clembench-playpen/DPO_turn_solved2

这是一个关于游戏互动的数据集，包含了游戏名称、游戏ID、版本和实验信息等字段。数据集记录了每个剧集的详细信息，包括模型在游戏中的成功与否，以及玩家在每个分支回合中的选择和拒绝信息。数据集被划分为训练集，大小为97479977字节，共有58859个示例。

This is a game interaction dataset that includes fields such as game name, game ID, version, and experiment information. The dataset records detailed information for each episode, including whether the model was successful in the game and the players choices and rejections at each branch turn. The dataset is split into a training set, which is 97479977 bytes in size and contains 58859 examples.