Xiaofeng77/reil_sokoban_long_horizon

该数据集是一个包含文本和数值信息的复合数据集，主要用于训练和测试某种模型。它包含了数据源、提示（包括内容和角色）、能力、奖励模型（包括地面真实值和风格）以及额外信息（包括索引和分割）。数据集分为训练集、测试集和测试环境，分别用于不同的训练和测试阶段。

This dataset is a composite dataset containing text and numerical information, primarily used for training and testing a certain model. It includes data source, prompt (including content and role), ability, reward model (including ground truth and style), and extra information (including index and split). The dataset is divided into training set, test set, and test environment, used for different training and testing stages.