config.yaml

feature_extractor:
  class_path: vocos.feature_extractors.MelSpectrogramFeatures
  init_args:
    sample_rate: 24000
    n_fft: 1024
    hop_length: 256
    n_mels: 100
    padding: center

backbone:
  class_path: vocos.models.VocosBackbone
  init_args:
    input_channels: 100
    dim: 768
    intermediate_dim: 2304
    num_layers: 12

head:
  class_path: vocos.heads.ISTFTHead
  init_args:
    dim: 768
    n_fft: 1024
    hop_length: 256
    padding: center