Hugging Face's logo

Datasets:
google
/
WaxalNLP

Tasks:
Automatic Speech Recognition
Text-to-Speech
Modalities:
Audio
Text
Formats:
parquet
Languages:
Acoli
Akan
Amharic
Size:
1M - 10M
ArXiv:
Tags:
audio
automatic-speech-recognition
text-to-speech
Libraries:
Datasets
Dask
Polars
License:
Dataset card Data Studio Files
xet
Community
6
perrynelson commited on
Commit
17894c6
·
verified ·
1 Parent(s): 7ad8eb1

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +528 -16
README.md CHANGED
@@ -43,26 +43,228 @@ source_datasets:
43
  - DigitalUmuganda/AfriVoice
44
  - original
45
  configs:
46
- - config_name: asr
47
  data_files:
48
  - split: train
49
- path: "data/ASR/**/*-train-*"
50
  - split: validation
51
- path: "data/ASR/**/*-validation-*"
52
  - split: test
53
- path: "data/ASR/**/*-test-*"
54
  - split: unlabeled
55
- path: "data/ASR/**/*-unlabeled-*"
56
- - config_name: tts
57
  data_files:
58
  - split: train
59
- path: "data/TTS/**/*-train-*"
60
  - split: validation
61
- path: "data/TTS/**/*-validation-*"
62
  - split: test
63
- path: "data/TTS/**/*-test-*"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
64
  dataset_info:
65
- - config_name: asr
66
  features:
67
  - name: id
68
  dtype: string
@@ -76,7 +278,315 @@ dataset_info:
76
  dtype: string
77
  - name: audio
78
  dtype: audio
79
- - config_name: tts
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
80
  features:
81
  - name: id
82
  dtype: string
@@ -108,7 +618,7 @@ dataset_info:
108
  - [Considerations for Using the Data](#considerations-for-using-the-data)
109
  - [Additional Information](#additional-information)
110
 
111
- ## Description
112
 
113
  The Waxal project provides datasets for both Automated Speech Recognition (ASR)
114
  and Text-to-Speech (TTS) for African languages. The goal of this dataset's
@@ -160,13 +670,14 @@ pip install datasets[audio]
160
 
161
  **Loading ASR Data**
162
 
163
- To load ASR data, point to the `data/ASR` directory.
 
164
 
165
  ```python
166
  from datasets import load_dataset, Audio
167
 
168
  # Load Shona (sna) ASR dataset
169
- asr_data = load_dataset("google/WaxalNLP", "sna", data_dir="data/ASR")
170
 
171
  # Access splits
172
  train = asr_data['train']
@@ -183,13 +694,14 @@ print(f"Audio Array Shape: {example['audio']['array'].shape}")
183
 
184
  **Loading TTS Data**
185
 
186
- To load TTS data, point to the `data/TTS` directory.
 
187
 
188
  ```python
189
  from datasets import load_dataset
190
 
191
  # Load Swahili (swa) TTS dataset
192
- tts_data = load_dataset("google/WaxalNLP", "swa", data_dir="data/TTS")
193
 
194
  # Access splits
195
  train = tts_data['train']
 
