mixing_style_transfer/networks/configs.yaml

# model architecture configurations


# Music Effects Encoder
Effects_Encoder:

  default:
    channels: [16, 32, 64, 128, 256, 256, 512, 512, 1024, 1024, 2048, 2048]
    kernels: [25, 25, 15, 15, 10, 10, 10, 10, 5, 5, 5, 5]
    strides: [4, 4, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1]
    dilation: [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]
    bias: True
    norm: 'batch'
    conv_block: 'res'
    activation: "relu"
    

# TCN
TCN:

  # receptive field = 5.2 seconds
  default:
    condition_dimension: 2048
    nblocks: 14
    dilation_growth: 2
    kernel_size: 15
    channel_width: 128
    stack_size: 15
    causal: False