eval_dit_posttrainv2_clare_dit_0_2_real_2_on_real_1_seed1000

该数据集是通过LeRobot创建的，主要用于机器人技术领域。数据集包含10个episodes，总计3877帧，涉及1个任务。数据以parquet文件格式存储，包含多种特征，如观察图像（primary和wrist）、状态（笛卡尔坐标、夹持器状态、关节状态等）、动作（笛卡尔坐标和夹持器状态）以及时间戳和索引信息。观察图像的分辨率为256x256，帧率为15fps。数据集的结构和特征详细信息可在meta/info.json文件中找到。This dataset was created using LeRobot, primarily targeting the robotics domain. It contains 10 episodes, totaling 3877 frames, and focuses on one single task. The dataset is stored in Parquet file format, including various features such as observation images (primary and wrist), states (Cartesian coordinates, gripper status, joint states, etc.), actions (Cartesian coordinates and gripper status), as well as timestamp and index information. The observation images have a resolution of 256×256 and a frame rate of 15 fps. Detailed information about the dataset structure and features can be found in the meta/info.json file.