configs/config.json

{
  "train": {
    "log_interval": 200,
    "eval_interval": 1000,
    "seed": 42,
    "epochs": 10000,
    "learning_rate": 0.0002,
    "betas": [
      0.8,
      0.99
    ],
    "eps": 1e-09,
    "batch_size": 24,
    "fp16_run": false,
    "lr_decay": 0.99995,
    "segment_size": 16384,
    "init_lr_ratio": 1,
    "warmup_epochs": 0,
    "c_mel": 45,
    "c_kl": 1.0,
    "skip_optimizer": true
  },
  "data": {
    "training_files": "filelists/train.list",
    "validation_files": "filelists/val.list",
    "max_wav_value": 32768.0,
    "sampling_rate": 22050,
    "filter_length": 2048,
    "hop_length": 512,
    "win_length": 2048,
    "n_mel_channels": 128,
    "mel_fmin": 0.0,
    "mel_fmax": null,
    "add_blank": true,
    "n_speakers": 700,
    "cleaned_text": true,
    "spk2id": {
      "老克勒": 0,
      "美琳": 1,
      "Jane": 2,
      "小贝": 3,
      "SSB0005": 4,
      "SSB0009": 5,
      "SSB0011": 6,
      "SSB0012": 7,
      "SSB0016": 8,
      "SSB0018": 9,
      "SSB0033": 10,
      "SSB0038": 11,
      "SSB0043": 12,
      "SSB0057": 13,
      "SSB0073": 14,
      "SSB0080": 15,
      "SSB0112": 16,
      "SSB0122": 17,
      "SSB0133": 18,
      "SSB0139": 19,
      "SSB0145": 20,
      "SSB0149": 21,
      "SSB0193": 22,
      "SSB0197": 23,
      "SSB0200": 24,
      "SSB0241": 25,
      "SSB0246": 26,
      "SSB0261": 27,
      "SSB0267": 28,
      "SSB0273": 29,
      "SSB0287": 30,
      "SSB0288": 31,
      "SSB0299": 32,
      "SSB0307": 33,
      "SSB0309": 34,
      "SSB0315": 35,
      "SSB0316": 36,
      "SSB0323": 37,
      "SSB0338": 38,
      "SSB0339": 39,
      "SSB0341": 40,
      "SSB0342": 41,
      "SSB0354": 42,
      "SSB0366": 43,
      "SSB0375": 44,
      "SSB0379": 45,
      "SSB0380": 46,
      "SSB0382": 47,
      "SSB0385": 48,
      "SSB0393": 49,
      "SSB0394": 50,
      "SSB0395": 51,
      "SSB0407": 52,
      "SSB0415": 53,
      "SSB0426": 54,
      "SSB0427": 55,
      "SSB0434": 56,
      "SSB0435": 57,
      "SSB0470": 58,
      "SSB0482": 59,
      "SSB0502": 60,
      "SSB0534": 61,
      "SSB0535": 62,
      "SSB0539": 63,
      "SSB0544": 64,
      "SSB0565": 65,
      "SSB0570": 66,
      "SSB0578": 67,
      "SSB0588": 68,
      "SSB0590": 69,
      "SSB0594": 70,
      "SSB0599": 71,
      "SSB0601": 72,
      "SSB0603": 73,
      "SSB0606": 74,
      "SSB0607": 75,
      "SSB0609": 76,
      "SSB0614": 77,
      "SSB0623": 78,
      "SSB0629": 79,
      "SSB0631": 80,
      "SSB0632": 81,
      "SSB0666": 82,
      "SSB0668": 83,
      "SSB0671": 84,
      "SSB0686": 85,
      "SSB0700": 86,
      "SSB0710": 87,
      "SSB0720": 88,
      "SSB0723": 89,
      "SSB0737": 90,
      "SSB0746": 91,
      "SSB0748": 92,
      "SSB0751": 93,
      "SSB0758": 94,
      "SSB0760": 95,
      "SSB0762": 96,
      "SSB0778": 97,
      "SSB0780": 98,
      "SSB0784": 99,
      "SSB0786": 100,
      "SSB0794": 101,
      "SSB0817": 102,
      "SSB0851": 103,
      "SSB0863": 104,
      "SSB0871": 105,
      "SSB0887": 106,
      "SSB0913": 107,
      "SSB0915": 108,
      "SSB0919": 109,
      "SSB0935": 110,
      "SSB0966": 111,
      "SSB0987": 112,
      "SSB1008": 113,
      "SSB1020": 114,
      "SSB1024": 115,
      "SSB1050": 116,
