{
  "feature_extractor": {
    "init_args": {
      "sample_rate": 24000,
      "n_fft": 1024,
      "hop_length": 256,
      "n_mels": 100
    }
  },
  "backbone": {
    "init_args": {
      "input_channels": 100,
      "dim": 512,
      "intermediate_dim": 1536,
      "num_layers": 8
    }
  },
  "head": {
    "init_args": {
      "dim": 512,
      "n_fft": 1024,
      "hop_length": 256
    }
  }
}