Nowadays, the versatile capabilities of Pre-trained Large Language Models (LLMs) have attracted much attention from the industry. However, some vertical domains are more interested in the in-domain capabilities of LLMs. For the Networks domain, we present NetEval, an evaluation set for measuring the comprehensive capabilities of LLMs in Network Operations (NetOps). NetEval is designed for evaluating the commonsense knowledge and inference ability in NetOps in a multi-lingual context. NetEval consists of 5,732 questions about NetOps, covering five different sub-domains of NetOps. With NetEval, we systematically evaluate the NetOps capability of 26 publicly available LLMs. The results show that only GPT-4 can achieve a performance competitive to humans. However, some open models like LLaMA 2 demonstrate significant potential.
@article{arxiv.2309.05557,
title = {An Empirical Study of NetOps Capability of Pre-Trained Large Language Models},
author = {Yukai Miao and Yu Bai and Li Chen and Dan Li and Haifeng Sun and Xizheng Wang and Ziqiu Luo and Yanyu Ren and Dapeng Sun and Xiuting Xu and Qi Zhang and Chao Xiang and Xinchi Li},
journal= {arXiv preprint arXiv:2309.05557},
year = {2023}
}