      "SSB1055": 117,
      "SSB1056": 118,
      "SSB1064": 119,
      "SSB1072": 120,
      "SSB1091": 121,
      "SSB1096": 122,
      "SSB1100": 123,
      "SSB1108": 124,
      "SSB1115": 125,
      "SSB1125": 126,
      "SSB1131": 127,
      "SSB1136": 128,
      "SSB1138": 129,
      "SSB1161": 130,
      "SSB1203": 131,
      "SSB1204": 132,
      "SSB1218": 133,
      "SSB1221": 134,
      "SSB1253": 135,
      "SSB1320": 136,
      "SSB1341": 137,
      "SSB1366": 138,
      "SSB1377": 139,
      "SSB1383": 140,
      "SSB1385": 141,
      "SSB1392": 142,
      "SSB1393": 143,
      "SSB1408": 144,
      "SSB1431": 145,
      "SSB1437": 146,
      "SSB1448": 147,
      "SSB1555": 148,
      "SSB1563": 149,
      "SSB1567": 150,
      "SSB1575": 151,
      "SSB1585": 152,
      "SSB1593": 153,
      "SSB1607": 154,
      "SSB1624": 155,
      "SSB1625": 156,
      "SSB1630": 157,
      "SSB1650": 158,
      "SSB1670": 159,
      "SSB1684": 160,
      "SSB1686": 161,
      "SSB1699": 162,
      "SSB1711": 163,
      "SSB1759": 164,
      "SSB1806": 165,
      "SSB1828": 166,
      "SSB1831": 167,
      "SSB1832": 168,
      "SSB1837": 169,
      "SSB1846": 170,
      "SSB1863": 171,
      "SSB1878": 172,
      "SSB1891": 173,
      "SSB1918": 174,
      "SSB1935": 175,
      "SSB1939": 176,
      "SSB1956": 177,
      "SSB0693": 178,
      "SSB0711": 179,
      "SSB0716": 180,
      "SSB0717": 181,
      "SSB0736": 182,
      "SSB0749": 183,
      "SSB0809": 184,
      "SSB0702": 185,
      "SSB0822": 186,
      "SSB1110": 187,
      "SSB1215": 188,
      "SSB1399": 189,
      "SSB1728": 190,
      "SSB0993": 191,
      "SSB0997": 192,
      "SSB1000": 193,
      "SSB1001": 194,
      "SSB1002": 195,
      "SSB1126": 196,
      "SSB1135": 197,
      "SSB1176": 198,
      "SSB1187": 199,
      "SSB1197": 200,
      "SSB1216": 201,
      "SSB1219": 202,
      "SSB1239": 203,
      "SSB1274": 204,
      "SSB1302": 205,
      "SSB1322": 206,
      "SSB1328": 207,
      "SSB1340": 208,
      "SSB1365": 209,
      "SSB1382": 210,
      "SSB1402": 211,
      "SSB1452": 212,
      "SSB1457": 213,
      "SSB1739": 214,
      "SSB1745": 215,
      "SSB1781": 216,
      "SSB1782": 217,
      "SSB1809": 218,
      "SSB1810": 219,
      "SSB1872": 220,
      "SSB1902": 221,
      "小庄": 222,
      "小嘟": 223
    }
  },
  "model": {
    "use_spk_conditioned_encoder": true,
    "use_noise_scaled_mas": true,
    "use_mel_posterior_encoder": false,
    "use_duration_discriminator": true,
    "inter_channels": 192,
    "hidden_channels": 192,
    "filter_channels": 768,
    "n_heads": 2,
    "n_layers": 6,
    "kernel_size": 3,
    "p_dropout": 0.1,
    "resblock": "1",
    "resblock_kernel_sizes": [
      3,
      7,
      11
    ],
    "resblock_dilation_sizes": [
      [
        1,
        3,
        5
      ],
      [
        1,
        3,
        5
      ],
      [
        1,
        3,
        5
      ]
    ],
    "upsample_rates": [
      8,
      8,
      2,
      2,
      2
    ],
    "upsample_initial_channel": 512,
    "upsample_kernel_sizes": [
      16,
      16,
      8,
      2,
      2
    ],
    "n_layers_q": 3,
    "use_spectral_norm": false,
    "gin_channels": 256
  },
  "version": "2.0"
}