unilab.envs.manipulation.allegro_inhand.grasp_gen

Allegro grasp-generation task built on top of rotation env.

Classes

AllegroRotationGrasp

AllegroRotationGraspCfg

AllegroRotationGraspCfg(scene: 'SceneCfg' = <factory>, sim_dt: 'float' = 0.005, max_episode_seconds: 'float' = 2.0, ctrl_dt: 'float' = 0.05, render_spacing: float = 1.0, render_offset_mode: str = 'grid', motrix_max_iterations: Optional[int] = None, post_step_forward_sensor: bool = False, noise_config: 'NoiseConfig' = <factory>, control_config: 'ControlConfig' = <factory>, reward_config: 'RewardConfigPPO' = <factory>, domain_rand: 'DomainRandConfig' = <factory>, rotation_axis: 'tuple[float, float, float]' = (0.0, 0.0, 1.0), grasp_cache_path: 'str' = 'caches/allegro_grasp_50k.npy', gen_grasp: 'bool' = True, grasp_collection_target: 'int' = 50000, grasp_auto_save: 'bool' = True, grasp_quality_check: 'bool' = True, grasp_min_contacts: 'int' = 2)

AllegroRotationGraspCfgAlias

alias of AllegroRotationGraspCfg

AllegroRotationGraspEnv

alias of AllegroRotationGrasp
class unilab.envs.manipulation.allegro_inhand.grasp_gen.AllegroRotationGraspCfg[source]

Bases: AllegroRotationPPOCfg

AllegroRotationGraspCfg(scene: ‘SceneCfg’ = <factory>, sim_dt: ‘float’ = 0.005, max_episode_seconds: ‘float’ = 2.0, ctrl_dt: ‘float’ = 0.05, render_spacing: float = 1.0, render_offset_mode: str = ‘grid’, motrix_max_iterations: Optional[int] = None, post_step_forward_sensor: bool = False, noise_config: ‘NoiseConfig’ = <factory>, control_config: ‘ControlConfig’ = <factory>, reward_config: ‘RewardConfigPPO’ = <factory>, domain_rand: ‘DomainRandConfig’ = <factory>, rotation_axis: ‘tuple[float, float, float]’ = (0.0, 0.0, 1.0), grasp_cache_path: ‘str’ = ‘caches/allegro_grasp_50k.npy’, gen_grasp: ‘bool’ = True, grasp_collection_target: ‘int’ = 50000, grasp_auto_save: ‘bool’ = True, grasp_quality_check: ‘bool’ = True, grasp_min_contacts: ‘int’ = 2)

Parameters:

  • scene (SceneCfg)

  • sim_dt (float)

  • max_episode_seconds (float)

  • ctrl_dt (float)

  • render_spacing (float)

  • render_offset_mode (str)

  • motrix_max_iterations (Optional[int])

  • post_step_forward_sensor (bool)

  • noise_config (NoiseConfig)

  • control_config (ControlConfig)

  • reward_config (RewardConfigPPO)

  • domain_rand (DomainRandConfig)

  • rotation_axis (tuple[float, float, float])

  • grasp_cache_path (str)

  • gen_grasp (bool)

  • grasp_collection_target (int)

  • grasp_auto_save (bool)

  • grasp_quality_check (bool)

  • grasp_min_contacts (int)

max_episode_seconds: float = 2.0
gen_grasp: bool = True
grasp_collection_target: int = 50000
grasp_auto_save: bool = True
grasp_quality_check: bool = True
grasp_min_contacts: int = 2
__init__(scene=<factory>, sim_dt=0.005, max_episode_seconds=2.0, ctrl_dt=0.05, render_spacing=1.0, render_offset_mode='grid', motrix_max_iterations=None, post_step_forward_sensor=False, noise_config=<factory>, control_config=<factory>, reward_config=<factory>, domain_rand=<factory>, rotation_axis=(0.0, 0.0, 1.0), grasp_cache_path='caches/allegro_grasp_50k.npy', gen_grasp=True, grasp_collection_target=50000, grasp_auto_save=True, grasp_quality_check=True, grasp_min_contacts=2)
Parameters:

  • scene (SceneCfg)

  • sim_dt (float)

  • max_episode_seconds (float)

  • ctrl_dt (float)

  • render_spacing (float)

  • render_offset_mode (str)

  • motrix_max_iterations (Optional[int])

  • post_step_forward_sensor (bool)

  • noise_config (NoiseConfig)

  • control_config (ControlConfig)

  • reward_config (RewardConfigPPO)

  • domain_rand (DomainRandConfig)

  • rotation_axis (tuple[float, float, float])

  • grasp_cache_path (str)

  • gen_grasp (bool)

  • grasp_collection_target (int)

  • grasp_auto_save (bool)

  • grasp_quality_check (bool)

  • grasp_min_contacts (int)

class unilab.envs.manipulation.allegro_inhand.grasp_gen.AllegroRotationGrasp[source]

Bases: AllegroRotationPPO

Parameters:
__init__(cfg, num_envs=1, backend_type='mujoco')[source]
Parameters:
apply_action(actions, state)[source]

子类实现：action → ctrl

Parameters:
Return type:

ndarray

update_state(state)[source]

子类实现：计算 obs/reward/terminated

Parameters:

state (NpEnvState)

Return type:

NpEnvState

close()[source]

关闭环境

Return type:

None

unilab.envs.manipulation.allegro_inhand.grasp_gen.AllegroRotationGraspEnv

alias of AllegroRotationGrasp

unilab.envs.manipulation.allegro_inhand.grasp_gen.AllegroRotationGraspCfgAlias

alias of AllegroRotationGraspCfg