Flova commited on
Commit
1a7f682
1 Parent(s): 8bdc6e0

Upload processor

Browse files
Files changed (3) hide show
  1. added_tokens.json +2 -76
  2. tokenizer.json +1 -667
  3. tokenizer_config.json +1 -0
added_tokens.json CHANGED
@@ -1,80 +1,6 @@
1
  {
 
2
  "<s_iitcdip>": 57523,
3
  "<s_synthdog>": 57524,
4
- "<sep/>": 57522,
5
- "a''1 ": 57580,
6
- "a''16 ": 57584,
7
- "a''2 ": 57581,
8
- "a''4 ": 57582,
9
- "a''8 ": 57583,
10
- "a'1 ": 57575,
11
- "a'16 ": 57579,
12
- "a'2 ": 57576,
13
- "a'4 ": 57577,
14
- "a'8 ": 57578,
15
- "b''1 ": 57590,
16
- "b''16 ": 57594,
17
- "b''2 ": 57591,
18
- "b''4 ": 57592,
19
- "b''8 ": 57593,
20
- "b'1 ": 57585,
21
- "b'16 ": 57589,
22
- "b'2 ": 57586,
23
- "b'4 ": 57587,
24
- "b'8 ": 57588,
25
- "c''1 ": 57530,
26
- "c''16 ": 57534,
27
- "c''2 ": 57531,
28
- "c''4 ": 57532,
29
- "c''8 ": 57533,
30
- "c'1 ": 57525,
31
- "c'16 ": 57529,
32
- "c'2 ": 57526,
33
- "c'4 ": 57527,
34
- "c'8 ": 57528,
35
- "d''1 ": 57540,
36
- "d''16 ": 57544,
37
- "d''2 ": 57541,
38
- "d''4 ": 57542,
39
- "d''8 ": 57543,
40
- "d'1 ": 57535,
41
- "d'16 ": 57539,
42
- "d'2 ": 57536,
43
- "d'4 ": 57537,
44
- "d'8 ": 57538,
45
- "e''1 ": 57550,
46
- "e''16 ": 57554,
47
- "e''2 ": 57551,
48
- "e''4 ": 57552,
49
- "e''8 ": 57553,
50
- "e'1 ": 57545,
51
- "e'16 ": 57549,
52
- "e'2 ": 57546,
53
- "e'4 ": 57547,
54
- "e'8 ": 57548,
55
- "f''1 ": 57560,
56
- "f''16 ": 57564,
57
- "f''2 ": 57561,
58
- "f''4 ": 57562,
59
- "f''8 ": 57563,
60
- "f'1 ": 57555,
61
- "f'16 ": 57559,
62
- "f'2 ": 57556,
63
- "f'4 ": 57557,
64
- "f'8 ": 57558,
65
- "g''1 ": 57570,
66
- "g''16 ": 57574,
67
- "g''2 ": 57571,
68
- "g''4 ": 57572,
69
- "g''8 ": 57573,
70
- "g'1 ": 57565,
71
- "g'16 ": 57569,
72
- "g'2 ": 57566,
73
- "g'4 ": 57567,
74
- "g'8 ": 57568,
75
- "r1 ": 57595,
76
- "r16 ": 57599,
77
- "r2 ": 57596,
78
- "r4 ": 57597,
79
- "r8 ": 57598
80
  }
 
1
  {
2
+ "1 ": 57525,
3
  "<s_iitcdip>": 57523,
4
  "<s_synthdog>": 57524,
5
+ "<sep/>": 57522
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6
  }
tokenizer.json CHANGED
@@ -77,673 +77,7 @@
77
  },
78
  {
79
  "id": 57525,
80
- "content": "c'1 ",
81
- "single_word": false,
82
- "lstrip": false,
83
- "rstrip": false,
84
- "normalized": true,
85
- "special": false
86
- },
87
- {
88
- "id": 57526,
89
- "content": "c'2 ",
90
- "single_word": false,
91
- "lstrip": false,
92
- "rstrip": false,
93
- "normalized": true,
94
- "special": false
95
- },
96
- {
97
- "id": 57527,
98
- "content": "c'4 ",
99
- "single_word": false,
100
- "lstrip": false,
101
- "rstrip": false,
102
- "normalized": true,
103
- "special": false
104
- },
105
- {
106
- "id": 57528,
107
- "content": "c'8 ",
108
- "single_word": false,
109
- "lstrip": false,
110
- "rstrip": false,
111
- "normalized": true,
112
- "special": false
113
- },
114
- {
115
- "id": 57529,
116
- "content": "c'16 ",
117
- "single_word": false,
118
- "lstrip": false,
119
- "rstrip": false,
120
- "normalized": true,
121
- "special": false
122
- },
123
- {
124
- "id": 57530,
125
- "content": "c''1 ",
126
- "single_word": false,
127
- "lstrip": false,
128
- "rstrip": false,
129
- "normalized": true,
130
- "special": false
131
- },
132
- {
133
- "id": 57531,
134
- "content": "c''2 ",
135
- "single_word": false,
136
- "lstrip": false,
137
- "rstrip": false,
138
