floriangardin commited on
Commit
0b9a7d6
1 Parent(s): a6aab70

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +9 -0
  2. tokenizer.json +638 -0
  3. tokenizer_config.json +80 -0
special_tokens_map.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[START]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[END]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": "[UNK]"
9
+ }
tokenizer.json ADDED
@@ -0,0 +1,638 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "version": "1.0",
3
+ "truncation": null,
4
+ "padding": null,
5
+ "added_tokens": [
6
+ {
7
+ "id": 0,
8
+ "content": "[UNK]",
9
+ "single_word": false,
10
+ "lstrip": false,
11
+ "rstrip": false,
12
+ "normalized": false,
13
+ "special": true
14
+ },
15
+ {
16
+ "id": 1,
17
+ "content": "[CLS]",
18
+ "single_word": false,
19
+ "lstrip": false,
20
+ "rstrip": false,
21
+ "normalized": false,
22
+ "special": true
23
+ },
24
+ {
25
+ "id": 2,
26
+ "content": "[SEP]",
27
+ "single_word": false,
28
+ "lstrip": false,
29
+ "rstrip": false,
30
+ "normalized": false,
31
+ "special": true
32
+ },
33
+ {
34
+ "id": 3,
35
+ "content": "[PAD]",
36
+ "single_word": false,
37
+ "lstrip": false,
38
+ "rstrip": false,
39
+ "normalized": false,
40
+ "special": true
41
+ },
42
+ {
43
+ "id": 4,
44
+ "content": "[MASK]",
45
+ "single_word": false,
46
+ "lstrip": false,
47
+ "rstrip": false,
48
+ "normalized": false,
49
+ "special": true
50
+ },
51
+ {
52
+ "id": 5,
53
+ "content": "[START]",
54
+ "single_word": false,
55
+ "lstrip": false,
56
+ "rstrip": false,
57
+ "normalized": false,
58
+ "special": true
59
+ },
60
+ {
61
+ "id": 556,
62
+ "content": "[END]",
63
+ "single_word": false,
64
+ "lstrip": false,
65
+ "rstrip": false,
66
+ "normalized": false,
67
+ "special": true
68
+ }
69
+ ],
70
+ "normalizer": null,
71
+ "pre_tokenizer": {
72
+ "type": "WhitespaceSplit"
73
+ },
74
+ "post_processor": null,
75
+ "decoder": null,
76
+ "model": {
77
+ "type": "WordLevel",
78
+ "vocab": {
79
+ "[UNK]": 0,
80
+ "[CLS]": 1,
81
+ "[SEP]": 2,
82
+ "[PAD]": 3,
83
+ "[MASK]": 4,
84
+ "[START]": 5,
85
+ "FM": 6,
86
+ "EM": 7,
87
+ "CM": 8,
88
+ "DM": 9,
89
+ "GM": 10,
90
+ "BbM": 11,
91
+ "AM": 12,
92
+ "EbM": 13,
93
+ "DbM": 14,
94
+ "AbM": 15,
95
+ "BM": 16,
96
+ "GbM": 17,
97
+ "Am": 18,
98
+ "Bm": 19,
99
+ "Gm": 20,
100
+ "Cm": 21,
101
+ "C#m": 22,
102
+ "Bbm": 23,
103
+ "Em": 24,
104
+ "F#m": 25,
105
+ "Fm": 26,
106
+ "Dm": 27,
107
+ "Ebm": 28,
108
