scottgeng00/olmo-3-preference-mix-deltas-yolo_lolwtf

该数据集包含了用户交互信息，如用户的IP hash、语言、国家、请求头部信息等，以及一些与交互相关的上下文信息，如对话的轮次标识符、对话内容、对话角色等。数据集还包含了内容是否被拒绝或编辑的标记，以及用于模型选择的模型名称。数据集分为训练集，包含约409111个样本，总大小约为3.38GB。

The dataset contains user interaction information such as user IP hash, language, country, request header information, etc., as well as context information related to interactions such as session turn identifier, conversation content, conversation role, etc. The dataset also includes marks indicating whether the content has been rejected or edited, and model names for model selection. The dataset is split into a training set with approximately 409111 samples, totaling about 3.38GB in size.