eval_vla_state_modulation_v6_nomodulation_nodense

该数据集由LeRobot创建，主要用于机器人领域。数据集包含多种数据类型，如动作（包括机械臂末端执行器的位置、旋转和夹爪位置）、观测状态（包括机械臂末端执行器的位置、旋转和夹爪位置）、图像（手腕和前端摄像头拍摄的图像，尺寸为256x256x3）、视觉特征（VFM和VLM特征，分别具有不同的形状和维度）、技能特征、物体特征等。数据以parquet格式存储，视频以mp4格式存储。数据集的结构信息详细描述了每个字段的数据类型、形状和名称。

This dataset was developed by LeRobot, and is primarily designed for robotics research and applications. It encompasses multiple data types, including actions (comprising the position, rotation, and gripper position of the robotic arm end-effector), observation states (also consisting of the position, rotation, and gripper position of the robotic arm end-effector), images captured by the wrist and front-facing cameras with a resolution of 256×256×3, visual features (VFM and VLM features with distinct shapes and dimensions respectively), skill features, and object features. Structured data is stored in Parquet format, while video data is stored in MP4 format. The dataset's structural information thoroughly details the data type, shape, and name of each field.