This model is based on insturction data generated from Korean national research reports and NAVER news related to Infectious diseases.
Korean national R&D projects and keywords are selected by experts in the field of infectious diseases.
Infectious disease experts selected the reports of Korean national R&D projects. We generated insturction data by using the R&D reports. To ensure the currency of the training data, instruction data was updated using NAVER news articles found with these expert-selected keywords.
The base model of KoRnDAlpaca is EleutherAI/polyglot-en-3.8b.
For more information about the training procedure and model, please contact gsjang@kisti.re.kr.
