LüQing-gong.Research on the Reform of Electronic Information Practice Base Construction under the Background of Industry-Education Integration.Experimental Science and Technology,2025,23(3). [4] Sun Jie,Liu Jing,Chang Xinyu.Research on the Construction of a Professional Practice Teaching System Based on Job Requirements.Educational Progress,2025,15(6). [5] Wang H.Research on the Construction and Practical Path of Artificial Intelligence Literacy Cultivation System for College Students in Guangxi.Educational Theory and Research,2025,1(6). [6] Xiao Wei,Li Jiandun,Ren Juhui.Research on Industry-Education Integration Courses in Applied Technical Universities Based on the Double Triple Helix Coupling Model.Advances in Social Sciences,2024. [7] Dong Shichang.Multimodal AI Empowering the Restructuring of Teaching Supervision System in Vocational Colleges for Cultural Tourism Integration:A Practical Path of Industry-Education Collaboration Based on Group-Based Education.Journal of Education,2025,3(1).

点击下载PDF