math-extraction-comp/princeton-nlp__Mistral-7B-Instruct-IPO

这是一个包含问题、答案及相关评分的数据集，适用于训练和评估问答系统。数据集中的每个样本包含了问题、标准答案、预测答案、数据来源子集，以及三个不同工具提取的答案和对应的评分。数据集分为训练集，提供了详细的文件大小和样本数量信息。

This dataset includes questions, answers, and related scores, which is suitable for training and evaluating question-answering systems. Each sample in the dataset contains a question, a standard answer, a predicted answer, the subset of data origin, and answers extracted by three different tools along with their corresponding scores. The dataset is split into a training set, and detailed information about file size and the number of samples is provided.