Dev-devadath's picture
Upload README.md with huggingface_hub
b81fa98 verified
|
Raw History Blame Contribute Delete
6.7 kB
---
license: apache-2.0
task_categories:
- robotics
tags:
- LeRobot
- chess
- so101
- so-101
- robot-arm
- manipulation
- pick-and-place
- teleoperation
- act
- chess-robot
configs:
- config_name: default
data_files: data/*/*.parquet
pretty_name: SO-101 Chess Pick, raw recording
---
This dataset was created using [LeRobot](https://github.com/huggingface/lerobot).
<a class="flex" href="https://huggingface.co/spaces/lerobot/visualize_dataset?path=Dev-devadath/so101-chess-pick-raw">
<img class="block dark:hidden" src="https://huggingface.co/datasets/huggingface/badges/resolve/main/visualize-this-dataset-xl.svg"/>
<img class="hidden dark:block" src="https://huggingface.co/datasets/huggingface/badges/resolve/main/visualize-this-dataset-xl-dark.svg"/>
</a>
## Dataset Description
**Robot arm playing chess: teleoperated demos of an SO-101 arm picking chess pieces off a real board.**
**The raw recording: every pick demo of the current round.** 271 episodes, 39,593 frames, wrist + overhead cameras.
Part of [so101-chess](https://github.com/Dev-devadath/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"`.
## What is in it
Recorded in five sessions, in this order (episode indices):
| Episodes | What |
|---|---|
| 0-123 | First round on a near-empty board: pawns on all 39 squares, tall pieces, knights/bishops on ranks 4-5 |
| 124-133 | Extra demos on the corner squares d1 and h1 |
| 134-168 | The full starting position, pawns picked with neighbours on both sides |
| 169-228 | Full board, every piece on ranks 1-2 |
| 229-270 | Mid-board (ranks 3-5), including pieces the arm had placed itself (off-centre) |
## Known issues
- **24 bad episodes are still in here**: 6, 18, 48, 50, 83, 89, 101, 102, 107, 110, 120, 121, 135, 140,
141, 142, 146, 147, 149, 151, 153, 155, 211, 220. Head grabs, demos that never lifted the piece, and
demos where the fingertips pressed more than 4 mm into the mat. Remove them before training
(`lerobot.datasets.dataset_tools.delete_episodes`); `Dev-devadath/so101-chess-pick-wrist-v4` is this set
with them removed.
- **Episodes 0-133 are on a near-empty board.** A policy trained on those alone does not transfer to a
real game (see `Dev-devadath/so101-chess-pick-v1`).
- The sideways grasp offsets at b1 are 10-16 mm, larger than elsewhere: the taught position of that
square was probably off when it was recorded.
- **Homepage:** https://github.com/Dev-devadath/so101-chess
- **Paper:** none
- **License:** apache-2.0
## Dataset Structure
[meta/info.json](meta/info.json):
```json
{
"codebase_version": "v3.0",
"robot_type": "so_follower",
"total_episodes": 271,
"total_frames": 39593,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 30,
"splits": {
"train": "0:271"
},
"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:**
```bibtex
none
```