llm40
收藏Hugging Face2025-03-05 更新2025-04-08 收录
下载链接:
https://huggingface.co/datasets/zijian2022/llm40
下载链接
链接失效反馈官方服务:
资源简介:
该数据集使用LeRobot框架创建,主要涉及机器人技术领域。数据集包含10个情节,3231帧和20个视频,数据以parquet格式存储。数据集结构包括观察状态(如机器人电机控制)、来自笔记本电脑和手机的图像数据、动作数据以及各种索引(如情节索引、帧索引等)。具体特征包括6个浮点型电机状态、480x640x3的视频图像数据(来自笔记本电脑和手机),以及时间戳、任务索引等辅助信息。
This dataset was developed using the LeRobot framework, with its core focus on the field of robotics. It consists of 10 episodes, 3231 frames, and 20 videos, with all data stored in Parquet format. The dataset structure encompasses observation states (e.g., robot motor controls), image data captured by laptops and mobile phones, action data, and various indices including episode index, frame index, and others. Specific features include 6 floating-point motor states, 480×640×3 video image data sourced from laptops and mobile phones, as well as auxiliary information such as timestamps and task indices.
提供机构:
zijian2022
创建时间:
2025-03-05
原始信息汇总
数据集概述
基本信息
- 许可证: Apache-2.0
- 任务类别: 机器人学 (Robotics)
- 标签: LeRobot, so100
- 代码库版本: v2.0
- 机器人类型: so100
数据集结构
- 总集数: 10
- 总帧数: 3231
- 总任务数: 1
- 总视频数: 20
- 总块数: 1
- 块大小: 1000
- 帧率 (FPS): 30
- 分割: 训练集 (train) 包含0到10集
数据文件路径
- 数据路径:
data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet - 视频路径:
videos/chunk-{episode_chunk:03d}/{video_key}/episode_{episode_index:06d}.mp4
特征描述
观测数据
- observation.state:
- 数据类型: float32
- 形状: [6]
- 名称: shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper
- observation.images.laptop:
- 数据类型: video
- 形状: [480, 640, 3]
- 名称: height, width, channels
- 视频信息:
- FPS: 30.0
- 编解码器: av1
- 像素格式: yuv420p
- 深度图: false
- 音频: false
- observation.images.phone:
- 数据类型: video
- 形状: [480, 640, 3]
- 名称: height, width, channels
- 视频信息:
- FPS: 30.0
- 编解码器: av1
- 像素格式: yuv420p
- 深度图: false
- 音频: false
动作数据
- action:
- 数据类型: float32
- 形状: [6]
- 名称: shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper
其他数据
- episode_index: int64, 形状 [1]
- frame_index: int64, 形状 [1]
- timestamp: float32, 形状 [1]
- next.done: bool, 形状 [1]
- index: int64, 形状 [1]
- task_index: int64, 形状 [1]
引用信息
- BibTeX: [More Information Needed]



