jessiejohnson commited on
Commit
ccb534e
1 Parent(s): b019dea

add tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +1 -0
  2. tokenizer_config.json +1 -0
  3. vocab.json +1 -0
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
1
+ {"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "tokenizer_class": "Wav2Vec2CTCTokenizer"}
vocab.json ADDED
@@ -0,0 +1 @@
 
1
+ {"aa": 0, "ae": 1, "ah": 2, "ao": 3, "aw": 4, "ax": 5, "ax-h": 6, "axr": 7, "ay": 8, "b": 9, "bcl": 10, "ch": 11, "d": 12, "dcl": 13, "dh": 14, "dx": 15, "eh": 16, "el": 17, "em": 18, "en": 19, "eng": 20, "epi": 21, "er": 22, "ey": 23, "f": 24, "g": 25, "gcl": 26, "hh": 27, "hv": 28, "ih": 29, "ix": 30, "iy": 31, "jh": 32, "k": 33, "kcl": 34, "l": 35, "m": 36, "n": 37, "ng": 38, "nx": 39, "ow": 40, "oy": 41, "p": 42, "pau": 43, "pcl": 44, "q": 45, "r": 46, "s": 47, "sh": 48, "t": 49, "tcl": 50, "th": 51, "uh": 52, "uw": 53, "ux": 54, "v": 55, "w": 56, "y": 57, "z": 58, "zh": 59, "|": 60, "[UNK]": 61, "[PAD]": 62}