Upload ChunkFormer model

README.md

@@ -0,0 +1,82 @@
+---
+tags:
+- speech-recognition
+- audio
+- chunkformer
+- ctc
+- pytorch
+- transformers
+- automatic-speech-recognition
+- long-form transcription
+- asr
+license: apache-2.0
+library_name: transformers
+pipeline_tag: automatic-speech-recognition
+---
+
+# ChunkFormer Model
+<style>
+img {
+display: inline;
+}
+</style>
+[![GitHub](https://img.shields.io/badge/GitHub-ChunkFormer-blue)](https://github.com/khanld/chunkformer)
+[![Paper](https://img.shields.io/badge/Paper-ICASSP%202025-green)](https://arxiv.org/abs/2502.14673)
+
+
+## Usage
+
+Install the package:
+
+```bash
+pip install chunkformer
+```
+
+```python
+from chunkformer import ChunkFormerModel
+
+# Load the model
+model = ChunkFormerModel.from_pretrained("khanhld/chunkformer-ctc-small-libri-100h")
+
+# For long-form audio transcription
+transcription = model.endless_decode(
+    audio_path="path/to/your/audio.wav",
+    chunk_size=64,
+    left_context_size=128,
+    right_context_size=128,
+    return_timestamps=True
+)
+print(transcription)
+
+# For batch processing
+audio_files = ["audio1.wav", "audio2.wav", "audio3.wav"]
+transcriptions = model.batch_decode(
+    audio_paths=audio_files,
+    chunk_size=64,
+    left_context_size=128,
+    right_context_size=128
+)
+```
+
+## Training
+
+This model was trained using the ChunkFormer framework. For more details about the training process and to access the source code, please visit: https://github.com/khanld/chunkformer
+
+Paper: https://arxiv.org/abs/2502.14673
+
+## Citation
+
+If you use this work in your research, please cite:
+
+```bibtex
+@INPROCEEDINGS{10888640,
+    author={Le, Khanh and Ho, Tuan Vu and Tran, Dung and Chau, Duc Thanh},
+    booktitle={ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
+    title={ChunkFormer: Masked Chunking Conformer For Long-Form Speech Transcription},
+    year={2025},
+    volume={},
+    number={},
+    pages={1-5},
+    keywords={Scalability;Memory management;Graphics processing units;Signal processing;Performance gain;Hardware;Resource management;Speech processing;Standards;Context modeling;chunkformer;masked batch;long-form transcription},
+    doi={10.1109/ICASSP49660.2025.10888640}}
+```

config.yaml

@@ -0,0 +1,121 @@
+accum_grad: 2
+cmvn: global_cmvn
+cmvn_conf:
+  cmvn_file: data/train-clean-100/global_cmvn
+  is_json_cmvn: true
+ctc: ctc
+ctc_conf:
+  ctc_blank_id: 0
+dataset: asr
+dataset_conf:
+  batch_conf:
+    batch_type: dynamic
+    max_frames_in_batch: 120000
+    pad_feat: true
+  fbank_conf:
+    dither: 1.0
+    frame_length: 25
+    frame_shift: 10
+    num_mel_bins: 80
+  filter_conf:
+    max_length: 40960
+    min_length: 0
+    token_max_length: 400
+    token_min_length: 1
+  resample_conf:
+    resample_rate: 16000
+  shuffle: true
+  shuffle_conf:
+    shuffle_size: 1000
