Recently, the "Qimin" large language model for the vertical field of ancient agricultural texts was launched in Beijing. This model was developed by the Key Laboratory of Agricultural Big Data, Ministry of Agriculture and Rural Affairs, the Institute of Agricultural Information, Chinese Academy of Agricultural Sciences, in collaboration with the Hubei Provincial Library, the Library of Huazhong Agricultural University, the School of Communication at Zhengzhou Normal University, and Gu联 (Beijing) Digital Media Technology Co., Ltd. of Zhonghua Book Company.
Dai Xin, Deputy Director of the Institute of Agricultural Information, Chinese Academy of Agricultural Sciences, stated that the model is trained on a vast corpus of ancient agricultural texts from China, extracting wisdom from these ancient agricultural books to deeply explore ancient agricultural technologies, farming experiences, and social ecological knowledge. It supports downstream applications such as agricultural knowledge question answering, text automatic processing, and semantic retrieval.
It is understood that the "Qimin" large language model possesses robust natural language processing capabilities, providing precise answers to various agricultural issues such as ancient agricultural cultivation, livestock and poultry farming, farmland water conservancy, and meteorological prediction. Tools based on this large model, such as text automatic processing and semantic retrieval, can offer efficient analytical support for scholars researching ancient society and agriculture using agricultural ancient texts.