omy_pnp_language

该数据集是一个机器人技术相关的数据集，采用Apache-2.0许可证发布，与LeRobot项目相关联。数据集包含3个完整的情节（episodes），共计894帧数据，覆盖2个不同的任务。数据以parquet格式存储，结构详细信息通过meta/info.json文件提供。数据集包含多种观测和动作数据，如图像观测（256x256像素的RGB图像）、手腕图像观测、状态观测（6维浮点向量）、动作数据（7维浮点向量）以及对象初始化状态（9维浮点向量）。此外，还包括时间戳、帧索引、情节索引等元数据。数据集适用于机器人学习、行为克隆、强化学习等任务。数据按训练集划分，所有情节均用于训练。

This is a robotics-related dataset released under the Apache-2.0 license and associated with the LeRobot project. It contains 3 complete episodes, totaling 894 frames, covering 2 distinct tasks. The data is stored in Parquet format, and its detailed structural specifications are provided in the meta/info.json file. The dataset includes various types of observation and action data: 256×256 pixel RGB image observations, wrist camera image observations, state observations (6-dimensional floating-point vectors), action data (7-dimensional floating-point vectors), and object initialization states (9-dimensional floating-point vectors). It also incorporates metadata such as timestamps, frame indices, and episode indices. This dataset is applicable to tasks including robotic learning, behavioral cloning, and reinforcement learning. The dataset is split into a training set, with all episodes used for training.