+  sort: false
+  sort_conf:
+    sort_size: 2000
+  spec_aug: true
+  spec_aug_conf:
+    max_f: 10
+    max_t: 50
+    num_f_mask: 2
+    num_t_mask: 2
+  spec_sub: false
+  spec_sub_conf:
+    max_t: 30
+    num_t_sub: 3
+  speed_perturb: true
+decoder: bitransformer
+decoder_conf:
+  attention_heads: 4
+  dropout_rate: 0.1
+  linear_units: 2048
+  num_blocks: 3
+  positional_dropout_rate: 0.1
+  r_num_blocks: 3
+  self_attention_dropout_rate: 0.1
+  src_attention_dropout_rate: 0.1
+dtype: fp16
+encoder: chunkformer
+encoder_conf:
+  activation_type: swish
+  attention_dropout_rate: 0.1
+  attention_heads: 4
+  cnn_module_kernel: 15
+  cnn_module_norm: layer_norm
+  dropout_rate: 0.1
+  dynamic_chunk_sizes:
+  - -1
+  - -1
+  - 64
+  - 128
+  - 256
+  dynamic_conv: true
+  dynamic_left_context_sizes:
+  - 64
+  - 128
+  - 256
+  dynamic_right_context_sizes:
+  - 64
+  - 128
+  - 256
+  input_layer: dw_striding
+  linear_units: 2048
+  normalize_before: true
+  num_blocks: 12
+  output_size: 256
+  pos_enc_layer_type: chunk_rel_pos
+  positional_dropout_rate: 0.1
+  selfattention_layer_type: chunk_rel_seflattn
+  use_cnn_module: true
+grad_clip: 5
+input_dim: 80
+log_interval: 100
+max_epoch: 200
+model: asr_model
+model_conf:
+  ctc_weight: 0.3
+  length_normalized_loss: false
+  lsm_weight: 0.1
+  reverse_weight: 0.3
+model_dir: exp/v1
+optim: adamw
+optim_conf:
+  lr: 0.001
+output_dim: 1024
+save_states: model_only
+scheduler: warmuplr
+scheduler_conf:
+  warmup_steps: 15000
+tokenizer: bpe
+tokenizer_conf:
+  bpe_path: data/lang_char/train-clean-100_bpe1024.model
+  is_multilingual: false
+  non_lang_syms_path: null
+  num_languages: 1
+  special_tokens:
+    <blank>: 0
+    <eos>: 2
+    <sos>: 2
+    <unk>: 1
+  split_with_space: false
+  symbol_table_path: data/lang_char/train-clean-100_bpe1024_units.txt
+train_engine: torch_ddp
+use_amp: true
+vocab_size: 1024

global_cmvn

@@ -0,0 +1 @@
+{"mean_stat": [377288640.0, 392184320.0, 415522112.0, 439642304.0, 462305440.0, 478134144.0, 486317792.0, 484854944.0, 481903008.0, 478719936.0, 484679808.0, 490927648.0, 501299424.0, 508010560.0, 509113632.0, 510138976.0, 509241344.0, 505738048.0, 507309600.0, 497438368.0, 489879040.0, 495238624.0, 486570304.0, 489920800.0, 486313792.0, 489624352.0, 486650080.0, 490423072.0, 490283296.0, 491844256.0, 493904384.0, 496209792.0, 499723104.0, 506249856.0, 511422688.0, 514613472.0, 519311744.0, 523896064.0, 528323104.0, 527253408.0, 531522400.0, 530094080.0, 532636160.0, 531560032.0, 532671776.0, 535229760.0, 538473664.0, 541965888.0, 544653120.0, 546927744.0, 547662464.0, 548900608.0, 550557184.0, 551221760.0, 550011072.0, 549681216.0, 549286208.0, 549093952.0, 549785856.0, 550029440.0, 550721856.0, 548635328.0, 544479872.0, 541057088.0, 537695744.0, 534047744.0, 530223392.0, 526621440.0, 524060832.0, 522416352.0, 520415168.0, 519538016.0, 