config.json

{
  "model_type": "zonos2",
  "dtype": "bfloat16",
  "n_layers": 28,
  "dim": 2048,
  "head_dim": 128,
  "n_heads": null,
  "n_kv_heads": 4,
  "ffn_dim_multiplier": 1.5,
  "multiple_of": 256,
  "norm_eps": 1e-05,
  "rope_theta": 10000.0,
  "max_seqlen": 6144,
  "n_codebooks": 9,
  "codebook_size": 1024,
  "eoa_id": 1024,
  "audio_pad_id": 1025,
  "text_vocab": 519,
  "loss_softcap": 15.0,
  "sample_rate": 44100,
  "dac_model_id": "mlx-community/descript-audio-codec-44khz",
  "speaker_enabled": true,
  "speaker_embedding_dim": 2048,
  "speaker_lda_dim": 1024,
  "speaker_encoder_model_id": "marksverdhei/Qwen3-Voice-Embedding-12Hz-1.7B",
  "speaker_encoder_path": "speaker_encoder",
  "speaker_encoder_sample_rate": 24000,
  "speaker_background_token_enabled": true,
  "accurate_mode_token_enabled": true,
  "speaking_rate_num_buckets": 8,
  "speaking_rate_buckets": [
    "0-8",
    "8-11",
    "11-14",
    "14-17",
    "17-21",
    "21-28",
    "28-40",
    "40+"
  ],
  "quality_num_buckets": 60,
  "quality_features": [
    "lufs",
    "estimated_snr",
    "max_pause",
    "estimated_bandlimit_hz",
    "leading_silence_s",
    "trailing_silence_s"
  ],
  "quality_buckets": {
    "lufs": [
      "-1000--50",
      "-50--45.5",
      "-45.5--41",
      "-41--36.5",
      "-36.5--32",
      "-32--27.5",
      "-27.5--23",
      "-23--18.5",
      "-18.5--14",
      "-14--9.5",
      "-9.5--5",
      "-5+"
    ],
    "estimated_snr": [
      "-1000-0",
      "0-6",
      "6-12",
      "12-18",
      "18-24",
      "24-30",
      "30-36",
      "36-42",
      "42-48",
      "48-54",
      "54-60",
      "60+"
    ],
    "max_pause": [
      "0-0.5",
      "0.5-1",
      "1-1.5",
      "1.5-2",
      "2-2.5",
      "2.5-3",
      "3-3.5",
      "3.5-4",
      "4-4.5",
      "4.5-5",
      "5-5.5",
      "5.5-6"
    ],
    "estimated_bandlimit_hz": [
      "495.3-3433",
      "3433-6371",
      "6371-9310",
      "9310-12248",
      "12248-15186",
      "15186-18124",
      "18124-21062",
      "21062-24000"
    ],
    "leading_silence_s": [
      "0-0.05",
      "0.05-0.1",
      "0.1-0.25",
      "0.25-0.5",
      "0.5-1",
      "1-2",
      "2-4",
      "4+"
    ],
    "trailing_silence_s": [
      "0-0.05",
      "0.05-0.1",
      "0.1-0.25",
      "0.25-0.5",
      "0.5-1",
      "1-2",
      "2-4",
      "4+"
    ]
  },
  "quality_dropout": {
    "lufs": 0.25,
    "estimated_snr": 0.25,
    "max_pause": 0.25,
    "estimated_bandlimit_hz": 0.25,
    "leading_silence_s": 0.25,
    "trailing_silence_s": 0.25
  },
  "moe_impl": "sonic",
  "moe_n_experts": 16,
  "moe_router_topk": 1,
  "special_topk_layers": {
    "26": 2
  },
  "moe_router_dim": 128,
  "moe_start_from_layer": 3,
  "moe_end_from_layer": 1,
  "norm_topk_prob": false,
  "moe_balancing_strategy": "legacy"
}