ICBCBench
收藏资源简介:
ICBCBench 是一个旨在评估下一代大型语言模型(LLMs)的基准数据集,这些模型因添加工具、高效提示、搜索访问等而具有增强能力。数据集包含120个问题,涵盖金融和政治领域,问题具有明确且无歧义的答案。数据集支持中文和英文,包含40个主观问题和80个客观问题,主要为文本形式,也有少量多模态示例。数据集分为公开验证集和带有私有答案及元数据的测试集。数据文件包括80个客观问题的JSON文件和40个主观报告问题的JSONL文件。
ICBCBench is a benchmark dataset designed to evaluate the next generation of large language models (LLMs) with enhanced capabilities due to the addition of tools, efficient prompts, search access, etc. The dataset contains 120 questions covering the financial and political domains, with clear and unambiguous answers. The dataset supports both Chinese and English, including 40 subjective questions and 80 objective questions, primarily in text form with a few multimodal examples. The dataset is divided into a public validation set and a test set with private answers and metadata. Data files include JSON files for 80 objective questions and JSONL files for 40 subjective report questions.





