features_extractor_kwargs: Optional[Dict[str, Any]] = None,
normalize_images: bool = True,
optimizer_class: Type[th.optim.Optimizer] = th.optim.Adam,
optimizer_kwargs: Optional[Dict[str, Any]] = None,
n_critics: int = 2,
share_features_extractor: bool = True,
):
super(CnnPolicy, self).__init__(
observation_space,
action_space,
lr_schedule,
net_arch,
activation_fn,
use_sde,
log_std_init,
sde_net_arch,
use_expln,
clip_mean,
features_extractor_class,
features_extractor_kwargs,
normalize_images,
optimizer_class,
optimizer_kwargs,
n_critics,
share_features_extractor,
)
register_policy("SAC", "MlpPolicy", MlpPolicy)
register_policy("SAC", "CnnPolicy", CnnPolicy)
"""
def __init__(
self,
observation_space: spaces.Space,
action_space: spaces.Space,
lr_schedule: Schedule,
net_arch: Optional[List[int]] = None,
activation_fn: Type[nn.Module] = nn.ReLU,
features_extractor_class: Type[BaseFeaturesExtractor] = NatureCNN,
features_extractor_kwargs: Optional[Dict[str, Any]] = None,
normalize_images: bool = True,
optimizer_class: Type[th.optim.Optimizer] = th.optim.Adam,
optimizer_kwargs: Optional[Dict[str, Any]] = None,
):
super(CnnPolicy, self).__init__(
observation_space,
action_space,
lr_schedule,
net_arch,
activation_fn,
features_extractor_class,
features_extractor_kwargs,
normalize_images,
optimizer_class,
optimizer_kwargs,
)
register_policy("DQN", "MlpPolicy", MlpPolicy)
register_policy("DQN", "CnnPolicy", CnnPolicy)
# This file is here just to define MlpPolicy/CnnPolicy
# that work for PPO
from hmlf.common.policies import ActorCriticCnnPolicy, ActorCriticPolicy
from hmlf.common.policy_register import register_policy
MlpPolicy = ActorCriticPolicy
CnnPolicy = ActorCriticCnnPolicy
register_policy("PPO", "MlpPolicy", MlpPolicy)
register_policy("PPO", "CnnPolicy", CnnPolicy)
self,
observation_space: Space,
action_space: Space,
lr_schedule: Schedule,
net_arch: Optional[Union[List[int], Dict[str, List[int]]]] = None,
activation_fn: Type[nn.Module] = nn.ReLU,
features_extractor_class: Type[BaseFeaturesExtractor] = NatureCNN,
features_extractor_kwargs: Optional[Dict[str, Any]] = None,
normalize_images: bool = True,
optimizer_class: Type[th.optim.Optimizer] = th.optim.Adam,
optimizer_kwargs: Optional[Dict[str, Any]] = None,
share_features_extractor: bool = True,
):
super().__init__(
observation_space,
action_space,
lr_schedule,
net_arch,
activation_fn,
features_extractor_class,
features_extractor_kwargs,
normalize_images,
optimizer_class,
optimizer_kwargs,
share_features_extractor,
)
register_policy("SDDPG", "MlpPolicy", MlpPolicy)
register_policy("SDDPG", "CnnPolicy", CnnPolicy)
action_space: spaces.Space,
lr_schedule: Schedule,
net_arch: Optional[Union[List[int], Dict[str, List[int]]]] = None,
activation_fn: Type[nn.Module] = nn.ReLU,
features_extractor_class: Type[BaseFeaturesExtractor] = NatureCNN,
features_extractor_kwargs: Optional[Dict[str, Any]] = None,
normalize_images: bool = True,
optimizer_class: Type[th.optim.Optimizer] = th.optim.Adam,
optimizer_kwargs: Optional[Dict[str, Any]] = None,
n_critics: int = 2,
share_features_extractor: bool = True,
):
super(CnnPolicy, self).__init__(
observation_space,
action_space,
lr_schedule,
net_arch,
activation_fn,
features_extractor_class,
features_extractor_kwargs,
normalize_images,
optimizer_class,
optimizer_kwargs,
n_critics,
share_features_extractor,
)
register_policy("TD3", "MlpPolicy", MlpPolicy)
register_policy("TD3", "CnnPolicy", CnnPolicy)