+ "F#M": 29,
109
+ "FM/C": 30,
110
+ "EM/B": 31,
111
+ "GM/D": 32,
112
+ "AM/E": 33,
113
+ "BbM/F": 34,
114
+ "DM/A": 35,
115
+ "CM/G": 36,
116
+ "END": 37,
117
+ "START": 38,
118
+ "AbM/Eb": 39,
119
+ "EbM/Bb": 40,
120
+ "DbM/Ab": 41,
121
+ "G#m": 42,
122
+ "CbM": 43,
123
+ "BM/F#": 44,
124
+ "GbM/Db": 45,
125
+ "Abm": 46,
126
+ "Bm/F#": 47,
127
+ "Am/E": 48,
128
+ "Bbm/F": 49,
129
+ "Cm/G": 50,
130
+ "C#m/G#": 51,
131
+ "Em/B": 52,
132
+ "Fm/C": 53,
133
+ "Dm/A": 54,
134
+ "Gm/D": 55,
135
+ "F#m/C#": 56,
136
+ "Ebm/Bb": 57,
137
+ "E7": 58,
138
+ "F#7": 59,
139
+ "C7": 60,
140
+ "F7": 61,
141
+ "A7": 62,
142
+ "B7": 63,
143
+ "D7": 64,
144
+ "Bb7": 65,
145
+ "G7": 66,
146
+ "DM/F#": 67,
147
+ "EM/G#": 68,
148
+ "FM/A": 69,
149
+ "GM/B": 70,
150
+ "CM/E": 71,
151
+ "BbM/D": 72,
152
+ "AM/C#": 73,
153
+ "EbM/G": 74,
154
+ "DbM/F": 75,
155
+ "AbM/C": 76,
156
+ "F#M/C#": 77,
157
+ "G#m/D#": 78,
158
+ "Db7": 79,
159
+ "Eb7": 80,
160
+ "Ab7": 81,
161
+ "CbM/Gb": 82,
162
+ "D#m": 83,
163
+ "GbM/Bb": 84,
164
+ "BM/D#": 85,
165
+ "Gm/Bb": 86,
166
+ "Bbm/Db": 87,
167
+ "Bm/D": 88,
168
+ "Am/C": 89,
169
+ "Em/G": 90,
170
+ "F#m/A": 91,
171
+ "Cm/Eb": 92,
172
+ "Fm/Ab": 93,
173
+ "C#m/E": 94,
174
+ "Dm/F": 95,
175
+ "C#M": 96,
176
+ "D#M": 97,
177
+ "G#M": 98,
178
+ "Am7": 99,
179
+ "Bm7": 100,
180
+ "Cm7": 101,
181
+ "C#m7": 102,
182
+ "Dm7": 103,
183
+ "Em7": 104,
184
+ "Fm7": 105,
185
+ "Bbm7": 106,
186
+ "F#m7": 107,
187
+ "Gm7": 108,
188
+ "Esus4": 109,
189
+ "Fsus4": 110,
190
+ "Gsus4": 111,
191
+ "Bsus4": 112,
192
+ "Asus4": 113,
193
+ "Bbsus4": 114,
194
+ "Csus4": 115,
195
+ "Dsus4": 116,
196
+ "G#m/B": 117,
197
+ "Asus2": 118,
198
+ "Gsus2": 119,
199
+ "Bbsus2": 120,
200
+ "Dsus2": 121,
201
+ "Esus2": 122,
202
+ "Fsus2": 123,
203
+ "Ebsus2": 124,
204
+ "Bsus2": 125,
205
+ "Csus2": 126,
206
+ "Ebm/Gb": 127,
207
+ "A7/E": 128,
208
+ "B7/F#": 129,
209
+ "Bb7/F": 130,
210
+ "C7/G": 131,
211
+ "D7/A": 132,
212
+ "E7/B": 133,
213
+ "F#7/C#": 134,
214
+ "F7/C": 135,
215
+ "G7/D": 136,
216
+ "Ebsus4": 137,
217
+ "Ebm7": 138,
218
+ "Esus4/A": 139,
219
+ "Fsus4/Bb": 140,
220
+ "Asus4/D": 141,
221
+ "Csus4/F": 142,
222
+ "Gsus4/C": 143,
223
+ "Bsus4/E": 144,
224
+ "Dsus4/G": 145,
225
+ "Bbsus4/Eb": 146,
226
+ "D7/C": 147,
227
+ "A7/G": 148,
228
+ "B7/A": 149,
229
+ "Bb7/Ab": 150,
230
+ "C7/Bb": 151,
231
+ "E7/D": 152,
232
+ "F#7/E": 153,
233
+ "F7/Eb": 154,
234
+ "G7/F": 155,
235
+ "Ebsus4/Ab": 156,
236
