Upload 4 files
Browse files- added_tokens.json +4 -0
- special_tokens_map (2).json +22 -0
- tokenizer_config (1).json +13 -0
- vocab (1).json +214 -0
added_tokens.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"</s>": 211,
|
3 |
+
"<s>": 210
|
4 |
+
}
|
special_tokens_map (2).json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
{
|
4 |
+
"content": "<s>",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": true,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false
|
9 |
+
},
|
10 |
+
{
|
11 |
+
"content": "</s>",
|
12 |
+
"lstrip": false,
|
13 |
+
"normalized": true,
|
14 |
+
"rstrip": false,
|
15 |
+
"single_word": false
|
16 |
+
}
|
17 |
+
],
|
18 |
+
"bos_token": "<s>",
|
19 |
+
"eos_token": "</s>",
|
20 |
+
"pad_token": "[PAD]",
|
21 |
+
"unk_token": "[UNK]"
|
22 |
+
}
|
tokenizer_config (1).json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"clean_up_tokenization_spaces": true,
|
4 |
+
"do_lower_case": false,
|
5 |
+
"eos_token": "</s>",
|
6 |
+
"model_max_length": 1000000000000000019884624838656,
|
7 |
+
"pad_token": "[PAD]",
|
8 |
+
"replace_word_delimiter_char": " ",
|
9 |
+
"target_lang": "kaz",
|
10 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
11 |
+
"unk_token": "[UNK]",
|
12 |
+
"word_delimiter_token": "|"
|
13 |
+
}
|
vocab (1).json
ADDED
@@ -0,0 +1,214 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"kaz": {
|
3 |
+
"!": 1,
|
4 |
+
"&": 2,
|
5 |
+
"'": 3,
|
6 |
+
"-": 4,
|
7 |
+
".": 5,
|
8 |
+
"/": 6,
|
9 |
+
":": 7,
|
10 |
+
"?": 8,
|
11 |
+
"A": 9,
|
12 |
+
"B": 10,
|
13 |
+
"C": 11,
|
14 |
+
"D": 12,
|
15 |
+
"E": 13,
|
16 |
+
"F": 14,
|
17 |
+
"G": 15,
|
18 |
+
"H": 16,
|
19 |
+
"I": 17,
|
20 |
+
"J": 18,
|
21 |
+
"K": 19,
|
22 |
+
"L": 20,
|
23 |
+
"M": 21,
|
24 |
+
"N": 22,
|
25 |
+
"O": 23,
|
26 |
+
"P": 24,
|
27 |
+
"Q": 25,
|
28 |
+
"R": 26,
|
29 |
+
"S": 27,
|
30 |
+
"T": 28,
|
31 |
+
"U": 29,
|
32 |
+
"V": 30,
|
33 |
+
"W": 31,
|
34 |
+
"X": 32,
|
35 |
+
"Y": 33,
|
36 |
+
"Z": 34,
|
37 |
+
"[": 35,
|
38 |
+
"[PAD]": 209,
|
39 |
+
"[UNK]": 208,
|
40 |
+
"]": 36,
|
41 |
+
"a": 37,
|
42 |
+
"b": 38,
|
43 |
+
"c": 39,
|
44 |
+
"d": 40,
|
45 |
+
"e": 41,
|
46 |
+
"f": 42,
|
47 |
+
"g": 43,
|
48 |
+
"h": 44,
|
49 |
+
"i": 45,
|
50 |
+
"j": 46,
|
51 |
+
"k": 47,
|
52 |
+
"l": 48,
|
53 |
+
"m": 49,
|
54 |
+
"n": 50,
|
55 |
+
"o": 51,
|
56 |
+
"p": 52,
|
57 |
+
"q": 53,
|
58 |
+
"r": 54,
|
59 |
+
"s": 55,
|
60 |
+
"t": 56,
|
61 |
+
"u": 57,
|
62 |
+
"v": 58,
|
63 |
+
"w": 59,
|
64 |
+
"x": 60,
|
65 |
+
"y": 61,
|
66 |
+
"z": 62,
|
67 |
+
"|": 0,
|
68 |
+
"}": 63,
|
69 |
+
"«": 64,
|
70 |
+
"®": 65,
|
71 |
+
"»": 66,
|
72 |
+
"Ü": 67,
|
73 |
+
"ã": 68,
|
74 |
+
"ä": 69,
|
75 |
+
"ó": 70,
|
76 |
+
