hyunseoki/iself-gsm8k

该数据集包含了问题回答的尝试和验证信息，旨在训练和测试模型对问题的理解和回答能力。数据集分为四种配置，每种配置都有训练集和测试集。数据集中的字段包括问题、答案、第一次尝试的提示和反应、第一次尝试的答案和正确性、第二次尝试的提示和反应（在后两种配置中）、第二次尝试的理由、验证提示和反应、验证答案和正确性。

The dataset contains information about attempts and validations of question answering, aimed at training and testing the models ability to understand and answer questions. It consists of four configurations, each with a training set and a test set. The fields in the dataset include the question, answer, first attempt prompt and response, first attempt answer and correctness, second attempt prompt and response (in the last two configurations), second attempt rationale, validation prompt and response, validation answer, and correctness.