{ "_description": "IPA phone label set used for training. id→symbol mapping from checkpoint phone_to_id.", "_n_classes": 45, "_notes": [ "Class 0 ('sil') covers silence, pauses, and boundary markers (TIMIT h#/pau/epi, Buckeye SIL).", "TIMIT stop closure+release pairs (bcl+b, dcl+d, …) are merged into a single release phone.", "TIMIT 'er/axr' → ɹ̩ (syllabic r); 'el' → l̩; 'em' → m̩; 'en' → n̩.", "TIMIT 'ah/ax/ax-h' → ə (schwa); 'dx' → ɾ (flap); 'q' → ʔ (glottal stop).", "Buckeye nasalised vowel variants (aen, own, …) are merged with their oral counterparts.", "IPA symbol sort order follows Python's default Unicode sort on the training code." ], "id2phone": { "0": "sil", "1": "aɪ", "2": "aʊ", "3": "b", "4": "d", "5": "dʒ", "6": "eɪ", "7": "f", "8": "g", "9": "h", "10": "i", "11": "j", "12": "k", "13": "l", "14": "l̩", "15": "m", "16": "m̩", "17": "n", "18": "n̩", "19": "oʊ", "20": "p", "21": "r", "22": "s", "23": "t", "24": "tʃ", "25": "u", "26": "v", "27": "w", "28": "z", "29": "æ", "30": "ð", "31": "ŋ", "32": "ɑ", "33": "ɔ", "34": "ɔɪ", "35": "ə", "36": "ɛ", "37": "ɪ", "38": "ɹ̩", "39": "ɾ", "40": "ʃ", "41": "ʊ", "42": "ʒ", "43": "ʔ", "44": "θ" }, "phone2id": { "sil": 0, "aɪ": 1, "aʊ": 2, "b": 3, "d": 4, "dʒ": 5, "eɪ": 6, "f": 7, "g": 8, "h": 9, "i": 10, "j": 11, "k": 12, "l": 13, "l̩": 14, "m": 15, "m̩": 16, "n": 17, "n̩": 18, "oʊ": 19, "p": 20, "r": 21, "s": 22, "t": 23, "tʃ": 24, "u": 25, "v": 26, "w": 27, "z": 28, "æ": 29, "ð": 30, "ŋ": 31, "ɑ": 32, "ɔ": 33, "ɔɪ": 34, "ə": 35, "ɛ": 36, "ɪ": 37, "ɹ̩": 38, "ɾ": 39, "ʃ": 40, "ʊ": 41, "ʒ": 42, "ʔ": 43, "θ": 44 } }