DanielHesslow commited on
Commit
50fbecb
1 Parent(s): 1747b98

add tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +1 -0
  2. tokenizer.json +328 -0
  3. tokenizer_config.json +1 -0
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
1
+ {}
tokenizer.json ADDED
@@ -0,0 +1,328 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "version": "1.0",
3
+ "truncation": null,
4
+ "padding": null,
5
+ "added_tokens": [
6
+ {
7
+ "id": 1,
8
+ "content": "<PAD>",
9
+ "single_word": false,
10
+ "lstrip": false,
11
+ "rstrip": false,
12
+ "normalized": false,
13
+ "special": true
14
+ },
15
+ {
16
+ "id": 2,
17
+ "content": "<EOS>",
18
+ "single_word": false,
19
+ "lstrip": false,
20
+ "rstrip": false,
21
+ "normalized": false,
22
+ "special": true
23
+ },
24
+ {
25
+ "id": 3,
26
+ "content": "L",
27
+ "single_word": false,
28
+ "lstrip": false,
29
+ "rstrip": false,
30
+ "normalized": true,
31
+ "special": false
32
+ },
33
+ {
34
+ "id": 4,
35
+ "content": "A",
36
+ "single_word": false,
37
+ "lstrip": false,
38
+ "rstrip": false,
39
+ "normalized": true,
40
+ "special": false
41
+ },
42
+ {
43
+ "id": 5,
44
+ "content": "G",
45
+ "single_word": false,
46
+ "lstrip": false,
47
+ "rstrip": false,
48
+ "normalized": true,
49
+ "special": false
50
+ },
51
+ {
52
+ "id": 6,
53
+ "content": "V",
54
+ "single_word": false,
55
+ "lstrip": false,
56
+ "rstrip": false,
57
+ "normalized": true,
58
+ "special": false
59
+ },
60
+ {
61
+ "id": 7,
62
+ "content": "E",
63
+ "single_word": false,
64
+ "lstrip": false,
65
+ "rstrip": false,
66
+ "normalized": true,
67
+ "special": false
68
+ },
69
+ {
70
+ "id": 8,
71
+ "content": "S",
72
+ "single_word": false,
73
+ "lstrip": false,
74
+ "rstrip": false,
75
+ "normalized": true,
76
+ "special": false
77
+ },
78
+ {
79
+ "id": 9,
80
+ "content": "I",
81
+ "single_word": false,
82
+ "lstrip": false,
83
+ "rstrip": false,
84
+ "normalized": true,
85
+ "special": false
86
+ },
87
+ {
88
+ "id": 10,
89
+ "content": "K",
90
+ "single_word": false,
91
+ "lstrip": false,
92
+ "rstrip": false,
93
+ "normalized": true,
94
+ "special": false
95
+ },
96
+ {
97
+ "id": 11,
98
+ "content": "R",
99
+ "single_word": false,
100
+ "lstrip": false,
101
+ "rstrip": false,
102
+ "normalized": true,
103
+ "special": false
104
+ },
105
+ {
106
+ "id": 12,
107
+ "content": "D",
108
+ "single_word": false,
109
+ "lstrip": false,
110
+ "rstrip": false,
111
+ "normalized": true,
112
+ "special": false
113
+ },
114
+ {
115
+ "id": 13,
116
+ "content": "T",
117
+ "single_word": false,
118
+ "lstrip": false,
119
+ "rstrip": false,
120
+ "normalized": true,
121
+ "special": false
122
+ },
123
+ {
124
+ "id": 14,
125
+ "content": "P",
126
+ "single_word": false,
127
+ "lstrip": false,
128
+ "rstrip": false,
129
+ "normalized": true,
130
+ "special": false
131
+ },
132
+ {
133
+ "id": 15,
134
+ "content": "N",
135
+ "single_word": false,
136
+ "lstrip": false,
137
+ "rstrip": false,
138
+ "normalized": true,
139
+ "special": false
140
+ },
141
+ {
142
+ "id": 16,
143
+ "content": "Q",
144
+ "single_word": false,
145
+ "lstrip": false,
146
+ "rstrip": false,
147
+ "normalized": true,
148
+ "special": false
149
+ },
150
+ {
151
+ "id": 17,
152
+ "content": "F",
153
+ "single_word": false,
154
