adzcai/babylm-eng-deu-50-50-stratified

BabyLM英语-德语50/50分层数据集是一个双语训练语料库，用于多语言BabyLM项目，由50%的英语和50%的德语BabyBabelLM语料库组成。采样方法是按类别分层，以保持每种语言中原始类别的比例。数据集包含总计103,460,378个令牌，其中英语占47.8%，德语占52.2%。类别细分包括儿童可用语音、儿童书籍、儿童导向语音、儿童新闻、儿童维基、教育内容、填充-开放字幕和填充-维基百科等，每种类别在英语和德语中的比例各不相同。

The BabyLM English–German 50/50 Stratified dataset is a bilingual training corpus for the Multilingual BabyLM project, combining 50% of the English and German BabyBabelLM corpora. Sampling is stratified by category to preserve the original category proportions within each language. The dataset contains a total of 103,460,378 tokens, with English accounting for 47.8% and German for 52.2%. The category breakdown includes child-available-speech, child-books, child-directed-speech, child-news, child-wiki, educational, padding-opensubtitles, and padding-wikipedia, with varying proportions in English and German.