Hugging Face's logo

pius-code
/
asha_twi_adapter

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Model card Files
xet
 Community
pius-code commited on
Commit
5b85f02
·
verified ·
1 Parent(s): cc1e595

Upload tokenizer

Browse files
Files changed (4) hide show
  1. added_tokens.json +2 -2
  2. special_tokens_map.json +6 -0
  3. tokenizer_config.json +5 -18
  4. vocab.json +52 -68
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
- "</s>": 70,
3
- "<s>": 69
4
  }
 
1
  {
2
+ "</s>": 54,
3
+ "<s>": 53
4
  }
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 CHANGED
@@ -1,14 +1,6 @@
1
  {
2
  "added_tokens_decoder": {
3
- "0": {
4
- "content": "|",
5
- "lstrip": false,
6
- "normalized": false,
7
- "rstrip": false,
8
- "single_word": false,
9
- "special": true
10
- },
11
- "67": {
12
  "content": "[UNK]",
13
  "lstrip": true,
14
  "normalized": false,
@@ -16,7 +8,7 @@
16
  "single_word": false,
17
  "special": false
18
  },
19
- "68": {
20
  "content": "[PAD]",
21
  "lstrip": true,
22
  "normalized": false,
@@ -24,7 +16,7 @@
24
  "single_word": false,
25
  "special": false
26
  },
27
- "69": {
28
  "content": "<s>",
29
  "lstrip": false,
30
  "normalized": false,
@@ -32,7 +24,7 @@
32
  "single_word": false,
33
  "special": true
34
  },
35
- "70": {
36
  "content": "</s>",
37
  "lstrip": false,
38
  "normalized": false,
@@ -41,16 +33,11 @@
41
  "special": true
42
  }
43
  },
44
- "backend": "custom",
45
  "bos_token": "<s>",
 
46
  "do_lower_case": false,
47
  "eos_token": "</s>",
48
- "is_local": true,
49
- "local_files_only": false,
50
  "model_max_length": 1000000000000000019884624838656,
51
- "model_specific_special_tokens": {
52
- "word_delimiter_token": "|"
53
- },
54
  "pad_token": "[PAD]",
55
  "replace_word_delimiter_char": " ",
56
  "target_lang": "aka",
 
1
  {
2
  "added_tokens_decoder": {
3
+ "51": {
 
 
 
 
 
 
 
 
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
 
8
  "single_word": false,
9
  "special": false
10
  },
11
+ "52": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
 
16
  "single_word": false,
17
  "special": false
18
  },
19
+ "53": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": false,
 
24
  "single_word": false,
25
  "special": true
26
  },
27
+ "54": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": false,
 
33
  "special": true
34
  }
35
  },
 
36
  "bos_token": "<s>",
37
+ "clean_up_tokenization_spaces": false,
38
  "do_lower_case": false,
39
  "eos_token": "</s>",
 
 
40
  "model_max_length": 1000000000000000019884624838656,
 
 
 
41
  "pad_token": "[PAD]",
42
  "replace_word_delimiter_char": " ",
43
  "target_lang": "aka",
vocab.json CHANGED
@@ -1,73 +1,57 @@
1
  {
2
  "aka": {
3
- "!": 1,
4
- "'": 2,
5
- ",": 3,
6
- ".": 4,
7
- ":": 5,
8
- "?": 6,
9
- "A": 7,
10
- "B": 8,
11
- "C": 9,
12
- "D": 10,
13
- "E": 11,
14
- "F": 12,
15
- "G": 13,
16
- "H": 14,
17
- "I": 15,
18
- "J": 16,
19
- "K": 17,
20
- "M": 18,
21
- "N": 19,
22
- "O": 20,
23
- "P": 21,
24
- "R": 22,
25
- "S": 23,
26
- "T": 24,
27
- "W": 25,
28
- "Y": 26,
29
- "[PAD]": 68,
30
- "[UNK]": 67,
31
- "a": 27,
32
- "b": 28,
33
- "c": 29,
34
- "d": 30,
35
- "e": 31,
36
- "f": 32,
37
- "g": 33,
38
- "h": 34,
39
- "i": 35,
40
- "k": 36,
41
- "l": 37,
42
- "m": 38,
43
- "n": 39,
44
- "o": 40,
45
- "p": 41,
46
- "r": 42,
47
- "s": 43,
48
- "t": 44,
49
- "u": 45,
50
- "v": 46,
51
- "w": 47,
52
- "y": 48,
53
- "z": 49,
54
  "|": 0,
55
- "·": 50,
56
- "á": 51,
57
- "ã": 52,
58
- "õ": 53,
59
- "ĩ": 54,
60
- "ũ": 55,
61
- "Ɔ": 56,
62
- "Ɛ": 57,
63
- "ɔ": 58,
64
- "ɛ": 59,
65
- "ʹ": 60,
66
- "́": 61,
67
- "—": 62,
68
- "‘": 63,
69
- "’": 64,
70
- "“": 65,
71
- "•": 66
 
72
  }
73
  }
 
1
  {
2
  "aka": {
3
+ ")": 1,
4
+ "0": 2,
5
+ "3": 3,
6
+ "4": 4,
7
+ "5": 5,
8
+ "6": 6,
9
+ "[PAD]": 52,
10
+ "[UNK]": 51,
11
+ "a": 7,
12
+ "b": 8,
13
+ "c": 9,
14
+ "d": 10,
15
+ "e": 11,
16
+ "f": 12,
17
+ "g": 13,
18
+ "h": 14,
19
+ "i": 15,
20
+ "j": 16,
21
+ "k": 17,
22
+ "l": 18,
23
+ "m": 19,
24
+ "n": 20,
25
+ "o": 21,
26
+ "p": 22,
27
+ "q": 23,
28
+ "r": 24,
29
+ "s": 25,
30
+ "t": 26,
31
+ "u": 27,
32
+ "v": 28,
33
+ "w": 29,
34
+ "x": 30,
35
+ "y": 31,
36
+ "z": 32,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
37
  "|": 0,
38
+ "·": 33,
39
+ "á": 34,
40
+ "ã": 35,
41
+ "õ": 36,
42
+ "ĩ": 37,
43
+ "ũ": 38,
44
+ "ɔ": 39,
45
+ "ɛ": 40,
46
+ "ʹ": 41,
47
+ "́": 42,
48
+ "—": 43,
49
+ "‘": 44,
50
+ "’": 45,
51
+ "“": 46,
52
+ "•": 47,
53
+ "…": 48,
54
+ "ℇ": 49,
55
+ "ↄ": 50
56
  }
57
  }