NewstaR/reasoning-graded

该数据集包含会话信息以及相关教育分数的评估。每个会话记录了信息的来源和内容。此外，数据集为每个会话的提示和回应提供了教育分数，以及计算了平均教育分数。数据集分为训练集，包含29857个示例，总大小为75118418字节。

The dataset consists of conversation information along with associated educational score evaluations. Each conversation records the source and content of the information. Additionally, the dataset provides educational scores for each conversations prompt and response, as well as calculates the average educational score. The dataset is split into a training set, which contains 29,857 examples and has a total size of 75,118,418 bytes.