Residual-Copilot
Collection
Datasets and checkpoints used for the paper: Efficient and Reliable Teleoperation through Real-to-Sim-to-Real Shared Autonomy • 9 items • Updated • 3
repo_id stringclasses 3
values | root stringclasses 3
values | tasks stringclasses 1
value | episodes null | stats dict | info dict | revision null | writer null | latest_episode dict | metadata_buffer listlengths 0 0 | metadata_buffer_size int64 10 10 |
|---|---|---|---|---|---|---|---|---|---|---|
0121_gearmesh_teleop_aug_2000 | /home/shuo/.cache/huggingface/lerobot/0121_gearmesh_teleop_aug_2000 | task_index
gear_mesh 0 | null | {
"observation.state": {
"min": "[ 0.27445084 -0.1873982 0.01454373 -0.06643009 0.9861852 -0.09685721\n -0.08673099 0. -0.08144522 -0.14883211 -0.0720678 -0.5000784\n -0.37609267 -0.24783607]",
"max": "[0.46427724 0.09578571 0.1017628 0.05926742 0.99999887 0.16411626\n 0.07516461 0.6999233 0.1... | {
"codebase_version": "v3.0",
"robot_type": "xarm",
"total_episodes": 2000,
"total_frames": 441968,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 15,
"splits": {
"train": "0:2000"
},
"data_path": "data/chunk-{chunk_index:03d}/file-... | null | null | {
"episode_index": [
1999
],
"tasks": [
[
"gear_mesh"
]
],
"length": [
246
],
"data/chunk_index": [
0
],
"data/file_index": [
0
],
"dataset_from_index": [
441722
],
"dataset_to_index": [
441968
],
"stats/action/min": [
"[ 0.39020696 -0.10425197 0.... | [] | 10 |
0121_nutthread_teleop_aug_2000 | /home/shuo/.cache/huggingface/lerobot/0121_nutthread_teleop_aug_2000 | task_index
gear_mesh 0 | null | {
"observation.state": {
"min": "[ 0.29610214 -0.22155188 0.00928816 -0.05808281 0.77398705 -0.632532\n -0.05035634 0. -0.09044845 -0.13223644 -0.05587521 -0.49977875\n -0.3091898 -1.8603337 ]",
"max": "[0.47568086 0.0867549 0.13170591 0.04305074 0.9999993 0.27761367\n 0.04826248 0.39880952 0.08... | {
"codebase_version": "v3.0",
"robot_type": "xarm",
"total_episodes": 2000,
"total_frames": 786308,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 15,
"splits": {
"train": "0:2000"
},
"data_path": "data/chunk-{chunk_index:03d}/file-... | null | null | {
"episode_index": [
1999
],
"tasks": [
[
"gear_mesh"
]
],
"length": [
414
],
"data/chunk_index": [
0
],
"data/file_index": [
1
],
"dataset_from_index": [
785894
],
"dataset_to_index": [
786308
],
"stats/action/min": [
"[ 0.375505 -0.19700827 0.... | [] | 10 |
0121_peginsert_teleop_aug_2000 | /home/shuo/.cache/huggingface/lerobot/0121_peginsert_teleop_aug_2000 | task_index
gear_mesh 0 | null | {
"observation.state": {
"min": "[ 0.2834762 -0.14622664 0.02682373 -0.12823594 0.9884889 -0.09908803\n -0.14057824 0. -0.05548669 -0.10661928 -0.06390861 -0.3587433\n -0.3525084 -0.27447733]",
"max": "[0.45719755 0.08806431 0.13214421 0.11149003 0.99999976 0.1398283\n 0.0674132 0.9464286 0.10... | {
"codebase_version": "v3.0",
"robot_type": "xarm",
"total_episodes": 2000,
"total_frames": 426688,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 15,
"splits": {
"train": "0:2000"
},
"data_path": "data/chunk-{chunk_index:03d}/file-... | null | null | {
"episode_index": [
1999
],
"tasks": [
[
"gear_mesh"
]
],
"length": [
206
],
"data/chunk_index": [
0
],
"data/file_index": [
0
],
"dataset_from_index": [
426482
],
"dataset_to_index": [
426688
],
"stats/action/min": [
"[ 0.37087697 -0.09327577 0.... | [] | 10 |
Residual Copilot — Data
Teleoperation demonstration data for Residual Shared Autonomy. Used by the kNN pilot for nearest-neighbor action retrieval and for sampling episode initial states during environment resets.
Contents
teleop/
├── gearmesh_train_data.npy # GearMesh teleop trajectories
├── peginsert_train_data.npy # PegInsert teleop trajectories
├── nutthread_train_data.npy # NutThread teleop trajectories
├── *_train_data_aug.npy # Augmented versions (with noise/perturbations)
└── *_train_data_aug_meta.json # Augmentation metadata
Each .npy file is a dictionary keyed by episode ID. Each episode contains obs.fingertip_pos, obs.fingertip_quat, obs.fingertip_pos_rel_fixed, obs.fingertip_pos_rel_held, and action trajectories.
For DiffusionPolicy BC training, use the separate LeRobot-format datasets instead (e.g., shashuo0104/0126_gearmesh_expert_2000).
Usage
Data is auto-downloaded at runtime via resolve_hf(). See the main repo for setup instructions.
Related
| Repo | Type | Contents |
|---|---|---|
| residual_copilot_assets | dataset | Robot URDF/USD and object meshes |
| residual_copilot_models | model | BC pilot + RL residual copilot checkpoints |
- Downloads last month
- 55