CohenQu/POPE-67-medium-no_guide

该数据集包含多个特征，包括数据来源（data_source）、提示（prompt）、能力（ability）、奖励模型（reward_model）、额外信息（extra_info）、级别（level）和问题ID（problem_id）。提示字段是一个列表，包含内容和角色；奖励模型字段包含真实值和风格；额外信息字段包含索引和分割信息。数据集分为训练集和测试集，训练集有256个样本，测试集有180个样本。

The dataset includes multiple features such as data source (data_source), prompt, ability, reward model (reward_model), extra information (extra_info), level, and problem ID (problem_id). The prompt field is a list containing content and role; the reward model field includes ground truth and style; the extra information field contains index and split information. The dataset is divided into training and test sets, with 256 samples in the training set and 180 samples in the test set.