{ "1": 3, "2": 4, "3": 5, "4": 6, "5": 7, "6": 8, "M": 9, "Ng": 10, "[PAD]": 0, "[UNK]": 1, "aa": 11, "aai": 12, "aak": 13, "aam": 14, "aan": 15, "aang": 16, "aap": 17, "aat": 18, "aau": 19, "ai": 20, "ak": 21, "am": 22, "an": 23, "ang": 24, "ap": 25, "at": 26, "au": 27, "b": 28, "c": 29, "d": 30, "e": 31, "ei": 32, "ek": 33, "eng": 34, "eoi": 35, "eon": 36, "eot": 37, "f": 38, "g": 39, "gw": 40, "h": 41, "i": 42, "ik": 43, "im": 44, "in": 45, "ing": 46, "ip": 47, "it": 48, "iu": 49, "j": 50, "k": 51, "kw": 52, "l": 53, "m": 54, "n": 55, "ng": 56, "o": 57, "oe": 58, "oek": 59, "oeng": 60, "oi": 61, "ok": 62, "on": 63, "ong": 64, "ot": 65, "ou": 66, "p": 67, "s": 68, "t": 69, "u": 70, "ui": 71, "uk": 72, "un": 73, "ung": 74, "ut": 75, "w": 76, "yu": 77, "yun": 78, "yut": 79, "z": 80, "|": 2 }