padmalcom commited on
Commit
a81d3fa
1 Parent(s): 78c722d

Upload 3 files

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +6 -0
  2. tokenizer_config.json +11 -0
  3. vocab.json +59 -0
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "<pad>",
5
+ "unk_token": "<unk>"
6
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "do_lower_case": false,
4
+ "eos_token": "</s>",
5
+ "model_max_length": 1000000000000000019884624838656,
6
+ "pad_token": "<pad>",
7
+ "replace_word_delimiter_char": " ",
8
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
9
+ "unk_token": "<unk>",
10
+ "word_delimiter_token": "|"
11
+ }
vocab.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</s>": 2,
3
+ "<pad>": 0,
4
+ "<s>": 1,
5
+ "<unk>": 3,
6
+ "A": 5,
7
+ "B": 6,
8
+ "C": 7,
9
+ "D": 8,
10
+ "E": 9,
11
+ "F": 10,
12
+ "G": 11,
13
+ "H": 12,
14
+ "I": 13,
15
+ "J": 14,
16
+ "K": 15,
17
+ "L": 16,
18
+ "M": 17,
19
+ "N": 18,
20
+ "O": 19,
21
+ "P": 20,
22
+ "Q": 21,
23
+ "R": 22,
24
+ "S": 23,
25
+ "T": 24,
26
+ "U": 25,
27
+ "V": 26,
28
+ "W": 27,
29
+ "X": 28,
30
+ "Y": 29,
31
+ "Z": 30,
32
+ "a": 31,
33
+ "b": 32,
34
+ "c": 33,
35
+ "d": 34,
36
+ "e": 35,
37
+ "f": 36,
38
+ "g": 37,
39
+ "h": 38,
40
+ "i": 39,
41
+ "j": 40,
42
+ "k": 41,
43
+ "l": 42,
44
+ "m": 43,
45
+ "n": 44,
46
+ "o": 45,
47
+ "p": 46,
48
+ "q": 47,
49
+ "r": 48,
50
+ "s": 49,
51
+ "t": 50,
52
+ "u": 51,
53
+ "v": 52,
54
+ "w": 53,
55
+ "x": 54,
56
+ "y": 55,
57
+ "z": 56,
58
+ "|": 4
59
+ }