decoders/config.yaml

backbone:
  class_path: vocos.models.VocosBackbone
  init_args:
    input_channels: 1024
    dim: 512
    intermediate_dim: 1536
    num_layers: 8

head:
  class_path: vocos.heads.ISTFTHead
  init_args:
    dim: 512
    n_fft: 3528
    hop_length: 882
    padding: same