Cartinoe5930/ust_sft_stage1

这是一个包含对话或问题的数据集，其中包括源(source)、问题(problem及其韩语版本problem_ko)、两个字符串s1和s2及其韩语版本(s1_ko和s2_ko)，以及索引(__index_level_0__)。此外，还有messages、chosen和rejected三个列表字段，每个字段包含content和role两个字符串。数据集分为训练集(train)，共有25000个示例。

This is a dataset containing dialogues or questions, which includes source, problem and its Korean version (problem_ko), two strings s1 and s2 and their Korean versions (s1_ko and s2_ko), and an index (__index_level_0__). Additionally, there are three list fields: messages, chosen, and rejected, each containing two strings: content and role. The dataset is split into a training set (train) with a total of 25,000 examples.