dreamer-robot/liftplace-ee-xyz-500-v2
This dataset was created using LeRobot. Dataset Description Language-conditioned smooth point-to-point Panda end-effector trajectories in robosuite LiftPlace. Episodes are continuous within each collection_segment_index: the cube and green placement goal change between episodes without changing the robot state. If planning or execution becomes unsafe, the incomplete episode is deleted and the Panda is reset to home before the next segment begins. Each fixed-orientation straight… See the full description on the dataset page: https://huggingface.co/datasets/dreamer-robot/liftplace-ee-xyz-500-v2.
0667
This dataset was created using LeRobot.
Dataset Description
Language-conditioned smooth point-to-point Panda end-effector trajectories in robosuite LiftPlace. Episodes are continuous within each collection_segment_index: the cube and green placement goal change between episodes without changing the robot state. If planning or execution becomes unsafe, the incomplete episode is deleted and the Panda is reset to home before the next segment begins. Each fixed-orientation straight path is screened with bounded damped-least-squares IK before collection, then tracked as a quintic minimum-jerk Cartesian reference. The green placement goal is scene context and is deliberately distinct from the language-conditioned end-effector target.
Reproduction metadata
- Dreamer Git commit:
766bdc378e2a15de7de30d1f1c28ab6991ea1bf6 - Dreamer worktree dirty:
False - Dreamer worktree SHA-256:
None - Dreamer provenance drift during collection:
False - Dreamer end-of-run Git commit:
766bdc378e2a15de7de30d1f1c28ab6991ea1bf6 - Dreamer end-of-run worktree dirty:
False - Dreamer end-of-run worktree SHA-256:
None uv.lockSHA-256:e7a7667144a24698a610ad8f0135b3f745c0cf807d777edee540bb7975ceb8d4- LeRobot release:
0.6.0 - LeRobot release commit:
30da8e687a6dfc617fcd94afc367ac7071c376ce - LeRobot dataset format:
v3.0 - Python:
3.13 - Seed:
0 - Saved collection segments: 24
- Rejected episode attempts / recovery resets: 3
- Maximum permitted recovery resets: 100
- Regular home-reset interval: 25 episodes
- Workspace: x=(-0.3, 0.3), y=(-0.3, 0.3), z=(0.95, 1.25) metres
- Target quantization: 0.001 m
- Minimum target displacement: 0.05 m
- Maximum target displacement: 0.35 m
- Endpoint tolerance: 0.002 m
- Endpoint joint-velocity tolerance: 0.01 rad/s (L2 norm)
- Motion profile: quintic minimum jerk (
10t^3 - 15t^4 + 6t^5) - Maximum Cartesian reference speed: 0.2 m/s
- Maximum Cartesian reference acceleration: 0.5 m/s^2
- Maximum Cartesian reference jerk: 2.0 m/s^3
- Minimum trajectory length: 20 steps
- Maximum trajectory length: 300 steps
- Minimum joint-limit margin: 0.2 rad
- Minimum pose-Jacobian singular value: 0.08
- Maximum pose-Jacobian condition number: 25.0
- Preflight joint-limit margin: 0.25 rad
- Preflight minimum pose-Jacobian singular value: 0.1
- Preflight maximum pose-Jacobian condition number: 20.0
- Shadow-rollout joint-limit margin: 0.22 rad
- Shadow-rollout minimum pose-Jacobian singular value: 0.09
- Shadow-rollout maximum pose-Jacobian condition number: 23.0
- Kinematic path samples: 16
- Control/video frequency: 20 Hz
- Homepage: https://huggingface.co/datasets/dreamer-robot/liftplace-ee-xyz-500-v2
- Paper: [More Information Needed]
- License: apache-2.0
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"fps": 20,
"features": {
"observation.images.agentview": {
"dtype": "video",