- "normalized": true,
139
- "special": false
140
- },
141
- {
142
- "id": 57532,
143
- "content": "c''4 ",
144
- "single_word": false,
145
- "lstrip": false,
146
- "rstrip": false,
147
- "normalized": true,
148
- "special": false
149
- },
150
- {
151
- "id": 57533,
152
- "content": "c''8 ",
153
- "single_word": false,
154
- "lstrip": false,
155
- "rstrip": false,
156
- "normalized": true,
157
- "special": false
158
- },
159
- {
160
- "id": 57534,
161
- "content": "c''16 ",
162
- "single_word": false,
163
- "lstrip": false,
164
- "rstrip": false,
165
- "normalized": true,
166
- "special": false
167
- },
168
- {
169
- "id": 57535,
170
- "content": "d'1 ",
171
- "single_word": false,
172
- "lstrip": false,
173
- "rstrip": false,
174
- "normalized": true,
175
- "special": false
176
- },
177
- {
178
- "id": 57536,
179
- "content": "d'2 ",
180
- "single_word": false,
181
- "lstrip": false,
182
- "rstrip": false,
183
- "normalized": true,
184
- "special": false
185
- },
186
- {
187
- "id": 57537,
188
- "content": "d'4 ",
189
- "single_word": false,
190
- "lstrip": false,
191
- "rstrip": false,
192
- "normalized": true,
193
- "special": false
194
- },
195
- {
196
- "id": 57538,
197
- "content": "d'8 ",
198
- "single_word": false,
199
- "lstrip": false,
200
- "rstrip": false,
201
- "normalized": true,
202
- "special": false
203
- },
204
- {
205
- "id": 57539,
206
- "content": "d'16 ",
207
- "single_word": false,
208
- "lstrip": false,
209
- "rstrip": false,
210
- "normalized": true,
211
- "special": false
212
- },
213
- {
214
- "id": 57540,
215
- "content": "d''1 ",
216
- "single_word": false,
217
- "lstrip": false,
218
- "rstrip": false,
219
- "normalized": true,
220
- "special": false
221
- },
222
- {
223
- "id": 57541,
224
- "content": "d''2 ",
225
- "single_word": false,
226
- "lstrip": false,
227
- "rstrip": false,
228
- "normalized": true,
229
- "special": false
230
- },
231
- {
232
- "id": 57542,
233
- "content": "d''4 ",
234
- "single_word": false,
235
- "lstrip": false,
236
- "rstrip": false,
237
- "normalized": true,
238
- "special": false
239
- },
240
- {
241
- "id": 57543,
242
- "content": "d''8 ",
243
- "single_word": false,
244
- "lstrip": false,
245
- "rstrip": false,
246
- "normalized": true,
247
- "special": false
248
- },
249
- {
250
- "id": 57544,
251
- "content": "d''16 ",
252
- "single_word": false,
253
- "lstrip": false,
254
- "rstrip": false,
255
- "normalized": true,
256
- "special": false
257
- },
258
- {
259
- "id": 57545,
260
- "content": "e'1 ",
261
- "single_word": false,
262
- "lstrip": false,
263
- "rstrip": false,
264
- "normalized": true,
265
- "special": false
266
- },
267
- {
268
- "id": 57546,
269
- "content": "e'2 ",
270
- "single_word": false,
271
- "lstrip": false,
272
- "rstrip": false,
273
- "normalized": true,
274
- "special": false
275
- },
276
- {
277
- "id": 57547,
278
- "content": "e'4 ",
279
- "single_word": false,
280
- "lstrip": false,
281
- "rstrip": false,
282
- "normalized": true,
283
- "special": false
284
- },
285
- {
286
- "id": 57548,
287
- "content": "e'8 ",
288
- "single_word": false,
289
- "lstrip": false,
290
- "rstrip": false,
291
- "normalized": true,
292
- "special": false
293
- },
294
- {
295
- "id": 57549,
296
- "content": "e'16 ",
297
- "single_word": false,
298
- "lstrip": false,
299
