CFBenchmark
收藏资源简介:
CFBenchmark是一个专为评估大型语言模型在中文金融助理领域性能而设计的数据集。由同济大学和上海人工智能实验室联合创建,该数据集包含3917个金融文本,覆盖金融识别、分类和生成三大方面,共计八个任务。数据集内容丰富,包括金融新闻和研究报告,文本长度从50字到超过1800字不等,旨在全面测试模型在金融文本处理中的基本能力。创建过程中,专业研究人员对文本进行了严格的筛选和标注,确保数据质量。CFBenchmark的应用领域广泛,主要用于提升金融决策的自动化和智能化水平,解决金融文本处理中的关键问题。
CFBenchmark is a dataset specifically designed for evaluating the performance of large language models (LLMs) in the Chinese financial assistant domain. Co-developed by Tongji University and the Shanghai Artificial Intelligence Laboratory, it contains 3,917 financial texts spanning three core areas: financial recognition, classification, and generation, with a total of eight tasks. The dataset features diverse content including financial news and research reports, with text lengths ranging from 50 words to over 1,800 words, and is designed to comprehensively assess the core capabilities of models in financial text processing. During its development, professional researchers conducted strict screening and annotation of all texts to ensure high data quality. CFBenchmark has a wide range of application scenarios, primarily used to enhance the automation and intelligence of financial decision-making and address key challenges in financial text processing.




