svla_so100_sorting

该数据集为HuggingFace LeRobot格式机器人数据集。

原始信息汇总

数据集概述

基本信息

• 名称: svla_so100_sorting
• 许可证: Apache-2.0
• 任务类别: 机器人技术
• 标签: LeRobot, tutorial
• 创建工具: LeRobot

数据集结构

• 总集数: 9
• 总帧数: 6633
• 总任务数: 1
• 总视频数: 18
• 总块数: 1
• 块大小: 1000
• 帧率: 30 fps
• 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

• 动作:
  • 数据类型: float32
  • 形状: [6]
  • 名称: main_shoulder_pan, main_shoulder_lift, main_elbow_flex, main_wrist_flex, main_wrist_roll, main_gripper
• 观测状态:
  • 数据类型: float32
  • 形状: [6]
  • 名称: main_shoulder_pan, main_shoulder_lift, main_elbow_flex, main_wrist_flex, main_wrist_roll, main_gripper
• 观测图像（顶部）:
  • 数据类型: video
  • 形状: [480, 640, 3]
  • 名称: height, width, channels
  • 视频信息: fps=30.0, height=480, width=640, channels=3, codec=av1, pix_fmt=yuv420p, is_depth_map=false, has_audio=false
• 观测图像（腕部）:
  • 数据类型: video
  • 形状: [480, 640, 3]
  • 名称: height, width, channels
  • 视频信息: fps=30.0, height=480, width=640, channels=3, codec=av1, pix_fmt=yuv420p, is_depth_map=false, has_audio=false
• 时间戳:
  • 数据类型: float32
  • 形状: [1]
• 帧索引:
  • 数据类型: int64
  • 形状: [1]
• 集索引:
  • 数据类型: int64
  • 形状: [1]
• 索引:
  • 数据类型: int64
  • 形状: [1]
• 任务索引:
  • 数据类型: int64
  • 形状: [1]

相关论文

• 标题: SmolVLA: A vision-language-action model for affordable and efficient robotics
• arXiv链接: https://arxiv.org/abs/2506.01844

引用

bibtex @misc{shukor2025smolvlavisionlanguageactionmodelaffordable, title={SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics}, author={Mustafa Shukor and Dana Aubakirova and Francesco Capuano and Pepijn Kooijmans and Steven Palma and Adil Zouitine and Michel Aractingi and Caroline Pascal and Martino Russi and Andres Marafioti and Simon Alibert and Matthieu Cord and Thomas Wolf and Remi Cadene}, year={2025}, eprint={2506.01844}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2506.01844}, }