curobo.wrap.model.curobo_robot_world module

This module has differentiable layers built from CuRobo’s core features for use in Pytorch.

class curobo.wrap.model.curobo_robot_world.CuroboRobotWorldConfig(kinematics: curobo.cuda_robot_model.cuda_robot_model.CudaRobotModel, sampler: curobo.util.sample_lib.HaltonGenerator, bound_scale: torch.Tensor, bound_cost: curobo.rollout.cost.bound_cost.BoundCost, pose_cost: curobo.rollout.cost.pose_cost.PoseCost, self_collision_cost: curobo.rollout.cost.self_collision_cost.SelfCollisionCost | None = None, collision_cost: curobo.rollout.cost.primitive_collision_cost.PrimitiveCollisionCost | None = None, collision_constraint: curobo.rollout.cost.primitive_collision_cost.PrimitiveCollisionCost | None = None, world_model: curobo.geom.sdf.world.WorldCollision | None = None, rejection_ratio: int = 10, tensor_args: curobo.types.base.TensorDeviceType = TensorDeviceType(device=device(type='cuda', index=0), dtype=torch.float32), contact_distance: float = 0.0)

Bases: RobotWorldConfig

Parameters:
class curobo.wrap.model.curobo_robot_world.CuroboRobotWorld(config)

Bases: RobotWorld

Parameters:

config (CuroboRobotWorldConfig) –