China's Nanjing Agricultural University (NAU) has introduced Sinong, the country's first open-source vertical large language model (LLM) dedicated to the general agricultural sector.
The launch marks a significant breakthrough in research on basic artificial intelligence (AI) models and their application to agriculture in China, as reported by Science and Technology Daily on Tuesday (January 13).
The model was trained using a vast and structured agricultural dataset, which includes specialized data from diverse fields such as animal science, agricultural economics and management, agricultural resources and environment, horticulture, smart agriculture, veterinary medicine, crop protection, and plant breeding.
The model, named after an ancient Chinese official who oversaw agriculture and finance, integrates content from nearly 9,000 books, more than 240,000 academic articles, about 20,000 policy documents and standards, and extensive web-based knowledge.
To address common challenges in specialized AI models such as hallucinations and knowledge obsolescence, which often result in misinformation, the team implemented a comprehensive technical solution. In addition to fine-tuning conventional instructions, the training process also incorporates multidimensional data, including chain-of-thought and contextual references, significantly enhancing the model's understanding and generative capabilities for professional agricultural knowledge, according to the report.
Sinong is now fully open-sourced on platforms such as ModelScope and GitHub. This open-source strategy aims to reduce barriers to AI adoption in agriculture, empowering research institutes, companies, and developers to pursue further development and innovation based on Sinong, thereby fostering a collaborative ecosystem for smart agricultural solutions, according to the NAU statement.