43
  - DigitalUmuganda/AfriVoice
44
  - original
45
  configs:
46
+ - config_name: ach_asr
47
  data_files:
48
  - split: train
49
+ path: "data/ASR/ach/ach-train-*"
50
  - split: validation
51
+ path: "data/ASR/ach/ach-validation-*"
52
  - split: test
53
+ path: "data/ASR/ach/ach-test-*"
54
  - split: unlabeled
55
+ path: "data/ASR/ach/ach-unlabeled-*"
56
+ - config_name: ach_tts
57
  data_files:
58
  - split: train
59
+ path: "data/TTS/ach/ach-train-*"
60
  - split: validation
61
+ path: "data/TTS/ach/ach-validation-*"
62
  - split: test
63
+ path: "data/TTS/ach/ach-test-*"
64
+ - config_name: aka_asr
65
+ data_files:
66
+ - split: train
67
+ path: "data/ASR/aka/aka-train-*"
68
+ - split: validation
69
+ path: "data/ASR/aka/aka-validation-*"
70
+ - split: test
71
+ path: "data/ASR/aka/aka-test-*"
72
+ - split: unlabeled
73
+ path: "data/ASR/aka/aka-unlabeled-*"
74
+ - config_name: dag_asr
75
+ data_files:
76
+ - split: train
77
+ path: "data/ASR/dag/dag-train-*"
78
+ - split: validation
79
+ path: "data/ASR/dag/dag-validation-*"
80
+ - split: test
81
+ path: "data/ASR/dag/dag-test-*"
82
+ - split: unlabeled
83
+ path: "data/ASR/dag/dag-unlabeled-*"
84
+ - config_name: dga_asr
85
+ data_files:
86
+ - split: train
87
+ path: "data/ASR/dga/dga-train-*"
88
+ - split: validation
89
+ path: "data/ASR/dga/dga-validation-*"
90
+ - split: test
91
+ path: "data/ASR/dga/dga-test-*"
92
+ - split: unlabeled
93
+ path: "data/ASR/dga/dga-unlabeled-*"
94
+ - config_name: ewe_asr
95
+ data_files:
96
+ - split: train
97
+ path: "data/ASR/ewe/ewe-train-*"
98
+ - split: validation
99
+ path: "data/ASR/ewe/ewe-validation-*"
100
+ - split: test
101
+ path: "data/ASR/ewe/ewe-test-*"
102
+ - split: unlabeled
103
+ path: "data/ASR/ewe/ewe-unlabeled-*"
104
+ - config_name: fat_tts
105
+ data_files:
106
+ - split: train
107
+ path: "data/TTS/fat/fat-train-*"
108
+ - split: validation
109
+ path: "data/TTS/fat/fat-validation-*"
110
+ - split: test
111
+ path: "data/TTS/fat/fat-test-*"
112
+ - config_name: ful_asr
113
+ data_files:
114
+ - split: train
115
+ path: "data/ASR/ful/ful-train-*"
116
+ - split: validation
117
+ path: "data/ASR/ful/ful-validation-*"
118
+ - split: test
119
+ path: "data/ASR/ful/ful-test-*"
120
+ - split: unlabeled
121
+ path: "data/ASR/ful/ful-unlabeled-*"
122
+ - config_name: ful_tts
123
+ data_files:
124
+ - split: train
125
+ path: "data/TTS/ful/ful-train-*"
126
+ - split: validation
127
+ path: "data/TTS/ful/ful-validation-*"
128
+ - split: test
129
+ path: "data/TTS/ful/ful-test-*"
130
+ - config_name: hau_tts
131
+ data_files:
132
+ - split: train
133
+ path: "data/TTS/hau/hau-train-*"
134
+ - split: validation
135
+ path: "data/TTS/hau/hau-validation-*"
136
+ - split: test
137
+ path: "data/TTS/hau/hau-test-*"
138
+ - config_name: ibo_tts
139
+ data_files:
140
+ - split: train
141
+ path: "data/TTS/ibo/ibo-train-*"
142
+ - split: validation
143
+ path: "data/TTS/ibo/ibo-validation-*"
144
+ - split: test
145
+ path: "data/TTS/ibo/ibo-test-*"
146
+ - config_name: kpo_asr
147
+ data_files:
148
+ - split: train
149
+ path: "data/ASR/kpo/kpo-train-*"
150
+ - split: validation
151
+ path: "data/ASR/kpo/kpo-validation-*"
152
+ - split: test
153
+ path: "data/ASR/kpo/kpo-test-*"
154
+ - split: unlabeled
155
+ path: "data/ASR/kpo/kpo-unlabeled-*"
156
+ - config_name: lin_asr
157
+ data_files:
158
+ - split: train
159
+ path: "data/ASR/lin/lin-train-*"
160
+ - split: validation
161
+ path: "data/ASR/lin/lin-validation-*"
162
+ - split: test
163
+ path: "data/ASR/lin/lin-test-*"
164
+ - split: unlabeled
165
+ path: "data/ASR/lin/lin-unlabeled-*"
166
+ - config_name: lug_asr
167
+ data_files:
168
+ - split: train
169
+ path: "data/ASR/lug/lug-train-*"
170