519445088.0, 519127552.0, 517873984.0, 516610112.0, 516178496.0, 513642656.0, 511779712.0, 509221632.0], "var_stat": [4320987136.0, 4707190784.0, 5358135296.0, 5987740672.0, 6634504704.0, 7109028352.0, 7372617216.0, 7347464192.0, 7241876480.0, 7122445824.0, 7274906624.0, 7460743680.0, 7764865024.0, 7959791616.0, 7985910272.0, 8000324096.0, 7956391424.0, 7843857408.0, 7875313152.0, 7581451776.0, 7354802688.0, 7481633280.0, 7224129024.0, 7295001088.0, 7178827776.0, 7249619968.0, 7157318656.0, 7249249792.0, 7242912768.0, 7282761216.0, 7336015872.0, 7396921856.0, 7479434240.0, 7617581056.0, 7766127616.0, 7900237824.0, 8038410752.0, 8170933760.0, 8297018368.0, 8260835840.0, 8380481024.0, 8333346816.0, 8399360000.0, 8366544896.0, 8400977920.0, 8480865280.0, 8584512000.0, 8697213952.0, 8783428608.0, 8851557376.0, 8868983808.0, 8906819584.0, 8952121344.0, 8969757696.0, 8927692800.0, 8911980544.0, 8898642944.0, 8894412800.0, 8916986880.0, 8925707264.0, 8938296320.0, 8863333376.0, 8754246656.0, 8649031680.0, 8546321920.0, 8433665024.0, 8317259264.0, 8210675712.0, 8133715456.0, 8079525888.0, 8018883584.0, 7984903680.0, 7976119296.0, 7963761664.0, 7922351104.0, 7879020544.0, 7862078976.0, 7796150272.0, 7744995840.0, 7673272832.0], "frame_num": 36162482}

pytorch_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99f4ee7b772ddf3b5d3a01bbbfd6fa85dbb687f653813d7fe8171ac949afaa04
+size 183009932

vocab.txt

@@ -0,0 +1,1024 @@
+<blank> 0
+<unk> 1
+<sos/eos> 2
+' 3
+▁ 4
+▁a 5
+a 6
+▁ab 7
+ab 8
+able 9
+ably 10
+▁about 11
+▁ac 12
+ac 13
+▁acc 14
+ace 15
+ach 16
+ached 17
+ack 18
+▁act 19
+act 20
+▁ad 21
+ad 22
+▁add 23
+ade 24
+▁adm 25
+▁adv 26
+ady 27
+▁af 28
+▁after 29
+▁ag 30
+ag 31
+▁again 32
+▁against 33
+age 34
+ail 35
+ain 36
+ained 37
+aint 38
+▁air 39
+air 40
+ak 41
+ake 42
+aken 43
+aking 44
+▁al 45
+al 46
+ale 47
+alk 48
+▁all 49
+all 50
+ally 51
+▁almost 52
+▁along 53
+als 54
+▁also 55
+▁always 56
+▁am 57
+am 58
+ame 59
+▁among 60
+amp 61
+▁an 62
+an 63
+anc 64
+ance 65
+▁and 66
+and 67
+ane 68
+ang 69
+ange 70
+ank 71
+▁another 72
+ans 73
+▁answ 74
+ant 75
+▁any 76
+▁anything 77
+ap 78
+▁app 79
+app 80
+▁appe 81
+aps 82
+▁ar 83
+ar 84
+ard 85
+▁are 86
+are 87
+ared 88
+ark 89
+▁arm 90
+arm 91
+ars 92
+art 93
+ary 94
+▁as 95
+as 96
+ase 97
+ash 98
+▁asked 99
+ason 100
+▁ass 101
+ass 102
+ast 103
+▁at 104
+at 105
+atch 106
+ate 107
+ated 108
+ately 109
+ater 110
+ates 111
+ath 112
+ather 113
+ating 114
+ation 115
+ations 116
+▁att 117
+au 118
+aught 119
+ause 120
+aut 121
+av 122
+ave 123
+▁aw 124
+▁away 125
+ay 126
+▁b 127
+b 128
+▁back 129
+▁bar 130
+▁be 131
+be 132
+▁beaut 133
+▁bec 134
+▁because 135
+▁been 136
+▁before 137
+▁beg 138
+▁began 139
+▁beh 140
+▁being 141
+▁bel 142
+▁belie 143
+▁believe 144
+ber 145
+▁best 146
+▁bet 147
+▁better 148
+▁betw 149
+▁between 150
+▁bl 151
+▁black 152
+ble 153
+▁bo 154
+▁bod 155
+▁both 156
