qfq/metamath_4o_sft

该数据集包含多个特征，如id、sample_id、question_statement、thinking_trajectory、golden_answer、answer、problem、orig_type、orig_problem、orig_original_question、orig_solution、orig_answer和text。这些特征涵盖了问题的各个方面，包括问题的陈述、思考轨迹、正确答案和用户提供的答案等。数据集分为训练集和测试集，分别包含146,567和7,715个样本。数据集的下载大小为385,726,339字节，总大小为1,020,645,429字节。

The dataset includes multiple features such as id, sample_id, question_statement, thinking_trajectory, golden_answer, answer, problem, orig_type, orig_problem, orig_original_question, orig_solution, orig_answer, and text. These features cover various aspects of questions, including the statement of the question, thinking trajectory, correct answers, and user-provided answers. The dataset is divided into training and test sets, containing 146,567 and 7,715 samples, respectively. The download size of the dataset is 385,726,339 bytes, and the total size is 1,020,645,429 bytes.