gvlassis/ClimbMix

ClimbMix是一个经过改进的文本数据集，解决了原始ClimbMix数据集在使用中的不便之处。它将数据集分为多个子集，每个子集代表一个特定的主题，如数学、历史、教育等。每个子集都包含了一定数量的文档和字节大小信息。此外，ClimbMix还提供了每个子集的下载大小和文档数量。数据集使用GPT-4.1-mini模型提取每个子集的主要主题，方便用户进行文本处理和自然语言处理任务。

ClimbMix is an improved text dataset that addresses the inconveniences of the original ClimbMix dataset. It is split into multiple subsets, each representing a specific topic such as mathematics, history, education, etc. Each subset includes information about the number of documents and byte sizes. Additionally, ClimbMix provides the download size and document count for each subset. The dataset uses the GPT-4.1-mini model to extract the main topics of each subset, facilitating text processing and natural language processing tasks.