scottgeng00/olmo-3-preference-mix-deltas-complement2-yolo_even_split_no_multilingual-DECON
收藏数据链接：
官方服务：
资源简介：
该数据集包含了文本内容、用户信息、地理位置等多种类型的特征。每个样本包括prompt、chosen和rejected三个部分，每个部分下有多个子字段，如文本内容、国家、语言等。数据集划分为训练集，可用于文本生成、信息抽取等任务。
The dataset contains various types of features such as text content, user information, and geographical location. Each sample includes three parts: prompt, chosen, and rejected, with multiple sub-fields under each part, such as text content, country, language, etc. The dataset is split into a training set, which can be used for tasks like text generation and information extraction.
提供机构：
scottgeng00
