rofunc.learning.RofuncRL.tasks.omniisaacgymenv.franka_cabinet

1.  Module Contents

1.1.  Classes

FrankaCabinetOmniTask

1.2.  API

class rofunc.learning.RofuncRL.tasks.omniisaacgymenv.franka_cabinet.FrankaCabinetOmniTask(name, sim_config, env, offset=None)

Bases: rofunc.learning.RofuncRL.tasks.omniisaacgym.base.rl_task.RLTask

Initialization

set_up_scene(scene) → None

get_franka()

get_cabinet()

get_props()

init_data() → None

get_observations() → dict

pre_physics_step(actions) → None

reset_idx(env_ids)

post_reset()

calculate_metrics() → None

is_done() → None

compute_grasp_transforms(hand_rot, hand_pos, franka_local_grasp_rot, franka_local_grasp_pos, drawer_rot, drawer_pos, drawer_local_grasp_rot, drawer_local_grasp_pos)

compute_franka_reward(reset_buf: Tensor, progress_buf: Tensor, actions: Tensor, cabinet_dof_pos: Tensor, franka_grasp_pos: Tensor, drawer_grasp_pos: Tensor, franka_grasp_rot: Tensor, drawer_grasp_rot: Tensor, franka_lfinger_pos: Tensor, franka_rfinger_pos: Tensor, gripper_forward_axis: Tensor, drawer_inward_axis: Tensor, gripper_up_axis: Tensor, drawer_up_axis: Tensor, num_envs: int, dist_reward_scale: float, rot_reward_scale: float, around_handle_reward_scale: float, open_reward_scale: float, finger_dist_reward_scale: float, action_penalty_scale: float, distX_offset: float, max_episode_length: float, joint_positions: Tensor, finger_close_reward_scale) → Tuple[Tensor, Tensor]