{
  "architectures": [
    "WavCoch"
  ],
  "auto_map": {
    "AutoConfig": "configuration_wavcoch.WavCochConfig",
    "AutoModel": "modeling_wavcoch.WavCoch"
  },
  "causal_convs": true,
  "causal_pad_mode": "repeat",
  "channels": [
    8,
    8,
    8,
    5,
    5,
    5
  ],
  "decoder_block_type": "plain",
  "decoder_dim": 512,
  "decoder_kernel_size": 9,
  "decoder_layers": 8,
  "decoder_residual_bottleneck_factor": 4,
  "decoder_residual_dilations": [
    1,
    3,
    9
  ],
  "decoder_residual_kernel_size": 7,
  "dtype": "float32",
  "encoder_block_type": "plain",
  "encoder_dim": 512,
  "encoder_kernel_size": 3,
  "encoder_layers": 8,
  "encoder_residual_bottleneck_factor": 4,
  "encoder_residual_dilations": [
    1,
    3,
    9
  ],
  "encoder_residual_kernel_size": 7,
  "has_vocoder": false,
  "hop_length": 320,
  "model_type": "wavcoch",
  "out_channels": 211,
  "quantizer": "FSQ",
  "sample_rate": 16000,
  "transformers_version": "5.10.2",
  "vocab_size": 64000,
  "vocoder_resblock": "1",
  "vocoder_resblock_dilation_sizes": [
    [
      1,
      3,
      5
    ],
    [
      1,
      3,
      5
    ],
    [
      1,
      3,
      5
    ]
  ],
  "vocoder_resblock_kernel_sizes": [
    11,
    7,
    3
  ],
  "vocoder_upsample_initial_channel": 512,
  "vocoder_upsample_kernel_sizes": [
    10,
    8,
    4,
    4
  ],
  "vocoder_upsample_rates": [
    5,
    4,
    2,
    2
  ],
  "window_padding": 1000,
  "window_size": 1001
}