"ö": 71,
|
77 |
+
"ü": 72,
|
78 |
+
"ğ": 73,
|
79 |
+
"ū": 74,
|
80 |
+
"ə": 75,
|
81 |
+
"ϵ": 76,
|
82 |
+
"І": 77,
|
83 |
+
"А": 78,
|
84 |
+
"Б": 79,
|
85 |
+
"В": 80,
|
86 |
+
"Г": 81,
|
87 |
+
"Д": 82,
|
88 |
+
"Е": 83,
|
89 |
+
"Ж": 84,
|
90 |
+
"З": 85,
|
91 |
+
"И": 86,
|
92 |
+
"Й": 87,
|
93 |
+
"К": 88,
|
94 |
+
"Л": 89,
|
95 |
+
"М": 90,
|
96 |
+
"Н": 91,
|
97 |
+
"О": 92,
|
98 |
+
"П": 93,
|
99 |
+
"Р": 94,
|
100 |
+
"С": 95,
|
101 |
+
"Т": 96,
|
102 |
+
"У": 97,
|
103 |
+
"Ф": 98,
|
104 |
+
"Х": 99,
|
105 |
+
"Ц": 100,
|
106 |
+
"Ч": 101,
|
107 |
+
"Ш": 102,
|
108 |
+
"Щ": 103,
|
109 |
+
"Ы": 104,
|
110 |
+
"Э": 105,
|
111 |
+
"Ю": 106,
|
112 |
+
"Я": 107,
|
113 |
+
"а": 108,
|
114 |
+
"б": 109,
|
115 |
+
"в": 110,
|
116 |
+
"г": 111,
|
117 |
+
"д": 112,
|
118 |
+
"е": 113,
|
119 |
+
"ж": 114,
|
120 |
+
"з": 115,
|
121 |
+
"и": 116,
|
122 |
+
"й": 117,
|
123 |
+
"к": 118,
|
124 |
+
"л": 119,
|
125 |
+
"м": 120,
|
126 |
+
"н": 121,
|
127 |
+
"о": 122,
|
128 |
+
"п": 123,
|
129 |
+
"р": 124,
|
130 |
+
"с": 125,
|
131 |
+
"т": 126,
|
132 |
+
"у": 127,
|
133 |
+
"ф": 128,
|
134 |
+
"х": 129,
|
135 |
+
"ц": 130,
|
136 |
+
"ч": 131,
|
137 |
+
"ш": 132,
|
138 |
+
"щ": 133,
|
139 |
+
"ъ": 134,
|
140 |
+
"ы": 135,
|
141 |
+
"ь": 136,
|
142 |
+
"э": 137,
|
143 |
+
"ю": 138,
|
144 |
+
"я": 139,
|
145 |
+
"ё": 140,
|
146 |
+
"і": 141,
|
147 |
+
"Ғ": 142,
|
148 |
+
"ғ": 143,
|
149 |
+
"Қ": 144,
|
150 |
+
"қ": 145,
|
151 |
+
"ң": 146,
|
152 |
+
"Ү": 147,
|
153 |
+
"ү": 148,
|
154 |
+
"Ұ": 149,
|
155 |
+
"ұ": 150,
|
156 |
+
"һ": 151,
|
157 |
+
"Ә": 152,
|
158 |
+
"ә": 153,
|
159 |
+
"Ө": 154,
|
160 |
+
"ө": 155,
|
161 |
+
"٪": 156,
|
162 |
+
"–": 157,
|
163 |
+
"™": 158,
|
164 |
+
"⏳": 159,
|
165 |
+
"⚐": 160,
|
166 |
+
"⚑": 161,
|
167 |
+
"⚖": 162,
|
168 |
+
"⚪": 163,
|
169 |
+
"️": 164,
|
170 |
+
"﹪": 165,
|
171 |
+
"🌍": 166,
|
172 |
+
"🌐": 167,
|
173 |
+
"🎨": 168,
|
174 |
+
"🎪": 169,
|
175 |
+
"🎭": 170,
|
176 |
+
"🎰": 171,
|
177 |
+
"🏠": 172,
|
178 |
+
"🏢": 173,
|
179 |
+
"🏬": 174,
|
180 |
+
"🏭": 175,
|
181 |
+
"🐾": 176,
|
182 |
+
"👅": 177,
|
183 |
+
"👑": 178,
|
184 |
+
"👤": 179,
|
185 |
+
"👽": 180,
|
186 |
+
"💊": 181,
|
187 |
+
"💯": 182,
|
188 |
+
"💱": 183,
|
189 |
+
"💲": 184,
|
190 |
+
"💻": 185,
|
191 |
+
"💼": 186,
|
192 |
+
"📃": 187,
|
193 |
+
"📅": 188,
|
194 |
+
"📆": 189,
|
195 |
+
"📈": 190,
|
196 |
+
"📌": 191,
|
197 |
+
"📍": 192,
|
198 |
+
"📖": 193,
|
199 |
+
"📘": 194,
|
200 |
+
"📝": 195,
|
201 |
+
"📞": 196,
|
202 |
+
"📱": 197,
|
203 |
+
"🔔": 198,
|
204 |
+
"🔠": 199,
|
205 |
+
"🔢": 200,
|
206 |
+
"🔤": 201,
|
207 |
+
"🔱": 202,
|
208 |
+
"🕑": 203,
|
209 |
+
"🗼": 204,
|
210 |
+
"😷": 205,
|
211 |
+
"🙍": 206,
|
212 |
+
"🚀": 207
|
213 |
+
}
|
214 |
+
}
|