APUS releases intelligent grass model to assist the intelligent transformation and development of traditional Chinese medicine industry
2024-08-27
Recently, artificial intelligence company Kirin Hesheng (APUS) announced the launch of a 30B traditional Chinese medicine model called "APUS Smart Grass". This model is based on APUS's self-developed 210 billion parameter scale universal large model base and has been pruned. It utilizes 20 million high-quality medical knowledge data and takes more than 1400 hours to train, achieving an accuracy rate of 80.3% in the practicing physician exam. It is reported that the name "Zhicao" comes from the combination of wisdom and traditional Chinese medicine, symbolizing the use of AI technology to integrate traditional Chinese medicine knowledge with modern technology to provide more convenient and intelligent services to users. In the development stage of the Zhicao large model, APUS is based on its self-developed universal large model base (registration number: Henan-apasi-202404290001), incorporating traditional Chinese medicine knowledge through large-scale incremental pre training, and aligning with the experience of traditional Chinese medicine experts through fine-tuning and RLHF training techniques. In addition, the APUS Smart Grass Model also includes core functions such as prescription knowledge consultation, toxic Chinese medicine consultation, Chinese medicine disease diagnosis, Chinese medicine prescription recommendation, Chinese medicine medication recommendation, Chinese medicine tea drink formula, Chinese medicine dietary therapy formula, and acupoint information consultation. (New Society)
Edit:HAN ZHUOLING Responsible editor:CAICAI
Source:xinhuanet.com
Special statement: if the pictures and texts reproduced or quoted on this site infringe your legitimate rights and interests, please contact this site, and this site will correct and delete them in time. For copyright issues and website cooperation, please contact through outlook new era email:lwxsd@liaowanghn.com