so101_test

该数据集基于LeRobot框架构建，专门用于so101_follower型机器人的任务数据收集。数据集包含10个episodes，总计7699帧图像数据，帧率为30fps。数据特征包括6维动作和状态信息（对应肩部平移、肩部提升、肘部弯曲、手腕弯曲、手腕旋转和夹爪位置），以及前视摄像头捕获的图像序列（分辨率为480x640，3通道）。数据以parquet文件存储动作和状态数据，mp4视频文件存储图像序列。数据分割方面，所有10个episodes均被划分为训练集。

This dataset is created using the LeRobot framework and contains task data for the so101_follower robot. It comprises 10 episodes with a total of 7699 frames at a frame rate of 30fps. The data features include 6-dimensional actions and states (corresponding to shoulder translation, shoulder elevation, elbow flexion, wrist flexion, wrist rotation, and gripper position), as well as front-view camera images (resolution 480x640, 3 channels). The action and state information is stored in parquet files, while the image sequences are stored in mp4 video files. The data is split into a training set, which includes all 10 episodes.