+ "lstrip": false,
155
+ "rstrip": false,
156
+ "normalized": true,
157
+ "special": false
158
+ },
159
+ {
160
+ "id": 18,
161
+ "content": "Y",
162
+ "single_word": false,
163
+ "lstrip": false,
164
+ "rstrip": false,
165
+ "normalized": true,
166
+ "special": false
167
+ },
168
+ {
169
+ "id": 19,
170
+ "content": "M",
171
+ "single_word": false,
172
+ "lstrip": false,
173
+ "rstrip": false,
174
+ "normalized": true,
175
+ "special": false
176
+ },
177
+ {
178
+ "id": 20,
179
+ "content": "H",
180
+ "single_word": false,
181
+ "lstrip": false,
182
+ "rstrip": false,
183
+ "normalized": true,
184
+ "special": false
185
+ },
186
+ {
187
+ "id": 21,
188
+ "content": "C",
189
+ "single_word": false,
190
+ "lstrip": false,
191
+ "rstrip": false,
192
+ "normalized": true,
193
+ "special": false
194
+ },
195
+ {
196
+ "id": 22,
197
+ "content": "W",
198
+ "single_word": false,
199
+ "lstrip": false,
200
+ "rstrip": false,
201
+ "normalized": true,
202
+ "special": false
203
+ },
204
+ {
205
+ "id": 23,
206
+ "content": "B",
207
+ "single_word": false,
208
+ "lstrip": false,
209
+ "rstrip": false,
210
+ "normalized": true,
211
+ "special": false
212
+ },
213
+ {
214
+ "id": 24,
215
+ "content": "O",
216
+ "single_word": false,
217
+ "lstrip": false,
218
+ "rstrip": false,
219
+ "normalized": true,
220
+ "special": false
221
+ },
222
+ {
223
+ "id": 25,
224
+ "content": "U",
225
+ "single_word": false,
226
+ "lstrip": false,
227
+ "rstrip": false,
228
+ "normalized": true,
229
+ "special": false
230
+ },
231
+ {
232
+ "id": 26,
233
+ "content": "Z",
234
+ "single_word": false,
235
+ "lstrip": false,
236
+ "rstrip": false,
237
+ "normalized": true,
238
+ "special": false
239
+ }
240
+ ],
241
+ "normalizer": {
242
+ "type": "Sequence",
243
+ "normalizers": [
244
+ {
245
+ "type": "Replace",
246
+ "pattern": {
247
+ "String": "X"
248
+ },
249
+ "content": "L"
250
+ },
251
+ {
252
+ "type": "Replace",
253
+ "pattern": {
254
+ "String": "B"
255
+ },
256
+ "content": "D"
257
+ },
258
+ {
259
+ "type": "Replace",
260
+ "pattern": {
261
+ "String": "Z"
262
+ },
263
+ "content": "E"
264
+ },
265
+ {
266
+ "type": "Replace",
267
+ "pattern": {
268
+ "String": "J"
269
+ },
270
+ "content": "L"
271
+ }
272
+ ]
273
+ },
274
+ "pre_tokenizer": null,
275
+ "post_processor": {
276
+ "type": "TemplateProcessing",
277
+ "single": [
278
+ {
279
+ "Sequence": {
280
+ "id": "A",
281
+ "type_id": 0
282
+ }
283
+ },
284
+ {
285
+ "SpecialToken": {
286
+ "id": "<EOS>",
287
+ "type_id": 0
288
+ }
289
+ }
290
+ ],
291
+ "pair": [
292
+ {
293
+ "Sequence": {
294
+ "id": "A",
295
+ "type_id": 0
296
+ }
297
+ },
298
+ {
299
+ "Sequence": {
300
+ "id": "B",
301
+ "type_id": 1
302
+ }
303
+ }
304
+ ],
305
+ "special_tokens": {
306
+ "<EOS>": {
307
+ "id": "<EOS>",
308
+ "ids": [
309
+ 2
310
+ ],
311
+ "tokens": [
312
+ "<EOS>"
313
+ ]
314
+ }
315
+ }
316
+ },
317
+ "decoder": null,
318
+ "model": {
319
+ "type": "Unigram",
320
+ "unk_id": 0,
321
+ "vocab": [
322
+ [
323
+ "<unk>",
324
+ 0.0
325
+ ]
326
+ ]
327
+ }
328
+ }
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
1
+ {"special_tokens_map_file": "Seledorn/RITA_tokenizer/special_tokens_map.json", "name_or_path": "Seledorn/RITA_tokenizer", "tokenizer_class": "PreTrainedTokenizerFast"}