+ "Absus4": 157,
237
+ "Db7/Ab": 158,
238
+ "Eb7/Bb": 159,
239
+ "Ab7/Eb": 160,
240
+ "Absus2": 161,
241
+ "Abm/Eb": 162,
242
+ "Absus4/Db": 163,
243
+ "Db7/Cb": 164,
244
+ "Eb7/Db": 165,
245
+ "Ab7/Gb": 166,
246
+ "Bm7/A": 167,
247
+ "C#m7/B": 168,
248
+ "Cm7/Bb": 169,
249
+ "Dm7/C": 170,
250
+ "Em7/D": 171,
251
+ "Am7/G": 172,
252
+ "F#m7/E": 173,
253
+ "Gm7/F": 174,
254
+ "Bbm7/Ab": 175,
255
+ "Fm7/Eb": 176,
256
+ "G#m7": 177,
257
+ "CbM/Eb": 178,
258
+ "Dbsus2": 179,
259
+ "Gbsus2": 180,
260
+ "Ebm7/Db": 181,
261
+ "Abm7": 182,
262
+ "F#M/A#": 183,
263
+ "A#dim": 184,
264
+ "Adim": 185,
265
+ "G#dim": 186,
266
+ "Bdim": 187,
267
+ "C#dim": 188,
268
+ "Edim": 189,
269
+ "F#dim": 190,
270
+ "D#dim": 191,
271
+ "Ddim": 192,
272
+ "F#sus4": 193,
273
+ "C#sus4": 194,
274
+ "Asus2/E": 195,
275
+ "Bbsus2/F": 196,
276
+ "Dsus2/A": 197,
277
+ "Ebsus2/Bb": 198,
278
+ "Gsus2/D": 199,
279
+ "Esus2/B": 200,
280
+ "Fsus2/C": 201,
281
+ "Csus2/G": 202,
282
+ "Bsus2/F#": 203,
283
+ "Dbsus4": 204,
284
+ "Gbsus4": 205,
285
+ "Gbsus4/Cb": 206,
286
+ "Dbsus4/Gb": 207,
287
+ "G#m7/F#": 208,
288
+ "G#7": 209,
289
+ "D#7": 210,
290
+ "C#7": 211,
291
+ "D#m/F#": 212,
292
+ "C#sus4/F#": 213,
293
+ "F#sus4/B": 214,
294
+ "Gdim": 215,
295
+ "Cdim": 216,
296
+ "Fdim": 217,
297
+ "Absus2/Eb": 218,
298
+ "Abm/Cb": 219,
299
+ "Bbsus4/F": 220,
300
+ "Dsus4/A": 221,
301
+ "Esus4/B": 222,
302
+ "Asus4/E": 223,
303
+ "Fsus4/C": 224,
304
+ "Gsus4/D": 225,
305
+ "Bsus4/F#": 226,
306
+ "Csus4/G": 227,
307
+ "Ebsus4/Bb": 228,
308
+ "Gbsus2/Db": 229,
309
+ "Dbsus2/Ab": 230,
310
+ "Am7/E": 231,
311
+ "Bbm7/F": 232,
312
+ "Bm7/F#": 233,
313
+ "Cm7/G": 234,
314
+ "C#m7/G#": 235,
315
+ "Gm7/D": 236,
316
+ "Dm7/A": 237,
317
+ "Em7/B": 238,
318
+ "F#m7/C#": 239,
319
+ "Fm7/C": 240,
320
+ "F#sus2": 241,
321
+ "C#sus2": 242,
322
+ "A#m7b5": 243,
323
+ "Am7b5": 244,
324
+ "Bm7b5": 245,
325
+ "C#m7b5": 246,
326
+ "Cm7b5": 247,
327
+ "D#m7b5": 248,
328
+ "Dm7b5": 249,
329
+ "Em7b5": 250,
330
+ "Fm7b5": 251,
331
+ "F#m7b5": 252,
332
+ "G#m7b5": 253,
333
+ "Gm7b5": 254,
334
+ "Absus4/Eb": 255,
335
+ "Ebm7/Bb": 256,
336
+ "D#M/A#": 257,
337
+ "C#M/G#": 258,
338
+ "G#M/D#": 259,
339
+ "Am7/C": 260,
340
+ "Em7/G": 261,
341
+ "F#m7/A": 262,
342
+ "Fm7/Ab": 263,
343
+ "Gm7/Bb": 264,
344
+ "Bbm7/Db": 265,
345
+ "Bm7/D": 266,
346
+ "Cm7/Eb": 267,
347
+ "C#m7/E": 268,
348
+ "Dm7/F": 269,
349
+ "D#m/A#": 270,
350
+ "D#m7": 271,
351
+ "Dsus2/E": 272,
352