"shape": [
256,
256,
3
],
"names": [
"height",
"width",
"channel"
],
"info": {
"video.height": 256,
"video.width": 256,
"video.codec": "av1",
"video.pix_fmt": "yuv420p",
"video.fps": 20,
"video.channels": 3,
"has_audio": false,
"video.g": 2,
"video.crf": 30,
"video.preset": 12,
"video.fast_decode": 0,
"video.video_backend": "pyav",
"video.extra_options": {},
"is_depth_map": false
}
},
"observation.images.wrist": {
"dtype": "video",
"shape": [
256,
256,
3
],
"names": [
"height",
"width",
"channel"
],
"info": {
"video.height": 256,
"video.width": 256,
"video.codec": "av1",
"video.pix_fmt": "yuv420p",
"video.fps": 20,
"video.channels": 3,
"has_audio": false,
"video.g": 2,
"video.crf": 30,
"video.preset": 12,
"video.fast_decode": 0,
"video.video_backend": "pyav",
"video.extra_options": {},
"is_depth_map": false
}
},
"observation.state": {
"dtype": "float32",
"shape": [
8
],
"names": [
"ee_position_x",
"ee_position_y",
"ee_position_z",
"ee_axis_angle_x",
"ee_axis_angle_y",
"ee_axis_angle_z",
"gripper_left_qpos",
"gripper_right_qpos"
]
},
"action": {
"dtype": "float32",
"shape": [
7
],
"names": [
"normalized_ee_position_delta_x",
"normalized_ee_position_delta_y",
"normalized_ee_position_delta_z",
"normalized_ee_rotation_delta_x",
"normalized_ee_rotation_delta_y",
"normalized_ee_rotation_delta_z",
"normalized_gripper_command"
]
},
"observation.ee_target_position": {
"dtype": "float32",
"shape": [
3
],
"names": [
"ee_target_x",
"ee_target_y",
"ee_target_z"
]
},
"observation.cartesian_reference_position": {
"dtype": "float32",
"shape": [
3
],
"names": [
"ee_reference_x",
"ee_reference_y",
"ee_reference_z"
]
},
"observation.distance_to_ee_target": {
"dtype": "float32",
"shape": [
1
],
"names": [
"distance_to_target"
]
},
"observation.cube_position": {
"dtype": "float32",
"shape": [
3
],
"names": [
"cube_x",
"cube_y",
"cube_z"
]
},
"observation.green_goal_position": {
"dtype": "float32",
"shape": [
3
],
"names": [
"green_goal_x",
"green_goal_y",
"green_goal_z"
]
},
"observation.joint_position": {
"dtype": "float32",
"shape": [
7
],
"names": [
"panda_joint_1_position",
"panda_joint_2_position",
"panda_joint_3_position",
"panda_joint_4_position",
"panda_joint_5_position",
"panda_joint_6_position",
"panda_joint_7_position"
]
},
"observation.joint_velocity": {
"dtype": "float32",
"shape": [
7
],
"names": [
"panda_joint_1_velocity",
"panda_joint_2_velocity",
"panda_joint_3_velocity",
"panda_joint_4_velocity",
"panda_joint_5_velocity",
"panda_joint_6_velocity",
"panda_joint_7_velocity"
]
},
"observation.pose_jacobian_singular_values": {
"dtype": "float32",
"shape": [
6
],
"names": [
"pose_jacobian_sigma_1",
"pose_jacobian_sigma_2",
"pose_jacobian_sigma_3",
"pose_jacobian_sigma_4",
"pose_jacobian_sigma_5",
"pose_jacobian_sigma_6"
]
},
"observation.pose_jacobian_condition_number": {
"dtype": "float32",
"shape": [
1
],
"names": [
"pose_jacobian_condition_number"
]
},
"observation.minimum_joint_limit_margin": {
"dtype": "float32",
"shape": [
1
],
"names": [
"minimum_joint_limit_margin_rad"
]
},
"collection_segment_index": {
"dtype": "int64",
"shape": [
1
],
"names": [
"collection_segment_index"
]
},
"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
}
},
"total_episodes": 500,
"total_frames": 27652,
"total_tasks": 500,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"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",
"robot_type": "panda",
"splits": {
"train": "0:500"
}
}Citation
BibTeX:
[More Information Needed]