espnet.nets.pytorch_backend.transformer.multi_layer_conv.MultiLayeredConv1d

source

class espnet.nets.pytorch_backend.transformer.multi_layer_conv.MultiLayeredConv1d(in_chans, hidden_chans, kernel_size, dropout_rate)

Bases: Module

Multi-layered conv1d for Transformer block.

This is a module of multi-leyered conv1d designed to replace positionwise feed-forward network in Transforner block, which is introduced in FastSpeech: Fast, Robust and Controllable Text to Speech.

Initialize MultiLayeredConv1d module.

• Parameters:
  • in_chans (int) – Number of input channels.
  • hidden_chans (int) – Number of hidden channels.
  • kernel_size (int) – Kernel size of conv1d.
  • dropout_rate (float) – Dropout rate.

forward(x)

Calculate forward propagation.

• Parameters:x (torch.Tensor) – Batch of input tensors (B, T, in_chans).
• Returns: Batch of output tensors (B, T, hidden_chans).
• Return type: torch.Tensor