kimic/pacman

这个数据集包含了三个不同难度级别的Pacman游戏的状态-动作对，每个级别的内容都包含在一个tar.gz文件中。数据集通过强化学习任务进行分类，并标记有GWM和Pacman。数据集大小在100K到1M之间，文件语言为英语。每个tar.gz文件包含输出文件夹、dataset.csv文件和results.txt文件。输出文件夹里有由强化学习代理生成的所有图像/帧。dataset.csv文件记录了与每个帧相对应的动作，动作类型包括北、南、东、西或停止，并有一个表示新游戏是否开始的标记列。results.txt文件记录了每局游戏的胜利或失败结果。

This dataset contains state-action pairs from three different Pacman levels, each contained within its own tar.gz file. The dataset is categorized under reinforcement learning tasks and tagged with GWM and Pacman. The size of the dataset is between 100K and 1M, with the language being English. Each tar.gz file includes an output folder, a dataset.csv file, and a results.txt file. The output folder contains all images/frames generated by a reinforcement learning agent based on Taylor Downeys RL Agent. The dataset.csv file records the actions corresponding to each frame, which include North, South, East, West, or Stop, along with a column indicating the start of a new game. The results.txt file records the Win or Lose outcome for each game.