corbt/art_e_sample_trajectories

这是一个包含问题、答案以及多种评价度量的数据集，适用于训练和评估自然语言处理模型。数据集中的字段包括问题ID、问题内容、答案、消息ID序列、真实性评分、邮箱地址、查询日期、数据集分割类型、使用的模型、轨迹序列、奖励值、评价度量等。数据集分为训练集，并提供了相应的配置信息。

This dataset includes questions, answers, and various evaluation metrics, suitable for training and evaluating natural language processing models. The fields in the dataset consist of question ID, question content, answer, message ID sequence, realism score, email address, query date, dataset split type, model used, trajectory sequence, reward, evaluation metrics, etc. The dataset is split into a training set and provides corresponding configuration information.