- "rstrip": false,
300
- "normalized": true,
301
- "special": false
302
- },
303
- {
304
- "id": 57550,
305
- "content": "e''1 ",
306
- "single_word": false,
307
- "lstrip": false,
308
- "rstrip": false,
309
- "normalized": true,
310
- "special": false
311
- },
312
- {
313
- "id": 57551,
314
- "content": "e''2 ",
315
- "single_word": false,
316
- "lstrip": false,
317
- "rstrip": false,
318
- "normalized": true,
319
- "special": false
320
- },
321
- {
322
- "id": 57552,
323
- "content": "e''4 ",
324
- "single_word": false,
325
- "lstrip": false,
326
- "rstrip": false,
327
- "normalized": true,
328
- "special": false
329
- },
330
- {
331
- "id": 57553,
332
- "content": "e''8 ",
333
- "single_word": false,
334
- "lstrip": false,
335
- "rstrip": false,
336
- "normalized": true,
337
- "special": false
338
- },
339
- {
340
- "id": 57554,
341
- "content": "e''16 ",
342
- "single_word": false,
343
- "lstrip": false,
344
- "rstrip": false,
345
- "normalized": true,
346
- "special": false
347
- },
348
- {
349
- "id": 57555,
350
- "content": "f'1 ",
351
- "single_word": false,
352
- "lstrip": false,
353
- "rstrip": false,
354
- "normalized": true,
355
- "special": false
356
- },
357
- {
358
- "id": 57556,
359
- "content": "f'2 ",
360
- "single_word": false,
361
- "lstrip": false,
362
- "rstrip": false,
363
- "normalized": true,
364
- "special": false
365
- },
366
- {
367
- "id": 57557,
368
- "content": "f'4 ",
369
- "single_word": false,
370
- "lstrip": false,
371
- "rstrip": false,
372
- "normalized": true,
373
- "special": false
374
- },
375
- {
376
- "id": 57558,
377
- "content": "f'8 ",
378
- "single_word": false,
379
- "lstrip": false,
380
- "rstrip": false,
381
- "normalized": true,
382
- "special": false
383
- },
384
- {
385
- "id": 57559,
386
- "content": "f'16 ",
387
- "single_word": false,
388
- "lstrip": false,
389
- "rstrip": false,
390
- "normalized": true,
391
- "special": false
392
- },
393
- {
394
- "id": 57560,
395
- "content": "f''1 ",
396
- "single_word": false,
397
- "lstrip": false,
398
- "rstrip": false,
399
- "normalized": true,
400
- "special": false
401
- },
402
- {
403
- "id": 57561,
404
- "content": "f''2 ",
405
- "single_word": false,
406
- "lstrip": false,
407
- "rstrip": false,
408
- "normalized": true,
409
- "special": false
410
- },
411
- {
412
- "id": 57562,
413
- "content": "f''4 ",
414
- "single_word": false,
415
- "lstrip": false,
416
- "rstrip": false,
417
- "normalized": true,
418
- "special": false
419
- },
420
- {
421
- "id": 57563,
422
- "content": "f''8 ",
423
- "single_word": false,
424
- "lstrip": false,
425
- "rstrip": false,
426
- "normalized": true,
427
- "special": false
428
- },
429
- {
430
- "id": 57564,
431
- "content": "f''16 ",
432
- "single_word": false,
433
- "lstrip": false,
434
- "rstrip": false,
435
- "normalized": true,
436
- "special": false
437
- },
438
- {
439
- "id": 57565,
440
- "content": "g'1 ",
441
- "single_word": false,
442
- "lstrip": false,
443
- "rstrip": false,
444
- "normalized": true,
445
- "special": false
446
- },
447
- {
448
- "id": 57566,
449
- "content": "g'2 ",
450
- "single_word": false,
451
- "lstrip": false,
452
- "rstrip": false,
453
- "normalized": true,
454
- "special": false
455
- },
456
- {
457
- "id": 57567,
458
- "content": "g'4 ",
459
- "single_word": false,
