domce20/c4-lithuanian-enhanced

该数据集包含文本、时间戳、URL、两种困惑度分数（perplexity_lrt和perplexity_wiki）以及详细的语言信息。语言信息包括结束索引、语言类型、开始索引和单词计数。数据集主要包含一个训练集，共有11,274,295个样本，总数据大小为40,448,045,198字节，下载大小为24,678,151,770字节。

The dataset includes text, timestamps, URLs, two types of perplexity scores (perplexity_lrt and perplexity_wiki), and detailed language information. The language information is further divided into end index, language type, start index, and word count. The dataset primarily consists of a training set with 11,274,295 samples, totaling 40,448,045,198 bytes in size, and a download size of 24,678,151,770 bytes.