open-cloth/so101-full-fold-merged-g8

该数据集是一个机器人控制数据集，使用LeRobot平台创建。数据集包含118个训练episodes，总计105071帧数据，帧率为30fps。数据特征包括：6维动作空间（肩部旋转、肩部提升、肘部弯曲、腕部弯曲、腕部旋转、夹爪位置），6维状态观测（与动作相同的关节位置），夹爪摄像头图像观测（480x640分辨率RGB视频），以及时间戳、帧索引、episode索引等元数据。数据以parquet格式存储，视频以mp4格式存储。数据集适用于机器人模仿学习、强化学习等任务，但具体任务类型未明确说明。

This dataset is a robotics control dataset created using the LeRobot platform. It contains 118 training episodes with a total of 105,071 frames at 30 fps. The features include: 6-dimensional action space (shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, gripper positions), 6-dimensional state observations (same joint positions as actions), gripper camera image observations (480x640 resolution RGB video), and metadata such as timestamps, frame indices, and episode indices. Data is stored in parquet format, and videos are stored in mp4 format. The dataset is suitable for tasks like robot imitation learning and reinforcement learning, though the specific task type is not explicitly stated.