+ "Ebsus2/F": 273,
353
+ "Fsus2/G": 274,
354
+ "Gsus2/A": 275,
355
+ "Asus2/B": 276,
356
+ "Bbsus2/C": 277,
357
+ "Esus2/F#": 278,
358
+ "Bsus2/C#": 279,
359
+ "Csus2/D": 280,
360
+ "Abm7/Gb": 281,
361
+ "G#sus4": 282,
362
+ "Dbsus4/Ab": 283,
363
+ "Gbsus4/Db": 284,
364
+ "G#m7/D#": 285,
365
+ "Absus2/Bb": 286,
366
+ "A#m7b5/G#": 287,
367
+ "Am7b5/G": 288,
368
+ "F#m7b5/E": 289,
369
+ "G#m7b5/F#": 290,
370
+ "Gm7b5/F": 291,
371
+ "Bm7b5/A": 292,
372
+ "C#m7b5/B": 293,
373
+ "Cm7b5/Bb": 294,
374
+ "D#m7b5/C#": 295,
375
+ "Dm7b5/C": 296,
376
+ "Em7b5/D": 297,
377
+ "Fm7b5/Eb": 298,
378
+ "D#dim/F#": 299,
379
+ "Edim/G": 300,
380
+ "F#dim/A": 301,
381
+ "Adim/C": 302,
382
+ "Bdim/D": 303,
383
+ "C#dim/E": 304,
384
+ "Ddim/F": 305,
385
+ "G#dim/B": 306,
386
+ "A#dim/C#": 307,
387
+ "Ebm7/Gb": 308,
388
+ "G#sus2": 309,
389
+ "G#m7/B": 310,
390
+ "Gbsus2/Ab": 311,
391
+ "Dbsus2/Eb": 312,
392
+ "G#sus4/C#": 313,
393
+ "Cdim/Eb": 314,
394
+ "Fdim/Ab": 315,
395
+ "Gdim/Bb": 316,
396
+ "C#dim/G": 317,
397
+ "D#dim/A": 318,
398
+ "Ddim/Ab": 319,
399
+ "Edim/Bb": 320,
400
+ "F#dim/C": 321,
401
+ "A#dim/E": 322,
402
+ "Adim/Eb": 323,
403
+ "G#dim/D": 324,
404
+ "Bdim/F": 325,
405
+ "D#7/A#": 326,
406
+ "C#7/G#": 327,
407
+ "G#7/D#": 328,
408
+ "A7/C#": 329,
409
+ "B7/D#": 330,
410
+ "Bb7/D": 331,
411
+ "C7/E": 332,
412
+ "D7/F#": 333,
413
+ "E7/G#": 334,
414
+ "F#7/A#": 335,
415
+ "F7/A": 336,
416
+ "G7/B": 337,
417
+ "F#sus4/C#": 338,
418
+ "C#sus4/G#": 339,
419
+ "C#sus2/G#": 340,
420
+ "F#sus2/C#": 341,
421
+ "Abm7/Eb": 342,
422
+ "A#m7b5/C#": 343,
423
+ "Am7b5/C": 344,
424
+ "Bm7b5/D": 345,
425
+ "C#m7b5/E": 346,
426
+ "Cm7b5/Eb": 347,
427
+ "D#m7b5/F#": 348,
428
+ "Dm7b5/F": 349,
429
+ "Em7b5/G": 350,
430
+ "F#m7b5/A": 351,
431
+ "Fm7b5/Ab": 352,
432
+ "G#m7b5/B": 353,
433
+ "Gm7b5/Bb": 354,
434
+ "D#m7/C#": 355,
435
+ "C#7/B": 356,
436
+ "D#7/C#": 357,
437
+ "G#7/F#": 358,
438
+ "Cdim/Gb": 359,
439
+ "Fdim/Cb": 360,
440
+ "Gdim/Db": 361,
441
+ "G#M/B#": 362,
442
+ "C#M/E#": 363,
443
+ "D#M/F##": 364,
444
+ "B#dim": 365,
445
+ "E#dim": 366,
446
+ "F##dim": 367,
447
+ "Ab7/C": 368,
448
+ "Db7/F": 369,
449
+ "Eb7/G": 370,
450
+ "AM7/G#": 371,
451
+ "AbM7/G": 372,
452
+ "BbM7/A": 373,
453
+ "CM7/B": 374,
454
+ "DM7/C#": 375,
455
+ "DbM7/C": 376,
456
+ "EM7/D#": 377,
457
+ "EbM7/D": 378,
458
+ "FM7/E": 379,
459
+ "GM7/F#": 380,
460
+ "GbM7/F": 381,
461
+ "Abm7/Cb": 382,
462
+ "D#sus4": 383,
463
+ "D#m7/F#": 384,
