prithwishdan/RHyME

该数据集包含多个配置，每个配置都是视频数据集，包括robot、sphere-easy、sphere-hard、sphere-medium等。每个数据集都包含了视频ID、视频帧（包括二进制数据和路径信息）、元数据、动作序列和状态序列等信息。数据集分为训练集和评估集，不同配置的数据集具有不同的数据量和文件大小。

The dataset consists of multiple configurations, each being a video dataset including robot, sphere-easy, sphere-hard, sphere-medium, etc. Each dataset contains video ID, video frames (including binary data and path information), metadata, action sequences, and state sequences. The dataset is split into training and evaluation sets, and different configurations have different amounts of data and file sizes.