460
- "lstrip": false,
461
- "rstrip": false,
462
- "normalized": true,
463
- "special": false
464
- },
465
- {
466
- "id": 57568,
467
- "content": "g'8 ",
468
- "single_word": false,
469
- "lstrip": false,
470
- "rstrip": false,
471
- "normalized": true,
472
- "special": false
473
- },
474
- {
475
- "id": 57569,
476
- "content": "g'16 ",
477
- "single_word": false,
478
- "lstrip": false,
479
- "rstrip": false,
480
- "normalized": true,
481
- "special": false
482
- },
483
- {
484
- "id": 57570,
485
- "content": "g''1 ",
486
- "single_word": false,
487
- "lstrip": false,
488
- "rstrip": false,
489
- "normalized": true,
490
- "special": false
491
- },
492
- {
493
- "id": 57571,
494
- "content": "g''2 ",
495
- "single_word": false,
496
- "lstrip": false,
497
- "rstrip": false,
498
- "normalized": true,
499
- "special": false
500
- },
501
- {
502
- "id": 57572,
503
- "content": "g''4 ",
504
- "single_word": false,
505
- "lstrip": false,
506
- "rstrip": false,
507
- "normalized": true,
508
- "special": false
509
- },
510
- {
511
- "id": 57573,
512
- "content": "g''8 ",
513
- "single_word": false,
514
- "lstrip": false,
515
- "rstrip": false,
516
- "normalized": true,
517
- "special": false
518
- },
519
- {
520
- "id": 57574,
521
- "content": "g''16 ",
522
- "single_word": false,
523
- "lstrip": false,
524
- "rstrip": false,
525
- "normalized": true,
526
- "special": false
527
- },
528
- {
529
- "id": 57575,
530
- "content": "a'1 ",
531
- "single_word": false,
532
- "lstrip": false,
533
- "rstrip": false,
534
- "normalized": true,
535
- "special": false
536
- },
537
- {
538
- "id": 57576,
539
- "content": "a'2 ",
540
- "single_word": false,
541
- "lstrip": false,
542
- "rstrip": false,
543
- "normalized": true,
544
- "special": false
545
- },
546
- {
547
- "id": 57577,
548
- "content": "a'4 ",
549
- "single_word": false,
550
- "lstrip": false,
551
- "rstrip": false,
552
- "normalized": true,
553
- "special": false
554
- },
555
- {
556
- "id": 57578,
557
- "content": "a'8 ",
558
- "single_word": false,
559
- "lstrip": false,
560
- "rstrip": false,
561
- "normalized": true,
562
- "special": false
563
- },
564
- {
565
- "id": 57579,
566
- "content": "a'16 ",
567
- "single_word": false,
568
- "lstrip": false,
569
- "rstrip": false,
570
- "normalized": true,
571
- "special": false
572
- },
573
- {
574
- "id": 57580,
575
- "content": "a''1 ",
576
- "single_word": false,
577
- "lstrip": false,
578
- "rstrip": false,
579
- "normalized": true,
580
- "special": false
581
- },
582
- {
583
- "id": 57581,
584
- "content": "a''2 ",
585
- "single_word": false,
586
- "lstrip": false,
587
- "rstrip": false,
588
- "normalized": true,
589
- "special": false
590
- },
591
- {
592
- "id": 57582,
593
- "content": "a''4 ",
594
- "single_word": false,
595
- "lstrip": false,
596
- "rstrip": false,
597
- "normalized": true,
598
- "special": false
599
- },
600
- {
601
- "id": 57583,
602
- "content": "a''8 ",
603
- "single_word": false,
604
- "lstrip": false,
605
- "rstrip": false,
606
- "normalized": true,
607
- "special": false
608
- },
609
- {
610
- "id": 57584,
611
- "content": "a''16 ",
612
- "single_word": false,
613
- "lstrip": false,
614
- "rstrip": false,
615
- "normalized": true,
616
- "special": false
617
- },
618
- {
619
- "id": 57585,
620
- "content": "b'1 ",
621
