DataCLUE

DataCLUE是首个应用于自然语言处理领域的数据中心基准，由CLUE团队创建。该数据集旨在通过优化数据集质量来提升模型性能，特别关注于提高数据集的实用性和关注度。DataCLUE包含多个代表性任务和评估指标，提供在线排行榜，并开放了相关资源，包括数据集、工具包和基线方法。此数据集适用于学术研究和工业应用，特别是在解决数据质量问题方面具有重要价值。

DataCLUE is the first data-centric benchmark in the field of natural language processing, created by the CLUE team. This benchmark aims to enhance model performance by optimizing dataset quality, with particular focus on improving the practicality and visibility of datasets. DataCLUE encompasses multiple representative tasks and evaluation metrics, provides an online leaderboard, and opens up relevant resources including datasets, toolkits, and baseline methods. This benchmark is suitable for both academic research and industrial applications, and is particularly valuable in addressing data quality issues.