+▁br 157
+▁bre 158
+▁bro 159
+▁brought 160
+▁bu 161
+▁bus 162
+▁but 163
+▁by 164
+▁c 165
+c 166
+▁call 167
+▁called 168
+▁came 169
+▁can 170
+▁cap 171
+▁car 172
+▁care 173
+cc 174
+ce 175
+ced 176
+cei 177
+cept 178
+▁certain 179
+ces 180
+cess 181
+▁ch 182
+ch 183
+▁char 184
+▁che 185
+ched 186
+▁child 187
+ci 188
+ck 189
+▁cl 190
+cl 191
+▁cle 192
+▁co 193
+▁col 194
+▁com 195
+▁come 196
+▁comm 197
+▁comp 198
+▁compan 199
+▁con 200
+▁conf 201
+▁cons 202
+▁cont 203
+▁contin 204
+▁cor 205
+▁cou 206
+▁could 207
+▁count 208
+▁country 209
+▁cour 210
+▁course 211
+▁cr 212
+ct 213
+ction 214
+▁cur 215
+▁d 216
+d 217
+▁dark 218
+▁day 219
+▁days 220
+dd 221
+dden 222
+▁de 223
+de 224
+▁dear 225
+▁del 226
+▁des 227
+▁did 228
+▁dif 229
+▁dis 230
+▁disc 231
+▁dist 232
+▁do 233
+▁doct 234
+▁does 235
+▁don 236
+▁done 237
+▁door 238
+▁down 239
+▁dr 240
+▁dra 241
+▁dre 242
+▁e 243
+e 244
+▁each 245
+▁ear 246
+ect 247
+▁ed 248
+ed 249
+een 250
+▁eight 251
+▁el 252
+el 253
+▁em 254
+em 255
+▁en 256
+en 257
+ence 258
+▁end 259
+ened 260
+▁eng 261
+▁enough 262
+ens 263
+ense 264
+▁ent 265
+ent 266
+ention 267
+ently 268
+ents 269
+enty 270
+ep 271
+er 272
+ered 273
+ering 274
+ers 275
+ertain 276
+▁es 277
+es 278
+ess 279
+est 280
+et 281
+eter 282
+▁ev 283
+▁even 284
+▁ever 285
+ever 286
+▁every 287
+ew 288
+▁ex 289
+▁exp 290
+▁ey 291
+▁eyes 292
+▁f 293
+f 294
+▁fa 295
+▁face 296
+▁fact 297
+▁fall 298
+▁fam 299
+▁far 300
+▁father 301
+▁fe 302
+fe 303
+fect 304
+▁feel 305
+▁feet 306
+▁felt 307
+▁few 308
+ff 309
+▁fin 310
+▁find 311
+▁fir 312
+▁first 313
+▁five 314
+▁fl 315
+▁fo 316
+▁follow 317
+▁for 318
+▁fore 319
+fore 320
+▁form 321
+fort 322
+▁found 323
+▁four 324
+▁fr 325
+▁friend 326
+▁from 327
+ft 328
+ful 329
+▁full 330
+fully 331
+▁g 332
+g 333
+▁gave 334
+ge 335
+ged 336
+▁gen 337
+▁gener 338
+▁gent 339
+▁get 340
+get 341
+gether 342
+gg 343
+gh 344
+ght 345
+▁gi 346
+▁girl 347
+▁give 348
+▁gl 349
+▁go 350
+▁god 351
+▁going 352
+▁good 353
+▁got 354
+▁gr 355
+▁gra 356
+▁gre 357
+▁great 358
+▁gu 359
+▁h 360
+h 361
+▁ha 362
+▁had 363
+▁half 364
+▁hand 365
+▁hands 366
+▁happ 367
+haps 368
+▁hard 369
+▁has 370
+▁have 371
+▁having 372
+▁he 373
+he 374
+▁head 375
+▁hear 376
+▁heard 377
+▁heart 378
+hed 379
+▁hel 380
+▁her 381
+her 382
+▁here 383
+▁herself 384
+hes 385
+▁high 386
+▁him 387
+▁himself 388
+hing 389
+▁his 390
+hn 391
+▁ho 392
+▁home 393
+▁hope 394
+▁hor 395
+▁hour 396
+▁house 397
+▁how 398
+▁however 399
+▁hu 400
+▁hum 401
+▁hundred 402
+▁i 403
+i 404
+ia 405
+ial 406
+ian 407
+ib 408
+ible 409
+ic 410
+ical 411
+ice 412
+ich 413
+ick 414
+ict 415
+id 416
+▁ide 417
+ide 418
+ie 419
+ied 420
+iend 421
+ier 422
+ies 423
+iet 424
+▁if 425
+if 426
+ife 427
+ig 428
+igh 429
+ight 430
+ign 431
+il 432
+ild 433
+ile 434
+iled 435
+ill 436
+ily 437
+▁im 438
+im 439
+▁imm 440
+▁imp 441
+▁in 442
+in 443
+▁inc 444
+▁ind 445
+ind 446
