robot-learning/ex2_all_v4

该数据集是使用LeRobot创建的机器人学习数据集，专门用于so_follower类型的机器人。数据集包含300个完整的情节（episodes），总计163,606帧数据，覆盖3个不同的任务。数据以30帧每秒（fps）的速率采集，并以parquet格式存储，数据文件总大小为100MB，视频文件总大小为200MB。数据集的特征包括机器人的动作（action）和观测状态（observation.state），动作和状态均包含6个关节位置（如肩部平移、肩部升降、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置）。观测部分还包括来自前置摄像头的图像数据，分辨率为480x640，3通道彩色视频。此外，数据集还包含时间戳、帧索引、情节索引、索引和任务索引等元数据。数据集仅包含训练集（train），划分为0到300的情节范围。This dataset is a robotics learning dataset created using LeRobot, specifically for the so_follower robot type. It contains 300 total episodes, with a total of 163,606 frames, covering 3 different tasks. The data is collected at 30 frames per second (fps) and stored in parquet format, with data files totaling 100MB and video files totaling 200MB. The dataset features include robot actions and observation states, both comprising 6 joint positions (e.g., shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, and gripper positions). Observations also include image data from a front camera, with a resolution of 480x640, 3-channel color video. Additionally, the dataset includes metadata such as timestamp, frame index, episode index, index, and task index. The dataset only contains a training split (train), covering episodes 0 to 300.