{
  "version": "1.0",
  "truncation": null,
  "padding": null,
  "added_tokens": [
    {
      "id": 0,
      "content": "UNK",
      "single_word": false,
      "lstrip": false,
      "rstrip": false,
      "normalized": false,
      "special": true
    },
    {
      "id": 1,
      "content": "PAD",
      "single_word": false,
      "lstrip": false,
      "rstrip": false,
      "normalized": false,
      "special": true
    },
    {
      "id": 154,
      "content": "<|endoftext|>",
      "single_word": false,
      "lstrip": false,
      "rstrip": false,
      "normalized": false,
      "special": true
    }
  ],
  "normalizer": {
    "type": "Sequence",
    "normalizers": [
      {
        "type": "Replace",
        "pattern": {
          "String": "\n"
        },
        "content": " UTT_BOUNDARY"
      },
      {
        "type": "Strip",
        "strip_left": true,
        "strip_right": true
      }
    ]
  },
  "pre_tokenizer": {
    "type": "Whitespace"
  },
  "post_processor": null,
  "decoder": null,
  "model": {
    "type": "WordLevel",
    "vocab": {
      "UNK": 0,
      "PAD": 1,
      "BOS": 2,
      "EOS": 3,
      "WORD_BOUNDARY": 4,
      "UTT_BOUNDARY": 5,
      "aː˧": 6,
      "t": 7,
      "ɐ˥": 8,
      "k": 9,
      "l": 10,
      "j": 11,
      "ʊ˥": 12,
      "aː˧˩̰": 13,
      "ɛː˥": 14,
      "n": 15,
      "ei˩˧": 16,
      "w": 17,
      "aː˨": 18,
      "ɐi˧˥": 19,
      "m̩˧˥": 20,
      "m": 21,
      "ou˥": 22,
      "aː˧˥": 23,
      "ei˥": 24,
      "iː˧": 25,
      "ts": 26,
      "ɔː˧˥": 27,
      "tʰ": 28,
      "iː˥": 29,
      "f": 30,
      "aːĭ˧": 31,
      "ɐ˨": 32,
      "p": 33,
      "h": 34,
      "ɵy˧": 35,
      "aː˥": 36,
      "ou˨": 37,
      "ɔː˧": 38,
      "ɐi˧˩̰": 39,
      "uː˧": 40,
      "ŋ": 41,
      "s": 42,
      "ɔːĭ˥": 43,
      "ɐu˨": 44,
      "iː˨": 45,
      "ei˧˥": 46,
      "ɐi˨": 47,
      "ʊ˧˩̰": 48,
      "ʊ˨": 49,
      "aː˩˧": 50,
      "aːĭ˧˥": 51,
      "ɔː˨": 52,
      "ɛː˩˧": 53,
      "ɪ˨": 54,
      "iːŭ˧": 55,
      "ɛː˧˩̰": 56,
      "ɪ˧˥": 57,
      "̩˧˩̰": 58,
      "ɵ˧˥": 59,
      "ei˧": 60,
      "ɐu˧˩̰": 61,
      "m̩˧": 62,
      "ɐu˧˥": 63,
      "ɐu˩˧": 64,
      "ɐi˥": 65,
      "ɔː˥": 66,
      "ɔːĭ˧": 67,
      "ou˧˥": 68,
      "ou˩˧": 69,
      "ɐ˧": 70,
      "tsʰ": 71,
      "ɛː˧˥": 72,
      "iː˧˥": 73,
      "ɔː˩˧": 74,
      "kʰ": 75,
      "ɐ˧˩̰": 76,
      "aːŭ˧˥": 77,
      "pʰ": 78,
      "aːĭ˧˩̰": 79,
      "ɵy˩˧": 80,
      "ɵ˧": 81,
      "ɛː˧": 82,
      "ei˧˩̰": 83,
      "uː˧˥": 84,
      "ɔː˧˩̰": 85,
      "ɛː˨": 86,
      "uː˥": 87,
      "ʊ˧": 88,
      "iː˧˩̰": 89,
      "yː˨": 90,
      "aːŭ˧": 91,
      "œː˩˧": 92,
      "ɐ˧˥": 93,
      "iː˩˧": 94,
      "ɪ˧˩̰": 95,
      "iːŭ˧˩̰": 96,
      "œː˧˥": 97,
      "yː˧": 98,
      "uːĭ˩˧": 99,
      "ɵy˧˥": 100,
      "yː˧˩̰": 101,
      "ɔːĭ˧˥": 102,
      "ɛː": 103,
      "u˨": 104,
      "ou˧": 105,
      "ei˨": 106,
      "ɐu˥": 107,
      "ɵ˥": 108,
      "uː˧˩̰": 109,
      "yː˥": 110,
      "ɪ˥": 111,
      "œː˥": 112,
      "œː˧˩̰": 113,
      "aːĭ˨": 114,
      "ɐ˩˧": 115,
      "œː˧": 116,
      "uːĭ˧˥": 117,
      "ɐu˧": 118,
      "ɐi˧": 119,
      "ou˧˩̰": 120,
      "aːĭ˥": 121,
      "aːŭ˥": 122,
      "yː˧˥": 123,
      "iːŭ˥": 124,
      "ɔːĭ˨": 125,
      "ʊ˧˥": 126,
      "m̩˥": 127,
      "iːŭ˧˥": 128,
      "ɐi˩˧": 129,
      "ɵy˥": 130,
      "uːĭ˧": 131,
      "ɵy˧˩̰": 132,
      "uːĭ˥": 133,
      "aːŭ˧˩̰": 134,
      "yː˩˧": 135,
      "ɔːĭ˧˩̰": 136,
      "aːŭ˩˧": 137,
      "aːĭ˩˧": 138,
      "uːĭ˨": 139,
      "œː˨": 140,
      "uː˨": 141,
      "ɵy˨": 142,
      "aːŭ˨": 143,
      "m̩˩˧": 144,
      "ŋ˩˧": 145,
      "ɪ˧": 146,
      "m̩˨": 147,
      "iːŭ˩˧": 148,
      "iːŭ˨": 149,
      "ɵ˨": 150,
      "uːĭ˧˩̰": 151,
      "uː˩˧": 152,
      "ɵ˧˩̰": 153
    },
    "unk_token": "UNK"
  }
}