Mxode/Fineweb-Edu-Chinese-V2.1-merged-score4_5

这是一个来自 Fineweb-Edu-Chinese-V2.1 数据集的子集，包含评分为 4 到 5 的数据。原始数据集被细分为非常小的片段，大约每 10MB 一个片段。本数据集对这些片段进行了集合，每 400 个片段合并为一个片段，新的片段大小约为 4GB。为了方便加载，数据集根据这些片段被分割为多个子集，子集名称来源于原数据集的片段范围。可以指定加载其中的一个子集，也可以加载多个子集或全部子集。

This is a subset of the Fineweb-Edu-Chinese-V2.1 dataset, focusing on data with scores between 4 and 5. The original dataset was sliced into very fine segments, approximately one segment per 10MB. This dataset aggregates those segments, combining every 400 segments into one. Each new segment is about 4GB in size. For easier loading, the dataset has been divided into subsets based on these segments. The subset names are derived from the range of segments in the original dataset. You can load a specific subset, multiple subsets, or all subsets.