LPAC neural network architecture. More...

Inheritance diagram for LPAC:

Public Member Functions
	__init__ (self, in_config)

torch.Tensor	forward (self, torch_geometric.data.Data data)
	Forward pass of the LPAC model.

None	load_compiled_state_dict (self, str model_state_dict_path)

None	load_model (self, str model_state_dict_path)
	Load the model from the state dict.

None	load_model_state_dict (self, dict model_state_dict)
	Load the model from the state dict.

None	load_cnn_backbone (self, str model_path)
	Load the CNN backbone from the model path.

None	load_gnn_backbone (self, str model_path)
	Load the GNN backbone from the model path.

Public Member Functions inherited from GNNConfigParser
None	parse (self, dict config)
	Parse the configuration for the GNN model.

Public Attributes
	cnn_config = in_config["CNNBackBone"]

	cnn_backbone = CNNBackBone(self.cnn_config)

	gnn_backbone = GNNBackBone(self.config, self.cnn_backbone.latent_size + 2)

	gnn_mlp = MLP([self.latent_size, 32, 32])

	output_linear = torch.nn.Linear(32, self.output_dim)

	output_dim

Public Attributes inherited from GNNConfigParser
	config = None

	input_dim = None

	output_dim = None

	num_hops = None

	num_layers = None

	latent_size = None

Detailed Description

LPAC neural network architecture.

Definition at line 40 of file lpac.py.

Constructor & Destructor Documentation

◆ init()

__init__	(		self,
			in_config )

Reimplemented from GNNConfigParser.

Definition at line 41 of file lpac.py.

Member Function Documentation

◆ forward()

torch.Tensor forward	(		self,
		torch_geometric.data.Data	data )

Forward pass of the LPAC model.

Definition at line 59 of file lpac.py.

◆ load_cnn_backbone()

None load_cnn_backbone	(		self,
		str	model_path )

Load the CNN backbone from the model path.

Definition at line 106 of file lpac.py.

◆ load_compiled_state_dict()

None load_compiled_state_dict	(		self,
		str	model_state_dict_path )

Definition at line 83 of file lpac.py.

◆ load_gnn_backbone()

None load_gnn_backbone	(		self,
		str	model_path )

Load the GNN backbone from the model path.

Definition at line 112 of file lpac.py.

◆ load_model()

None load_model	(		self,
		str	model_state_dict_path )

Load the model from the state dict.

Definition at line 94 of file lpac.py.

◆ load_model_state_dict()

None load_model_state_dict	(		self,
		dict	model_state_dict )

Load the model from the state dict.

Definition at line 100 of file lpac.py.

Member Data Documentation

◆ cnn_backbone

cnn_backbone = CNNBackBone(self.cnn_config)

Definition at line 45 of file lpac.py.

◆ cnn_config

cnn_config = in_config["CNNBackBone"]

Definition at line 43 of file lpac.py.

◆ gnn_backbone

gnn_backbone = GNNBackBone(self.config, self.cnn_backbone.latent_size + 2)

Definition at line 46 of file lpac.py.

◆ gnn_mlp

gnn_mlp = MLP([self.latent_size, 32, 32])

Definition at line 50 of file lpac.py.

◆ output_dim

output_dim

Definition at line 53 of file lpac.py.

◆ output_linear

output_linear = torch.nn.Linear(32, self.output_dim)

Definition at line 51 of file lpac.py.

Public Member Functions

Public Attributes