mani_skill.vector.wrappers.sb3#

Classes#

ManiSkillSB3VectorEnv

A wrapper for to make ManiSkill parallel simulation compatible with SB3 VecEnv and auto adds the monitor wrapper

Functions#

select_index_from_dict(data, i)

Module Contents#

class mani_skill.vector.wrappers.sb3.ManiSkillSB3VectorEnv(env)[source]#

Bases: stable_baselines3.common.vec_env.base_vec_env.VecEnv

A wrapper for to make ManiSkill parallel simulation compatible with SB3 VecEnv and auto adds the monitor wrapper

Parameters:

env (mani_skill.envs.sapien_env.BaseEnv) –

close()[source]#
Return type:

None

env_is_wrapped(wrapper_class, indices=None)[source]#
Parameters:

  • wrapper_class (Type[gymnasium.Wrapper]) –

  • indices (stable_baselines3.common.vec_env.base_vec_env.VecEnvIndices) –

Return type:

list[bool]

env_method(method_name, *method_args, indices=None, **method_kwargs)[source]#
Parameters:

  • method_name (str) –

  • indices (stable_baselines3.common.vec_env.base_vec_env.VecEnvIndices) –

Return type:

list[Any]

get_attr(attr_name, indices=None)[source]#
Parameters:

  • attr_name (str) –

  • indices (stable_baselines3.common.vec_env.base_vec_env.VecEnvIndices) –

Return type:

list[Any]

reset()[source]#
Return type:

stable_baselines3.common.vec_env.base_vec_env.VecEnvObs

seed(seed=None)[source]#
Parameters:

seed (Optional[int]) –

Return type:

list[Union[None, int]]

set_attr(attr_name, value, indices=None)[source]#
Parameters:

  • attr_name (str) –

  • value (Any) –

  • indices (stable_baselines3.common.vec_env.base_vec_env.VecEnvIndices) –

Return type:

None

step_async(actions)[source]#
Parameters:

actions (numpy.ndarray) –

Return type:

None

step_wait()[source]#
Return type:

stable_baselines3.common.vec_env.base_vec_env.VecEnvStepReturn

_env[source]#
_last_seed = None[source]#
property base_env: mani_skill.envs.sapien_env.BaseEnv[source]#
Return type:

mani_skill.envs.sapien_env.BaseEnv

episode_lengths: torch.Tensor[source]#
episode_returns: torch.Tensor[source]#
episode_times: torch.Tensor[source]#
t_start[source]#
total_steps = 0[source]#
mani_skill.vector.wrappers.sb3.select_index_from_dict(data, i)[source]#
Parameters:

  • data (Union[dict, numpy.ndarray]) –

  • i (int) –