Skip to content

projects.pointnav_baselines.models.point_nav_models#

[view_source]

ResnetTensorPointNavActorCritic#

class ResnetTensorPointNavActorCritic(ActorCriticModel[CategoricalDistr])

[view_source]

ResnetTensorPointNavActorCritic.recurrent_hidden_state_size#

 | @property
 | recurrent_hidden_state_size() -> int

[view_source]

The recurrent hidden state size of the model.

ResnetTensorPointNavActorCritic.is_blind#

 | @property
 | is_blind() -> bool

[view_source]

True if the model is blind (e.g. neither 'depth' or 'rgb' is an input observation type).

ResnetTensorPointNavActorCritic.num_recurrent_layers#

 | @property
 | num_recurrent_layers() -> int

[view_source]

Number of recurrent hidden layers.

ResnetTensorGoalEncoder#

class ResnetTensorGoalEncoder(nn.Module)

[view_source]

ResnetTensorGoalEncoder.get_object_type_encoding#

 | get_object_type_encoding(observations: Dict[str, torch.FloatTensor]) -> torch.FloatTensor

[view_source]

Get the object type encoding from input batched observations.

ResnetDualTensorGoalEncoder#

class ResnetDualTensorGoalEncoder(nn.Module)

[view_source]

ResnetDualTensorGoalEncoder.get_object_type_encoding#

 | get_object_type_encoding(observations: Dict[str, torch.FloatTensor]) -> torch.FloatTensor

[view_source]

Get the object type encoding from input batched observations.