selfcorrexp2/llama31_additional_prompt_first_cort_math_23_regular_processed

这个数据集包含了问题、答案、代码以及对话等信息。具体字段包括问题索引(idx)、问题(question)、正确答案(gt_cot和gt)、难度级别(level)、问题类型(type)、解决方案(solution)、我的解决方案(my_solu)、预测(pred)、代码(code)、对话轮数(turn)和对话内容(conversations)。数据集被划分为训练集(train)，共有342350个示例，总大小为1,877,878,740字节。

The dataset includes information such as questions, answers, code, and conversations. Specific fields include question index (idx), question (question), correct answers (gt_cot and gt), difficulty level (level), question type (type), solution (solution), my solution (my_solu), prediction (pred), code (code), number of conversation turns (turn), and conversation content (conversations). The dataset is split into a training set (train) with a total of 342,350 examples and a total size of 1,877,878,740 bytes.