Skip to content

allenact_plugins.robothor_plugin.robothor_models#

[view_source]

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.

ResnetTensorObjectNavActorCritic#

class ResnetTensorObjectNavActorCritic(ActorCriticModel[CategoricalDistr])

[view_source]

ResnetTensorObjectNavActorCritic.recurrent_hidden_state_size#

 | @property
 | recurrent_hidden_state_size() -> Union[int, Dict[str, Tuple[Sequence[Tuple[str, Optional[int]]], torch.dtype]]]

[view_source]

The recurrent hidden state size of the model.

ResnetTensorObjectNavActorCritic.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).

ResnetTensorObjectNavActorCritic.num_recurrent_layers#

 | @property
 | num_recurrent_layers() -> int

[view_source]

Number of recurrent hidden layers.

ResnetTensorObjectNavActorCritic.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.

ResnetFasterRCNNTensorsGoalEncoder#

class ResnetFasterRCNNTensorsGoalEncoder(nn.Module)

[view_source]

ResnetFasterRCNNTensorsGoalEncoder.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.

ResnetFasterRCNNTensorsObjectNavActorCritic#

class ResnetFasterRCNNTensorsObjectNavActorCritic(ActorCriticModel[CategoricalDistr])

[view_source]

ResnetFasterRCNNTensorsObjectNavActorCritic.recurrent_hidden_state_size#

 | @property
 | recurrent_hidden_state_size() -> int

[view_source]

The recurrent hidden state size of the model.

ResnetFasterRCNNTensorsObjectNavActorCritic.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).

ResnetFasterRCNNTensorsObjectNavActorCritic.num_recurrent_layers#

 | @property
 | num_recurrent_layers() -> int

[view_source]

Number of recurrent hidden layers.

ResnetFasterRCNNTensorsObjectNavActorCritic.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.