464
+ "G#sus2/D#": 385,
465
+ "C#sus2/D#": 386,
466
+ "F#sus2/G#": 387,
467
+ "C#m7b5/G": 388,
468
+ "Dm7b5/Ab": 389,
469
+ "Em7b5/Bb": 390,
470
+ "F#m7b5/C": 391,
471
+ "Fm7b5/Cb": 392,
472
+ "D#m7b5/A": 393,
473
+ "A#m7b5/E": 394,
474
+ "Am7b5/Eb": 395,
475
+ "Bm7b5/F": 396,
476
+ "Cm7b5/Gb": 397,
477
+ "G#m7b5/D": 398,
478
+ "Gm7b5/Db": 399,
479
+ "CbM7/Bb": 400,
480
+ "G#sus4/D#": 401,
481
+ "AM7": 402,
482
+ "AbM7": 403,
483
+ "BbM7": 404,
484
+ "CM7": 405,
485
+ "DM7": 406,
486
+ "DbM7": 407,
487
+ "EM7": 408,
488
+ "EbM7": 409,
489
+ "FM7": 410,
490
+ "GM7": 411,
491
+ "GbM7": 412,
492
+ "A#o": 413,
493
+ "Ao": 414,
494
+ "B#o": 415,
495
+ "Bo": 416,
496
+ "C#o": 417,
497
+ "D#o": 418,
498
+ "Do": 419,
499
+ "E#o": 420,
500
+ "Eo": 421,
501
+ "F##o": 422,
502
+ "F#o": 423,
503
+ "G#o": 424,
504
+ "BM7/A#": 425,
505
+ "D#sus4/G#": 426,
506
+ "D#m7/A#": 427,
507
+ "C#7/E#": 428,
508
+ "D#7/F##": 429,
509
+ "G#7/B#": 430,
510
+ "G#sus2/A#": 431,
511
+ "B#dim/F#": 432,
512
+ "E#dim/B": 433,
513
+ "F##dim/C#": 434,
514
+ "Aaug": 435,
515
+ "Abaug": 436,
516
+ "Bbaug": 437,
517
+ "Caug": 438,
518
+ "Daug": 439,
519
+ "Dbaug": 440,
520
+ "Eaug": 441,
521
+ "Ebaug": 442,
522
+ "Faug": 443,
523
+ "Gaug": 444,
524
+ "Gbaug": 445,
525
+ "Baug": 446,
526
+ "CbM7": 447,
527
+ "E#dim/G#": 448,
528
+ "F##dim/A#": 449,
529
+ "B#dim/D#": 450,
530
+ "BM7": 451,
531
+ "AM7/E": 452,
532
+ "AbM7/Eb": 453,
533
+ "BbM7/F": 454,
534
+ "CM7/G": 455,
535
+ "DM7/A": 456,
536
+ "DbM7/Ab": 457,
537
+ "EM7/B": 458,
538
+ "EbM7/Bb": 459,
539
+ "FM7/C": 460,
540
+ "GM7/D": 461,
541
+ "GbM7/Db": 462,
542
+ "A#o/C#": 463,
543
+ "Ao/C": 464,
544
+ "B#o/D#": 465,
545
+ "Bo/D": 466,
546
+ "C#o/E": 467,
547
+ "D#o/F#": 468,
548
+ "Do/F": 469,
549
+ "E#o/G#": 470,
550
+ "Eo/G": 471,
551
+ "F##o/A#": 472,
552
+ "F#o/A": 473,
553
+ "G#o/B": 474,
554
+ "Aaug/C#": 475,
555
+ "Abaug/C": 476,
556
+ "Bbaug/D": 477,
557
+ "Caug/E": 478,
558
+ "Daug/F#": 479,
559
+ "Dbaug/F": 480,
560
+ "Eaug/G#": 481,
561
+ "Ebaug/G": 482,
562
+ "Faug/A": 483,
563
+ "Gaug/B": 484,
564
+ "Gbaug/Bb": 485,
565
+ "Baug/D#": 486,
566
+ "A#o/E": 487,
567
+ "Ao/Eb": 488,
568
+ "B#o/F#": 489,
569
+ "Bo/F": 490,
570
+ "C#o/G": 491,
571
+ "D#o/A": 492,
572
+ "Do/Ab": 493,
573
+ "E#o/B": 494,
574
+ "Eo/Bb": 495,
575
+ "F##o/C#": 496,
576
+ "F#o/C": 497,
577
+ "G#o/D": 498,
578
+ "BM7/F#": 499,
579
+ "A#o/G": 500,
580
+ "Ao/Gb": 501,
581
+ "B#o/A": 502,
582
