TAUR-dev/raw__BC__s1k_gpt4omini_CoT

该数据集包含了问题和解决方案的相关信息，以及一些关于思考和尝试的轨迹记录。具体包括问题的描述、解决方案、思考轨迹、尝试次数、评分和评分原因等。此外，还包含了与用户行为相关的提示、类型、跨度等信息。数据集分为训练集，共有400个样本。

The dataset contains information related to questions and solutions, as well as records of thinking and attempt trajectories. It includes descriptions of the questions, solutions, thinking trajectories, number of attempts, grades, and reasons for the grades. In addition, it contains prompts, types, spans related to user behavior. The dataset is split into a training set with a total of 400 samples.