qfq/cotmath

该数据集包含多个字段，主要用于记录和细化思维轨迹。字段包括唯一标识符（id）、细化后的思维轨迹（refined_thinking_trajectory）、原始标识符（orig_id）、原始样本标识符（orig_sample_id）、原始问题陈述（orig_question_statement）、原始思维轨迹（orig_thinking_trajectory）、原始正确答案（orig_golden_answer）、原始答案（orig_answer）、原始问题（orig_problem）、原始类型（orig_orig_type）、原始问题（orig_orig_problem）、原始原始问题（orig_orig_original_question）、原始解决方案（orig_orig_solution）和原始答案（orig_orig_answer）。数据集主要用于训练模型，包含1,147,653个样本，总大小为4,592,760,926字节。

This dataset includes multiple fields primarily used for recording and refining thinking trajectories. The fields include a unique identifier (id), refined thinking trajectory (refined_thinking_trajectory), original identifier (orig_id), original sample identifier (orig_sample_id), original question statement (orig_question_statement), original thinking trajectory (orig_thinking_trajectory), original golden answer (orig_golden_answer), original answer (orig_answer), original problem (orig_problem), original type (orig_orig_type), original problem (orig_orig_problem), original original question (orig_orig_original_question), original solution (orig_orig_solution), and original answer (orig_orig_answer). The dataset is primarily used for training models and contains 1,147,653 samples with a total size of 4,592,760,926 bytes.