{
  "model_type": "cosyvoice3",
  "sample_rate": 24000,
  "dtype": "float16",
  "llm": {
    "llm_input_size": 896,
    "llm_output_size": 896,
    "speech_token_size": 6561,
    "extended_vocab_size": 200
  },
  "flow": {
    "input_size": 80,
    "output_size": 80,
    "spk_embed_dim": 192,
    "vocab_size": 6561,
    "token_mel_ratio": 2,
    "pre_lookahead_len": 3,
    "dit": {
      "dim": 1024,
      "depth": 22,
      "heads": 16,
      "dim_head": 64,
      "ff_mult": 2,
      "static_chunk_size": 50
    }
  },
  "hifigan": {
    "in_channels": 80,
    "base_channels": 512,
    "nb_harmonics": 8,
    "sampling_rate": 24000,
    "upsample_rates": [
      8,
      5,
      3
    ],
    "upsample_kernel_sizes": [
      16,
      11,
      7
    ],
    "istft_n_fft": 16,
    "istft_hop_len": 4,
    "resblock_kernel_sizes": [
      3,
      7,
      11
    ],
    "resblock_dilation_sizes": [
      [
        1,
        3,
        5
      ],
      [
        1,
        3,
        5
      ],
      [
        1,
        3,
        5
      ]
    ],
    "source_resblock_kernel_sizes": [
      7,
      7,
      11
    ],
    "source_resblock_dilation_sizes": [
      [
        1,
        3,
        5
      ],
      [
        1,
        3,
        5
      ],
      [
        1,
        3,
        5
      ]
    ],
    "conv_pre_look_right": 4,
    "causal": true
  }
}