{"id":3875,"date":"2025-12-15T11:17:24","date_gmt":"2025-12-15T11:17:24","guid":{"rendered":"https:\/\/jita-au.com\/?p=3875"},"modified":"2025-12-15T13:48:22","modified_gmt":"2025-12-15T13:48:22","slug":"a-small-language-ai-model-in-the-bosnian-language","status":"publish","type":"post","link":"https:\/\/jita-au.com\/index.php\/2025\/12\/15\/a-small-language-ai-model-in-the-bosnian-language\/","title":{"rendered":"A Small Language AI Model in the Bosnian Language"},"content":{"rendered":"\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Vol. 15 No. 2 (2025): JITA - APEIRON<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Bo\u0161ko Jefi\u0107, Vlatko Bodul, Admir Agi\u0107<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t\n\t\t\t\t\t\t<\/span>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

A Small Language AI Model in the Bosnian Language<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t\n\t\t\t\t\t\t<\/span>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Review paper
DOI: https:\/\/doi.org\/10.7251\/JIT2502128J<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\n\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t\tDownload Article PDF<\/span>\n\t\t\t\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Abstract<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

This study presents the development and evaluation of Mali Mujo, a small-scale language model optimized for the\nBosnian language, designed to operate efficiently on devices with limited computational resources. Leveraging the TinyLlama\narchitecture, the model demonstrates the feasibility of deploying natural language processing (NLP) applications in environments\nwith constrained memory and processing capabilities, specifically devices with 1 GB storage and 8 GB RAM. The system integrates\nLangchain agents and the DuckDuckGo API to enable real-time information retrieval, enhancing the model\u2019s responsiveness and\naccuracy in practical applications. The methodology involved training the TinyLlama model on a curated Bosnian dataset, followed\nby testing across diverse real-world scenarios in industry and administration. Performance metrics focused on accuracy, response\ntime, and computational efficiency, while additional evaluation considered user experience and adaptability to domain-specific\ntasks. The results indicate that Mali Mujo delivers rapid and reliable responses to user queries, with significant advantages in speed\nand resource efficiency compared to larger language models. The model effectively processes administrative requests, generates\ntechnical and market-related insights, and supports educational and governmental applications, highlighting its versatility. While\nsmall-scale models exhibit lower absolute accuracy than their larger counterparts, the study demonstrates that careful optimization\nand integration with external APIs can mitigate limitations, providing a balance between performance and accessibility.\nFurthermore, the model\u2019s design ensures user privacy and low energy consumption, contributing to sustainable and secure AI\ndeployment. Mali Mujo exemplifies the potential of small language models to enhance efficiency, accessibility, and usability in locallanguage\ncontexts. Its deployment provides a scalable, cost-effective solution for organizations with limited infrastructure, offering\nopportunities for further enhancement through expanded datasets, multilingual support, adaptive learning, and integration with\nemerging AI technologies. The findings underscore the practicality of small AI models in bridging the gap between advanced NLP\ncapabilities and resource-constrained environments.\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t

Keywords: Small language models, TinyLlama, Bosnian language, Langchain agents, Real-time information retrieval, AI in industry.<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Paper received:<\/b> 31.10.2025.
Paper accepted:<\/b> 27.11.2025.<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\n\t\t\t