Volume 458 - International Symposium on Grids & Clouds (ISGC) 2024 (ISGC2024) - Artificial Intelligence (AI)
OMAI: A Specialized Large Language Model for Operational Maintenance in Institute of High Energy Physics
S. Chen*, H. Li, Z. Zhang, Z. Sun and Y. Cheng
*: corresponding author
Full text: pdf
Published on: October 29, 2024
Abstract
This study presents the integration of Artificial Intelligence Operations (AIOps) at the Institute of High Energy Physics (IHEP) Computing Center to address the challenges of managing a complex IT infrastructure and providing efficient user support. We propose a question-answering (QA) system utilizing advancements in large language models and Natural Language Processing (NLP) to facilitate rapid problem resolution and document queries, aiming to reduce the operational staff's workload. The paper highlights the limitations of applying large language models in specialized domains due to insufficient exposure to domain-specific texts. It discusses using targeted fine-tuning and Retrieval-Augmented Generation (RAG) technology to enhance model performance. By compiling Helpdesk QA datasets for fine-tuning the Xiwu model, specializing in high-energy physics, and developing an RAG framework to integrate external knowledge, we significantly improve operational efficiency and support within the IHEP's computing environment.
DOI: https://doi.org/10.22323/1.458.0034
How to cite

Metadata are provided both in article format (very similar to INSPIRE) as this helps creating very compact bibliographies which can be beneficial to authors and readers, and in proceeding format which is more detailed and complete.

Open Access
Creative Commons LicenseCopyright owned by the author(s) under the term of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.