daikin-industries-ltd/ja-fineweb-2-hvac-fastText-scored-v3

本数据集是基于daikin-industries-ltd/ja-fineweb-2-hvac-fastText-filtered-v3的扩展版本，增加了LLM的质量评分。数据集包含空调（HVAC）相关的日文文本数据，每个数据点包含文本内容、URL、来源信息、语言代码、语言评分、FastText分类评分、LLM质量评分（1-5分）及评分依据。FastText评分用于判断文本是否与空調相关，LLM评分则基于文本对空調技术人员的教育价值进行评分。数据集大小为100K到1M之间，主要用于文本生成任务。

This dataset is an extended version of daikin-industries-ltd/ja-fineweb-2-hvac-fastText-filtered-v3, with added LLM quality scoring. It contains Japanese text data related to HVAC (Heating, Ventilation, and Air Conditioning), with each data point including text content, URL, source information, language code, language score, FastText classification score, LLM quality score (1-5), and scoring rationale. The FastText score determines whether the text is HVAC-related, while the LLM score evaluates the educational value of the text for HVAC technicians. The dataset size ranges from 100K to 1M and is primarily used for text generation tasks.