+ "Bo/Ab": 503,
583
+ "C#o/Bb": 504,
584
+ "D#o/C": 505,
585
+ "Do/Cb": 506,
586
+ "E#o/D": 507,
587
+ "Eo/Db": 508,
588
+ "F##o/E": 509,
589
+ "F#o/Eb": 510,
590
+ "G#o/F": 511,
591
+ "Aaug/E#": 512,
592
+ "Abaug/E": 513,
593
+ "Baug/F##": 514,
594
+ "Bbaug/F#": 515,
595
+ "Caug/G#": 516,
596
+ "Daug/A#": 517,
597
+ "Eaug/B#": 518,
598
+ "Ebaug/B": 519,
599
+ "Faug/C#": 520,
600
+ "Gaug/D#": 521,
601
+ "Gbaug/D": 522,
602
+ "Dbaug/A": 523,
603
+ "D#sus4/A#": 524,
604
+ "CbM7/Gb": 525,
605
+ "AM7/C#": 526,
606
+ "AbM7/C": 527,
607
+ "BbM7/D": 528,
608
+ "CM7/E": 529,
609
+ "DM7/F#": 530,
610
+ "DbM7/F": 531,
611
+ "EM7/G#": 532,
612
+ "EbM7/G": 533,
613
+ "FM7/A": 534,
614
+ "GM7/B": 535,
615
+ "GbM7/Bb": 536,
616
+ "BM7/D#": 537,
617
+ "CbM7/Eb": 538,
618
+ "Cbaug": 539,
619
+ "F#aug": 540,
620
+ "Aaug/B": 541,
621
+ "Abaug/Bb": 542,
622
+ "Bbaug/C": 543,
623
+ "Caug/D": 544,
624
+ "Daug/E": 545,
625
+ "Dbaug/Eb": 546,
626
+ "Eaug/F#": 547,
627
+ "Ebaug/F": 548,
628
+ "Faug/G": 549,
629
+ "Gaug/A": 550,
630
+ "Baug/C#": 551,
631
+ "Cbaug/Db": 552,
632
+ "Cbaug/Eb": 553,
633
+ "F#aug/G#": 554,
634
+ "Gbaug/Ab": 555
635
+ },
636
+ "unk_token": "[UNK]"
637
+ }
638
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "[UNK]",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "[CLS]",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "2": {
21
+ "content": "[SEP]",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "3": {
29
+ "content": "[PAD]",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "4": {
37
+ "content": "[MASK]",
38
+ "lstrip": false,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ },
44
+ "5": {
45
+ "content": "[START]",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false,
50
+ "special": true
51
+ },
52
+ "556": {
53
+ "content": "[END]",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": false,
58
+ "special": true
59
+ }
60
+ },
61
+ "bos_token": "[START]",
62
+ "bos_token_id": 5,
63
+ "clean_up_tokenization_spaces": true,
64
+ "cls_token": "[CLS]",
65
+ "cls_token_id": 1,
66
+ "eos_token": "[END]",
67
+ "eos_token_id": 556,
68
+ "mask_token": "[MASK]",
69
+ "mask_token_id": 4,
70
+ "model_max_len": 4096,
71
+ "model_max_length": 4096,
72
+ "pad_token": "[PAD]",
73
+ "pad_token_id": 3,
74
+ "seo_token_id": 2,
75
+ "sep_token": "[SEP]",
76
+ "tokenizer_class": "PreTrainedTokenizerFast",
77
+ "type": "Split",
78
+ "unk_token": "[UNK]",
79
+ "unk_token_id": 0
80
+ }