nlee-208/UF-Mistral-Self8

该数据集包含多个字段，用于存储不同来源的对话数据及其评分。主要字段包括source（来源）、prompt（提示）、chosen（选择的对话内容及其角色）、chosen-rating（选择的对话评分）、chosen-model（选择的对话模型）、rejected（拒绝的对话内容及其角色）、rejected-rating（拒绝的对话评分）、rejected-model（拒绝的对话模型）和generated（生成的序列）。数据集被分割为训练集，包含5000个样本，总大小为94723871字节。

This dataset contains multiple fields for storing dialogue data from various sources along with their ratings. The main fields include source, prompt, chosen (selected dialogue content and its role), chosen-rating (rating of the selected dialogue), chosen-model (model used for the selected dialogue), rejected (rejected dialogue content and its role), rejected-rating (rating of the rejected dialogue), rejected-model (model used for the rejected dialogue), and generated (generated sequence). The dataset is split into a training set containing 5000 samples, with a total size of 94723871 bytes.