def create_for_env(
cls,
env: EnvWrapper,
policy: Optional[Policy],
*,
device: Union[str, torch.device] = "cpu",
obs_preprocessor=None,
action_extractor=None,
**kwargs,
):
"""
If `policy` is not given, we will try to create a random policy
"""
if isinstance(device, str):
device = torch.device(device)
if obs_preprocessor is None:
obs_preprocessor = env.get_obs_preprocessor(device=device)
if action_extractor is None:
action_extractor = env.get_action_extractor()
if policy is None:
policy = make_random_policy_for_env(env)
return cls(
policy,
obs_preprocessor=obs_preprocessor,
action_extractor=action_extractor,
**kwargs,
)
def create_for_env(
cls,
env: EnvWrapper,
policy: Policy,
*,
device: Union[str, torch.device] = "cpu",
obs_preprocessor=None,
action_extractor=None,
**kwargs,
):
if isinstance(device, str):
device = torch.device(device)
if obs_preprocessor is None:
obs_preprocessor = env.get_obs_preprocessor(device=device)
if action_extractor is None:
action_extractor = env.get_action_extractor()
return cls(
policy,
obs_preprocessor=obs_preprocessor,
action_extractor=action_extractor,
**kwargs,
)
