daikin-industries-ltd/ja-fineweb-2-hvac-fastText-filtered-v6

该数据集是从llm-jp/llm-jp-corpus-v3的ja_fineweb_2数据中，使用fastText分类器提取HVAC（空調、暖房、換気、空気調和）相关文本构建的。数据集仅包含被分类为`__label__positive`的记录，每条记录包含文本内容、元数据、fastText分类标签和分类分数。数据集总记录数为756,463条，总字符数为904,726,356，平均每记录1,196.0字符。fastText分类分数最小为0.5000，最大为1.0000，平均为0.8120，中位数为0.8472。数据集采用CC BY-SA 4.0许可证发布。

This dataset is constructed by extracting HVAC (heating, ventilation, air conditioning, and air conditioning) related texts from the ja_fineweb_2 data of llm-jp/llm-jp-corpus-v3 using a fastText classifier. The dataset only includes records classified as `__label__positive`. Each record contains the text content, metadata, fastText classification label, and classification score. The dataset has a total of 756,463 records, with a total of 904,726,356 characters, averaging 1,196.0 characters per record. The fastText classification scores range from 0.5000 to 1.0000, with an average of 0.8120 and a median of 0.8472. The dataset is released under the CC BY-SA 4.0 license.