Method __init__

modules/fastspeech/tts_modules.py:193–220

Initilize pitch predictor module. Args: idim (int): Input dimension. n_layers (int, optional): Number of convolutional layers. n_chans (int, optional): Number of channels of convolutional layers. kernel_size (int, optional): Kernel size of conv

(self, idim, n_layers=5, n_chans=384, odim=2, kernel_size=5,
                 dropout_rate=0.1, padding='SAME')

class PitchPredictor(torch.nn.Module):
    def __init__(self, idim, n_layers=5, n_chans=384, odim=2, kernel_size=5,
                 dropout_rate=0.1, padding='SAME'):
        """Initilize pitch predictor module.
        Args:
            idim (int): Input dimension.
            n_layers (int, optional): Number of convolutional layers.
            n_chans (int, optional): Number of channels of convolutional layers.
            kernel_size (int, optional): Kernel size of convolutional layers.
            dropout_rate (float, optional): Dropout rate.
        """
        super(PitchPredictor, self).__init__()
        self.conv = torch.nn.ModuleList()
        self.kernel_size = kernel_size
        self.padding = padding
        for idx in range(n_layers):
            in_chans = idim if idx == 0 else n_chans
            self.conv += [torch.nn.Sequential(
                torch.nn.ConstantPad1d(((kernel_size - 1) // 2, (kernel_size - 1) // 2)
                                       if padding == 'SAME'
                                       else (kernel_size - 1, 0), 0),
                torch.nn.Conv1d(in_chans, n_chans, kernel_size, stride=1, padding=0),
                torch.nn.ReLU(),
                LayerNorm(n_chans, dim=1),
                torch.nn.Dropout(dropout_rate)
            )]
        self.linear = torch.nn.Linear(n_chans, odim)
        self.embed_positions = SinusoidalPositionalEmbedding(idim, 0, init_size=4096)
        self.pos_embed_alpha = nn.Parameter(torch.Tensor([1]))

    def forward(self, xs):
        """