Episodes Preview SO-101 Visualizer
112 episodes · 30 fps · 2 cameras · 480×480 h264

This dataset was created using LeRobot.

Dataset Description

Robot arm playing chess: teleoperated demos of an SO-101 arm picking chess pieces off a real board.

The first cleaned training set. 112 episodes, 13,526 frames, wrist + overhead cameras. Trained Dev-devadath/so101-chess-act-pick-v1.

Part of so101-chess: an SO-101 arm that plays chess on a real board. Everything except the grasp is inverse kinematics; a learned ACT policy does only the pick. Every episode starts with the arm hovering 8 cm above one piece, jaws open and pointing down, and ends with the piece lifted. Demos were teleoperated with an SO-101 leader arm.

  • 30 fps; state and action are 6-D (5 joints + gripper, lerobot degrees, so_follower)
  • observation.images.wrist: gripper camera, 640x480
  • observation.images.top (where present): overhead camera warped to a top-down board view, 480x480, a1 bottom-left
  • Gripper: red TPU V-groove pads glued to both jaws (except so101-chess-pick-v0). The pads are in the wrist view, so a policy trained here expects them.
  • Board: 49 mm squares, ranks 1-5 reachable. White is on the arm's side.
  • Recorded with lerobot at commit 5c43fa1c (0.5.2 dev). Load with video_backend="pyav".

It is episodes 0-123 of Dev-devadath/so101-chess-pick-raw minus 12 bad ones (6, 18, 48, 50, 83, 89, 101, 102, 107, 110, 120, 121: head grabs, no lift, fingertips >4 mm into the mat).

Known issues

  • Recorded on a near-empty board. The far ranks were empty in every overhead frame. The policy trained on it picked 10 of 10 in isolated tests, then in a real game the black pieces on ranks 7-8 shifted its grasps about half a square toward the arm: it had learned to aim partly from the overhead view of an empty board. The later sets drop the overhead camera and add full-board demos.

  • Mixed video codecs. delete_episodes re-encoded the files it cut as AV1 and copied the rest as h264. Both decode with pyav, but do not record more episodes into this copy: record into Dev-devadath/so101-chess-pick-raw and rebuild.

  • h1 is weak: a cable sagged into the wrist camera at that pose during play, which never happened in the demos.

  • Homepage: https://github.com/Dev-devadath/so101-chess

  • Paper: none

  • License: apache-2.0

Dataset Structure

meta/info.json:

{
    "codebase_version": "v3.0",
    "robot_type": "so_follower",
    "total_episodes": 112,
    "total_frames": 13526,
    "total_tasks": 1,
    "chunks_size": 1000,
    "data_files_size_in_mb": 100,
    "video_files_size_in_mb": 200,
    "fps": 30,
    "splits": {
        "train": "0:112"
    },
    "data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
    "video_path": "videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4",
    "features": {
        "action": {
            "dtype": "float32",
            "names": [
                "shoulder_pan.pos",
                "shoulder_lift.pos",
                "elbow_flex.pos",
                "wrist_flex.pos",
                "wrist_roll.pos",
                "gripper.pos"
            ],
            "shape": [
                6
            ]
        },
        "observation.state": {
            "dtype": "float32",
            "names": [
                "shoulder_pan.pos",
                "shoulder_lift.pos",
                "elbow_flex.pos",
                "wrist_flex.pos",
                "wrist_roll.pos",
                "gripper.pos"
            ],
            "shape": [
                6
            ]
        },
        "observation.images.wrist": {
            "dtype": "video",
            "shape": [
                480,
                640,
                3
            ],
            "names": [
                "height",
                "width",
                "channels"
            ],
            "info": {
                "video.height": 480,
                "video.width": 640,
                "video.codec": "h264",
                "video.pix_fmt": "yuv420p",
                "video.is_depth_map": false,
                "video.fps": 30,
                "video.channels": 3,
                "has_audio": false
            }
        },
        "observation.images.top": {
            "dtype": "video",
            "shape": [
                480,
                480,
                3
            ],
            "names": [
                "height",
                "width",
                "channels"
            ],
            "info": {
                "video.height": 480,
                "video.width": 480,
                "video.codec": "h264",
                "video.pix_fmt": "yuv420p",
                "video.is_depth_map": false,
                "video.fps": 30,
                "video.channels": 3,
                "has_audio": false
            }
        },
        "timestamp": {
            "dtype": "float32",
            "shape": [
                1
            ],
            "names": null
        },
        "frame_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "episode_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "task_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        }
    }
}

Citation

BibTeX:

none
Downloads last month
55

Models trained or fine-tuned on Dev-devadath/so101-chess-pick-v1