Aneesha/eval_so101_blueblock_run50

该数据集是使用LeRobot工具创建的机器人学数据集，专门用于评估或训练机器人任务。数据集包含1个完整的情节（episode），共277帧数据，帧率为30fps。数据格式为parquet文件，并包含视频文件。特征包括机器人的动作（action）和观察状态（observation.state），其中动作和状态均涉及6个关节位置（如肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部滚动和夹爪位置）。观察部分还包含两个摄像头视角的图像（正面和侧面），分辨率为480x640，3通道彩色视频，编码为av1。数据集适用于机器人控制、模仿学习或强化学习任务。

This dataset is a robotics dataset created using the LeRobot tool, designed for evaluating or training robot tasks. It contains 1 complete episode with 277 frames at a frame rate of 30fps. The data is stored in parquet format and includes video files. Features include robot actions and observation states, both involving 6 joint positions (e.g., shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, gripper position). Observations also include images from two camera perspectives (front and side) with a resolution of 480x640, 3-channel color video encoded in av1. The dataset is suitable for robot control, imitation learning, or reinforcement learning tasks.