thinhpt/test_merged3

该数据集是使用LeRobot创建的机器人数据集，专门用于机器人任务，具体针对so101_follower机器人类型。数据集包含60个episodes，总计23175帧，帧率为30fps。数据以parquet文件格式存储，总数据文件大小为100MB，视频文件大小为200MB。数据集仅包含一个任务，并划分为训练集（全部60个episodes）。数据特征包括动作（6个浮点数值，对应机器人关节位置：肩部平移、肩部抬升、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置）、观察状态（与动作相同的6个关节位置）、以及来自三个视角的图像观察：顶部摄像头、腕部摄像头和侧面摄像头，每个图像分辨率为480x640，3通道，视频编码为av1。此外，还包括时间戳、帧索引、episode索引、索引和任务索引等元数据。

This dataset is created using LeRobot and is a robotics dataset specifically for robot tasks, targeting the so101_follower robot type. It contains 60 episodes with a total of 23175 frames at 30 fps. The data is stored in parquet format, with a total data file size of 100MB and video file size of 200MB. The dataset includes only one task and is split into a training set (all 60 episodes). Features include actions (6 float32 values for robot joint positions: shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, and gripper), observation states (same 6 joint positions as actions), and image observations from three perspectives: top camera, wrist camera, and side camera, each with a resolution of 480x640, 3 channels, and video codec av1. Additionally, metadata such as timestamp, frame index, episode index, index, and task index are included.