def __init__(self, max_len, latent_dim): super(Regressor, self).__init__() self.latent_dim = latent_dim self.max_len = max_len self.conv1 = nn.Conv1d(DECISION_DIM, 9, 9) self.conv2 = nn.Conv1d(9, 9, 9) self.conv3 = nn.Conv1d(9, 10, 11) self.last_conv_size = max_len - 9 + 1 - 9 + 1 - 11 + 1 self.w1 = nn.Linear(self.last_conv_size * 10, 435) self.mean_w = nn.Linear(435, latent_dim) self.label_linear_1 = nn.Linear(latent_dim, cmd_args.output_dim) weights_init(self)
def __init__(self, max_len, latent_dim): super(StateDecoder, self).__init__() self.latent_dim = latent_dim self.max_len = max_len self.z_to_latent = nn.Linear(self.latent_dim, self.latent_dim) if cmd_args.rnn_type == 'gru': self.gru = nn.GRU(self.latent_dim, cmd_args.hidden, 3) else: raise NotImplementedError self.decoded_logits = nn.Linear(cmd_args.hidden, DECISION_DIM) weights_init(self)
def __init__(self, max_len, latent_dim): super(CNNEncoder, self).__init__() self.latent_dim = latent_dim self.max_len = max_len self.conv1 = nn.Conv1d(DECISION_DIM, 9, 9) self.conv2 = nn.Conv1d(9, 9, 9) self.conv3 = nn.Conv1d(9, 10, 11) self.last_conv_size = max_len - 9 + 1 - 9 + 1 - 11 + 1 self.w1 = nn.Linear(self.last_conv_size * 10, 435) self.mean_w = nn.Linear(435, latent_dim) self.log_var_w = nn.Linear(435, latent_dim) weights_init(self)
def __init__(self, max_len, latent_dim): super(CNNEncoder, self).__init__() self.latent_dim = latent_dim self.max_len = max_len self.conv1 = nn.Conv1d(DECISION_DIM, cmd_args.c1, cmd_args.c1) self.conv2 = nn.Conv1d(cmd_args.c1, cmd_args.c2, cmd_args.c2) self.conv3 = nn.Conv1d(cmd_args.c2, cmd_args.c3, cmd_args.c3) self.last_conv_size = max_len - cmd_args.c1 + 1 - cmd_args.c2 + 1 - cmd_args.c3 + 1 self.w1 = nn.Linear(self.last_conv_size * cmd_args.c3, cmd_args.dense) self.mean_w = nn.Linear(cmd_args.dense, latent_dim) self.log_var_w = nn.Linear(cmd_args.dense, latent_dim) weights_init(self)