def __init__()

in Experiments/PolicyNetworks.py [0:0]

• 22 lines of code
• 8 McCabe index (conditional complexity)

	def __init__(self, input_size, hidden_size, z_dimensions, args, number_layers=4):
		# Ensures inheriting from torch.nn.Module goes nicely and cleanly. 	
		# super().__init__()
		super(ContinuousVariationalPolicyNetwork, self).__init__()
	
		self.args = args	
		self.input_size = input_size
		self.hidden_size = hidden_size
		self.output_size = z_dimensions
		self.num_layers = number_layers	
		self.z_exploration_bias = self.args.z_exploration_bias
		self.b_exploration_bias = self.args.b_exploration_bias
		self.z_probability_factor = self.args.z_probability_factor
		self.b_probability_factor = self.args.b_probability_factor
		self.batch_size = self.args.batch_size

		# Define a bidirectional LSTM now.
		self.lstm = torch.nn.LSTM(input_size=self.input_size,hidden_size=self.hidden_size,num_layers=self.num_layers, bidirectional=True)

		# Transform to output space - Latent z and Latent b. 
		# THIS OUTPUT LAYER TAKES 2*HIDDEN SIZE as input because it's bidirectional. 
		self.termination_output_layer = torch.nn.Linear(2*self.hidden_size,2)

		# Softmax activation functions for Bernoulli termination probability and latent z selection .
		self.batch_softmax_layer = torch.nn.Softmax(dim=-1)
		self.batch_logsoftmax_layer = torch.nn.LogSoftmax(dim=-1)

		# Define output layers for the LSTM, and activations for this output layer. 
		self.mean_output_layer = torch.nn.Linear(2*self.hidden_size,self.output_size)
		self.variances_output_layer = torch.nn.Linear(2*self.hidden_size, self.output_size)

		self.activation_layer = torch.nn.Tanh()

		self.variance_activation_layer = torch.nn.Softplus()
		self.variance_activation_bias = 0.
			
		self.variance_factor = 0.01