+▁inde 447
+▁indeed 448
+ine 449
+ined 450
+iness 451
+▁inf 452
+ing 453
+ings 454
+ink 455
+▁ins 456
+ins 457
+▁inst 458
+▁int 459
+int 460
+▁inte 461
+▁inter 462
+▁into 463
+ion 464
+ions 465
+ious 466
+ip 467
+ir 468
+ire 469
+ired 470
+irl 471
+▁is 472
+is 473
+ise 474
+ised 475
+ish 476
+ished 477
+iss 478
+ist 479
+▁it 480
+it 481
+ite 482
+ited 483
+ith 484
+ither 485
+ities 486
+ition 487
+▁its 488
+its 489
+itt 490
+ittle 491
+ity 492
+iv 493
+ive 494
+iver 495
+ix 496
+▁j 497
+j 498
+ject 499
+▁jo 500
+▁john 501
+▁just 502
+▁k 503
+k 504
+▁ke 505
+ke 506
+ked 507
+▁keep 508
+▁kind 509
+▁king 510
+▁kn 511
+▁knew 512
+▁know 513
+ks 514
+▁l 515
+l 516
+▁la 517
+▁land 518
+land 519
+▁lar 520
+▁last 521
+▁lay 522
+ld 523
+▁le 524
+le 525
+led 526
+▁left 527
+les 528
+▁less 529
+less 530
+▁let 531
+lf 532
+▁li 533
+lic 534
+▁life 535
+▁light 536
+▁like 537
+ling 538
+▁little 539
+ll 540
+llow 541
+▁lo 542
+▁long 543
+▁loo 544
+▁look 545
+▁looked 546
+▁looking 547
+▁love 548
+lt 549
+ly 550
+▁m 551
+m 552
+▁ma 553
+▁made 554
+▁make 555
+▁man 556
+man 557
+▁many 558
+▁mar 559
+▁mas 560
+▁mat 561
+▁may 562
+▁me 563
+▁mean 564
+▁men 565
+ment 566
+▁mer 567
+▁met 568
+▁might 569
+▁mil 570
+▁min 571
+▁mind 572
+▁mis 573
+▁miss 574
+▁missus 575
+▁mister 576
+▁mo 577
+▁moment 578
+▁mon 579
+▁money 580
+▁more 581
+▁morn 582
+▁morning 583
+▁most 584
+most 585
+▁mother 586
+▁mount 587
+▁mu 588
+▁much 589
+▁must 590
+▁my 591
+▁myself 592
+▁n 593
+n 594
+▁name 595
+▁nat 596
+nd 597
+nded 598
+nder 599
+▁ne 600
+ne 601
+▁near 602
+ned 603
+▁need 604
+ner 605
+ness 606
+▁never 607
+▁new 608
+▁next 609
+ng 610
+▁night 611
+ning 612
+▁no 613
+▁nor 614
+▁not 615
+not 616
+▁nothing 617
+▁now 618
+nt 619
+▁o 620
+o 621
+▁ob 622
+▁occ 623
+ock 624
+od 625
+▁of 626
+▁off 627
+og 628
+▁oh 629
+ol 630
+▁old 631
+om 632
+ome 633
+omet 634
+▁on 635
+on 636
+▁once 637
+ond 638
+▁one 639
+one 640
+oney 641
+ong 642
+▁only 643
+ons 644
+ont 645
+oo 646
+ood 647
+ook 648
+ool 649
+▁op 650
+op 651
+▁open 652
+ople 653
+▁or 654
+or 655
+▁ord 656
+ord 657
+ore 658
+orm 659
+orn 660
+ors 661
+ort 662
+ory 663
+os 664
+ose 665
+osed 666
+oss 667
+ost 668
+ot 669
+oth 670
+▁other 671
+other 672
+▁others 673
+ou 674
+oub 675
+oud 676
+ough 677
+ought 678
+ould 679
+ound 680
+ount 681
+▁our 682
+our 683
+ous 684
+ouse 685
+▁out 686
+out 687
+ove 688
+▁over 689
+ow 690
+owed 691
+ower 692
+▁own 693
+own 694
+ows 695
+oy 696
+▁p 697
+p 698
+▁par 699
+▁part 700
+▁pass 701
+▁passed 702
+▁pat 703
+▁pe 704
+pe 705
+pect 706
+▁people 707
+▁per 708
+per 709
+▁perhaps 710
+▁pers 711
+▁person 712
+▁ph 713
+ph 714
+▁pl 715
+pl 716
+▁place 717
+▁play 718
+▁ple 719
+ple 720
+▁pleas 721
+▁po 722
+▁point 723
+▁poor 724
+▁poss 725
+▁power 726
+pp 727
+pped 728
+▁pr 729
+▁pre 730
+▁pres 731
+▁present 732
+▁prin 733
+▁pro 734
+ps 735
+pt 736
+▁pur 737
+▁put 738
+q 739
+▁qu 740
