fractal20220817_data

  • Description:

Table-top manipulation with 17 objects

Split Examples
'train' 87,212
  • Feature structure:
FeaturesDict({
    'aspects': FeaturesDict({
        'already_success': bool,
        'feasible': bool,
        'has_aspects': bool,
        'success': bool,
        'undesirable': bool,
    }),
    'attributes': FeaturesDict({
        'collection_mode': int64,
        'collection_mode_name': string,
        'data_type': int64,
        'data_type_name': string,
        'env': int64,
        'env_name': string,
        'location': int64,
        'location_name': string,
        'objects_family': int64,
        'objects_family_name': string,
        'task_family': int64,
        'task_family_name': string,
    }),
    'steps': Dataset({
        'action': FeaturesDict({
            'base_displacement_vector': Tensor(shape=(2,), dtype=float32),
            'base_displacement_vertical_rotation': Tensor(shape=(1,), dtype=float32),
            'gripper_closedness_action': Tensor(shape=(1,), dtype=float32, description=continuous gripper position),
            'rotation_delta': Tensor(shape=(3,), dtype=float32, description=rpy commanded orientation displacement, in base-relative frame),
            'terminate_episode': Tensor(shape=(3,), dtype=int32),
            'world_vector': Tensor(shape=(3,), dtype=float32, description=commanded end-effector displacement, in base-relative frame),
        }),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'observation': FeaturesDict({
            'base_pose_tool_reached': Tensor(shape=(7,), dtype=float32, description=end-effector base-relative position+quaternion pose),
            'gripper_closed': Tensor(shape=(1,), dtype=float32),
            'gripper_closedness_commanded': Tensor(shape=(1,), dtype=float32, description=continuous gripper position),
            'height_to_bottom': Tensor(shape=(1,), dtype=float32, description=height of end-effector from ground),
            'image': Image(shape=(256, 320, 3), dtype=uint8),
            'natural_language_embedding': Tensor(shape=(512,), dtype=float32),
            'natural_language_instruction': string,
            'orientation_box': Tensor(shape=(2, 3), dtype=float32),
            'orientation_start': Tensor(shape=(4,), dtype=float32),
            'robot_orientation_positions_box': Tensor(shape=(3, 3), dtype=float32),
            'rotation_delta_to_go': Tensor(shape=(3,), dtype=float32, description=rotational displacement from current orientation to target),
            'src_rotation': Tensor(shape=(4,), dtype=float32),
            'vector_to_go': Tensor(shape=(3,), dtype=float32, description=displacement from current end-effector position to target),
            'workspace_bounds': Tensor(shape=(3, 3), dtype=float32),
        }),
        'reward': Scalar(shape=(), dtype=float32),
    }),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
aspects FeaturesDict Session Aspects for crowdcompute ratings
aspects/already_success Tensor bool
aspects/feasible Tensor bool
aspects/has_aspects Tensor bool
aspects/success Tensor bool
aspects/undesirable Tensor bool
attributes FeaturesDict
attributes/collection_mode Tensor int64
attributes/collection_mode_name Tensor string
attributes/data_type Tensor int64
attributes/data_type_name Tensor string
attributes/env Tensor int64
attributes/env_name Tensor string
attributes/location Tensor int64
attributes/location_name Tensor string
attributes/objects_family Tensor int64
attributes/objects_family_name Tensor string
attributes/task_family Tensor int64
attributes/task_family_name Tensor string
steps Dataset
steps/action FeaturesDict
steps/action/base_displacement_vector Tensor (2,) float32
steps/action/base_displacement_vertical_rotation Tensor (1,) float32
steps/action/gripper_closedness_action Tensor (1,) float32 continuous gripper position
steps/action/rotation_delta Tensor (3,) float32 rpy commanded orientation displacement, in base-relative frame
steps/action/terminate_episode Tensor (3,) int32
steps/action/world_vector Tensor (3,) float32 commanded end-effector displacement, in base-relative frame
steps/is_first Tensor bool
steps/is_last Tensor bool
steps/is_terminal Tensor bool
steps/observation FeaturesDict
steps/observation/base_pose_tool_reached Tensor (7,) float32 end-effector base-relative position+quaternion pose
steps/observation/gripper_closed Tensor (1,) float32
steps/observation/gripper_closedness_commanded Tensor (1,) float32 continuous gripper position
steps/observation/height_to_bottom Tensor (1,) float32 height of end-effector from ground
steps/observation/image Image (256, 320, 3) uint8
steps/observation/natural_language_embedding Tensor (512,) float32
steps/observation/natural_language_instruction Tensor string
steps/observation/orientation_box Tensor (2, 3) float32
steps/observation/orientation_start Tensor (4,) float32
steps/observation/robot_orientation_positions_box Tensor (3, 3) float32
steps/observation/rotation_delta_to_go Tensor (3,) float32 rotational displacement from current orientation to target
steps/observation/src_rotation Tensor (4,) float32
steps/observation/vector_to_go Tensor (3,) float32 displacement from current end-effector position to target
steps/observation/workspace_bounds Tensor (3, 3) float32
steps/reward Scalar float32
  • Citation:
@article{brohan2022rt,
  title={Rt-1: Robotics transformer for real-world control at scale},
  author={Brohan, Anthony and Brown, Noah and Carbajal, Justice and Chebotar, Yevgen and Dabis, Joseph and Finn, Chelsea and Gopalakrishnan, Keerthana and Hausman, Karol and Herzog, Alex and Hsu, Jasmine and others},
  journal={arXiv preprint arXiv:2212.06817},
  year={2022}
}