regent-project/regent-subset-of-jat-dataset-tokenized

该数据集包含多个Atari游戏的数据集配置，每个配置包含不同的特征（如distances、indices、image_observations等）和训练集的分割信息（如num_bytes、num_examples等）。这些数据集主要用于游戏相关的机器学习任务，如强化学习。

This dataset contains multiple configurations of Atari game datasets, each with different features (e.g., distances, indices, image_observations, etc.) and training set splits (e.g., num_bytes, num_examples, etc.). These datasets are primarily used for game-related machine learning tasks, such as reinforcement learning.