- "single_word": false,
622
- "lstrip": false,
623
- "rstrip": false,
624
- "normalized": true,
625
- "special": false
626
- },
627
- {
628
- "id": 57586,
629
- "content": "b'2 ",
630
- "single_word": false,
631
- "lstrip": false,
632
- "rstrip": false,
633
- "normalized": true,
634
- "special": false
635
- },
636
- {
637
- "id": 57587,
638
- "content": "b'4 ",
639
- "single_word": false,
640
- "lstrip": false,
641
- "rstrip": false,
642
- "normalized": true,
643
- "special": false
644
- },
645
- {
646
- "id": 57588,
647
- "content": "b'8 ",
648
- "single_word": false,
649
- "lstrip": false,
650
- "rstrip": false,
651
- "normalized": true,
652
- "special": false
653
- },
654
- {
655
- "id": 57589,
656
- "content": "b'16 ",
657
- "single_word": false,
658
- "lstrip": false,
659
- "rstrip": false,
660
- "normalized": true,
661
- "special": false
662
- },
663
- {
664
- "id": 57590,
665
- "content": "b''1 ",
666
- "single_word": false,
667
- "lstrip": false,
668
- "rstrip": false,
669
- "normalized": true,
670
- "special": false
671
- },
672
- {
673
- "id": 57591,
674
- "content": "b''2 ",
675
- "single_word": false,
676
- "lstrip": false,
677
- "rstrip": false,
678
- "normalized": true,
679
- "special": false
680
- },
681
- {
682
- "id": 57592,
683
- "content": "b''4 ",
684
- "single_word": false,
685
- "lstrip": false,
686
- "rstrip": false,
687
- "normalized": true,
688
- "special": false
689
- },
690
- {
691
- "id": 57593,
692
- "content": "b''8 ",
693
- "single_word": false,
694
- "lstrip": false,
695
- "rstrip": false,
696
- "normalized": true,
697
- "special": false
698
- },
699
- {
700
- "id": 57594,
701
- "content": "b''16 ",
702
- "single_word": false,
703
- "lstrip": false,
704
- "rstrip": false,
705
- "normalized": true,
706
- "special": false
707
- },
708
- {
709
- "id": 57595,
710
- "content": "r1 ",
711
- "single_word": false,
712
- "lstrip": false,
713
- "rstrip": false,
714
- "normalized": true,
715
- "special": false
716
- },
717
- {
718
- "id": 57596,
719
- "content": "r2 ",
720
- "single_word": false,
721
- "lstrip": false,
722
- "rstrip": false,
723
- "normalized": true,
724
- "special": false
725
- },
726
- {
727
- "id": 57597,
728
- "content": "r4 ",
729
- "single_word": false,
730
- "lstrip": false,
731
- "rstrip": false,
732
- "normalized": true,
733
- "special": false
734
- },
735
- {
736
- "id": 57598,
737
- "content": "r8 ",
738
- "single_word": false,
739
- "lstrip": false,
740
- "rstrip": false,
741
- "normalized": true,
742
- "special": false
743
- },
744
- {
745
- "id": 57599,
746
- "content": "r16 ",
747
  "single_word": false,
748
  "lstrip": false,
749
  "rstrip": false,
 
77
  },
78
  {
79
  "id": 57525,
80
+ "content": "1 ",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
81
  "single_word": false,
82
  "lstrip": false,
83
  "rstrip": false,
tokenizer_config.json CHANGED
@@ -11,6 +11,7 @@
11
  "single_word": false
12
  },
13
  "model_max_length": 1000000000000000019884624838656,
 
14
  "pad_token": "<pad>",
15
  "processor_class": "DonutProcessor",
16
  "sep_token": "</s>",
 
11
  "single_word": false
12
  },
13
  "model_max_length": 1000000000000000019884624838656,
14
+ "name_or_path": "./model_extended_2",
15
  "pad_token": "<pad>",
16
  "processor_class": "DonutProcessor",
17
  "sep_token": "</s>",