+qu 741
+▁quest 742
+▁quite 743
+▁r 744
+r 745
+ra 746
+▁rather 747
+▁re 748
+re 749
+▁read 750
+reat 751
+▁rec 752
+▁red 753
+red 754
+ree 755
+▁ref 756
+▁reg 757
+▁rel 758
+▁rem 759
+ren 760
+rent 761
+▁rep 762
+▁repl 763
+▁replied 764
+▁res 765
+res 766
+ress 767
+ressed 768
+▁rest 769
+rest 770
+▁ret 771
+ret 772
+▁return 773
+ri 774
+ried 775
+▁right 776
+▁ro 777
+ro 778
+▁room 779
+ross 780
+▁round 781
+row 782
+ru 783
+ry 784
+▁s 785
+s 786
+▁sa 787
+▁said 788
+▁same 789
+▁sat 790
+▁saw 791
+▁say 792
+▁sc 793
+▁se 794
+se 795
+▁sec 796
+▁see 797
+▁seem 798
+▁seemed 799
+▁seen 800
+sel 801
+self 802
+selves 803
+▁sent 804
+▁ser 805
+▁serv 806
+▁set 807
+▁sh 808
+▁shall 809
+▁she 810
+▁short 811
+▁should 812
+▁side 813
+▁sil 814
+▁sim 815
+▁sin 816
+▁since 817
+▁sir 818
+▁six 819
+▁sk 820
+▁sl 821
+▁sle 822
+▁sm 823
+▁small 824
+▁sn 825
+▁so 826
+so 827
+▁some 828
+▁somet 829
+▁something 830
+▁soon 831
+▁sou 832
+▁sp 833
+▁spe 834
+ss 835
+▁st 836
+st 837
+▁stand 838
+▁start 839
+▁ste 840
+▁still 841
+▁stood 842
+▁str 843
+▁stre 844
+▁su 845
+▁sub 846
+▁such 847
+▁sudden 848
+▁sun 849
+▁supp 850
+▁sur 851
+▁sure 852
+▁sw 853
+sw 854
+▁sy 855
+▁t 856
+t 857
+tain 858
+▁take 859
+▁te 860
+te 861
+ted 862
+▁tell 863
+ten 864
+▁ter 865
+ter 866
+ters 867
+▁th 868
+th 869
+▁than 870
+▁that 871
+▁the 872
+▁their 873
+▁them 874
+▁then 875
+ther 876
+▁there 877
+▁these 878
+▁they 879
+▁thing 880
+thing 881
+▁things 882
+▁think 883
+▁thir 884
+▁this 885
+▁those 886
+▁though 887
+▁thought 888
+▁thr 889
+▁three 890
+▁through 891
+▁tim 892
+▁time 893
+ting 894
+▁to 895
+▁together 896
+▁told 897
+ton 898
+▁too 899
+▁took 900
+▁toward 901
+▁tr 902
+▁tra 903
+▁tre 904
+▁tri 905
+▁turn 906
+▁turned 907
+▁tw 908
+▁twenty 909
+▁two 910
+ty 911
+▁u 912
+u 913
+ual 914
+ub 915
+uck 916
+ud 917
+ue 918
+ued 919
+ul 920
+ular 921
+ull 922
+ully 923
+ult 924
+um 925
+▁un 926
+un 927
+▁unc 928
+▁under 929
+▁underst 930
+ung 931
+▁unt 932
+▁until 933
+▁up 934
+up 935
+▁upon 936
+ur 937
+ure 938
+ured 939
+ures 940
+uring 941
+urn 942
+▁us 943
+us 944
+▁use 945
+used 946
+ust 947
+ut 948
+ute 949
+uth 950
+▁v 951
+v 952
+ve 953
+ved 954
+vel 955
+ven 956
+vent 957
+ver 958
+▁very 959
+very 960
+ves 961
+ving 962
+▁vis 963
+▁vo 964
+▁voice 965
+▁w 966
+w 967
+▁wa 968
+▁want 969
+▁war 970
+ward 971
+▁was 972
+▁water 973
+▁way 974
+way 975
+ways 976
+▁we 977
+▁well 978
+▁went 979
+▁were 980
+▁wh 981
+▁what 982
+▁whe 983
+▁when 984
+▁where 985
+▁which 986
+▁while 987
+▁white 988
+▁who 989
+▁whole 990
+▁whom 991
+▁why 992
+▁will 993
+▁wind 994
+▁with 995
+▁without 996
+wn 997
+▁wom 998
+▁woman 999
+▁wond 1000
+▁wood 1001
+▁wor 1002
+▁word 1003
+▁words 1004
+▁work 1005
+▁world 1006
+▁would 1007
+▁wr 1008
+x 1009
+xt 1010
+▁y 1011
+y 1012
+▁ye 1013
+▁years 1014
+▁yes 1015
+▁yet 1016
+ying 1017
+▁you 1018
+▁young 1019
+▁your 1020
+ys 1021
+z 1022
+zz 1023