rollout_dagger_smolvla_20260515_150001
收藏Hugging Face2026-05-15 更新2026-05-15 收录
下载链接:
https://huggingface.co/datasets/robot-learning-team43/rollout_dagger_smolvla_20260515_150001
下载链接
链接失效反馈官方服务:
资源简介:
该数据集由LeRobot创建,专注于机器人学领域。它包含一个单一episode,总时长为286帧,帧率为30 fps。数据特征包括机器人的动作(action)和状态观察(observation.state),涉及6个关节位置:肩部平移、肩部升降、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置。观察数据还包括来自camera1的图像,分辨率为480x640像素,3通道,视频编码为av1。数据集还包含干预标志(intervention)、时间戳、帧索引、episode索引、索引和任务索引等元数据。机器人类型为so_follower,数据以Parquet文件格式存储,总数据文件大小为100 MB,视频文件大小为200 MB。该数据集适用于训练和评估机器人控制模型。
This dataset is created by LeRobot and focuses on the field of robotics. It contains a single episode with a total duration of 286 frames at a frame rate of 30 fps. The data features include robot actions (action) and state observations (observation.state), involving 6 joint positions: shoulder translation, shoulder elevation, elbow flexion, wrist flexion, wrist rotation, and gripper position. Observation data also includes images from camera1 with a resolution of 480x640 pixels, 3 channels, and video encoding as av1. The dataset includes metadata such as intervention flags (intervention), timestamps, frame indices, episode indices, indices, and task indices. The robot type is so_follower, and the data is stored in Parquet format, with a total data file size of 100 MB and a video file size of 200 MB. This dataset is suitable for training and evaluating robot control models.
提供机构:
robot-learning-team43
创建时间:
2026-05-15
原始信息汇总
数据集概述
- 数据集名称: rollout_dagger_smolvla_20260515_150001
- 许可证: Apache-2.0
- 任务类别: 机器人学(robotics)
- 创建工具: 基于 LeRobot 框架构建
数据集结构
- 总集数(episodes): 1
- 总帧数(frames): 286
- 总任务数(tasks): 1
- 帧率(FPS): 30
- 数据类型: 包含动作(action)、状态观测(observation.state)、图像观测(observation.images.camera1)、干预标志(intervention)等
- 数据格式: 数据文件采用 Parquet 格式,视频文件采用 MP4 格式(AV1 编码)
特征字段说明
| 字段 | 数据类型 | 形状 | 说明 |
|---|---|---|---|
| action | float32 | [6] | 6维动作指令(如 shoulder_pan.pos 等) |
| observation.state | float32 | [6] | 6维关节状态(同动作名称) |
| observation.images.camera1 | video | [480, 640, 3] | 来自 camera1 的视频图像(480x640,3通道,每秒30帧) |
| intervention | bool | [1] | 是否为人工干预步骤 |
| timestamp | float32 | [1] | 时间戳 |
| frame_index | int64 | [1] | 帧索引 |
| episode_index | int64 | [1] | 集索引 |
| index | int64 | [1] | 全局索引 |
| task_index | int64 | [1] | 任务索引 |
数据划分
- 训练集(train): 包含全部数据(索引范围 0:1)
物理存储
- 数据文件大小: 约 100 MB
- 视频文件大小: 约 200 MB
机器人类型
- 机器人: so_follower(SO系列跟随机器人)