+ - split: validation
171
+ path: "data/ASR/lug/lug-validation-*"
172
+ - split: test
173
+ path: "data/ASR/lug/lug-test-*"
174
+ - split: unlabeled
175
+ path: "data/ASR/lug/lug-unlabeled-*"
176
+ - config_name: lug_tts
177
+ data_files:
178
+ - split: train
179
+ path: "data/TTS/lug/lug-train-*"
180
+ - split: validation
181
+ path: "data/TTS/lug/lug-validation-*"
182
+ - split: test
183
+ path: "data/TTS/lug/lug-test-*"
184
+ - config_name: mas_asr
185
+ data_files:
186
+ - split: train
187
+ path: "data/ASR/mas/mas-train-*"
188
+ - split: validation
189
+ path: "data/ASR/mas/mas-validation-*"
190
+ - split: test
191
+ path: "data/ASR/mas/mas-test-*"
192
+ - split: unlabeled
193
+ path: "data/ASR/mas/mas-unlabeled-*"
194
+ - config_name: mlg_asr
195
+ data_files:
196
+ - split: train
197
+ path: "data/ASR/mlg/mlg-train-*"
198
+ - split: validation
199
+ path: "data/ASR/mlg/mlg-validation-*"
200
+ - split: test
201
+ path: "data/ASR/mlg/mlg-test-*"
202
+ - split: unlabeled
203
+ path: "data/ASR/mlg/mlg-unlabeled-*"
204
+ - config_name: nyn_asr
205
+ data_files:
206
+ - split: train
207
+ path: "data/ASR/nyn/nyn-train-*"
208
+ - split: validation
209
+ path: "data/ASR/nyn/nyn-validation-*"
210
+ - split: test
211
+ path: "data/ASR/nyn/nyn-test-*"
212
+ - split: unlabeled
213
+ path: "data/ASR/nyn/nyn-unlabeled-*"
214
+ - config_name: nyn_tts
215
+ data_files:
216
+ - split: train
217
+ path: "data/TTS/nyn/nyn-train-*"
218
+ - split: validation
219
+ path: "data/TTS/nyn/nyn-validation-*"
220
+ - split: test
221
+ path: "data/TTS/nyn/nyn-test-*"
222
+ - config_name: sna_asr
223
+ data_files:
224
+ - split: train
225
+ path: "data/ASR/sna/sna-train-*"
226
+ - split: validation
227
+ path: "data/ASR/sna/sna-validation-*"
228
+ - split: test
229
+ path: "data/ASR/sna/sna-test-*"
230
+ - split: unlabeled
231
+ path: "data/ASR/sna/sna-unlabeled-*"
232
+ - config_name: sog_asr
233
+ data_files:
234
+ - split: train
235
+ path: "data/ASR/sog/sog-train-*"
236
+ - split: validation
237
+ path: "data/ASR/sog/sog-validation-*"
238
+ - split: test
239
+ path: "data/ASR/sog/sog-test-*"
240
+ - split: unlabeled
241
+ path: "data/ASR/sog/sog-unlabeled-*"
242
+ - config_name: swa_tts
243
+ data_files:
244
+ - split: train
245
+ path: "data/TTS/swa/swa-train-*"
246
+ - split: validation
247
+ path: "data/TTS/swa/swa-validation-*"
248
+ - split: test
249
+ path: "data/TTS/swa/swa-test-*"
250
+ - config_name: twi_tts
251
+ data_files:
252
+ - split: train
253
+ path: "data/TTS/twi/twi-train-*"
254
+ - split: validation
255
+ path: "data/TTS/twi/twi-validation-*"
256
+ - split: test
257
+ path: "data/TTS/twi/twi-test-*"
258
+ - config_name: yor_tts
259
+ data_files:
260
+ - split: train
261
+ path: "data/TTS/yor/yor-train-*"
262
+ - split: validation
263
+ path: "data/TTS/yor/yor-validation-*"
264
+ - split: test
265
+ path: "data/TTS/yor/yor-test-*"
266
  dataset_info:
267
+ - config_name: ach_asr
268
  features:
269
  - name: id
270
  dtype: string
 
278
  dtype: string
279
  - name: audio
280
  dtype: audio
281
+ - config_name: ach_tts
282
+ features:
283
+ - name: id
284
+ dtype: string
285
+ - name: speaker_id
286
+ dtype: string
287
+ - name: transcription
288
+ dtype: string
289
+ - name: locale
290
+ dtype: string
291
+ - name: gender
292
+ dtype: string
293
+ - name: audio
294
+ dtype: audio
295
+ - config_name: aka_asr
296
+ features:
297
+ - name: id
298
+ dtype: string
299
+ - name: speaker_id
300
+ dtype: string
301
+ - name: transcription
302
+ dtype: string
303
+ - name: language
304
+ dtype: string
305
+ - name: gender
306
+ dtype: string
307
+ - name: audio
308
+ dtype: audio
309
+ - config_name: dag_asr
310
+ features:
311
+ - name: id
312
+ dtype: string
313
+ - name: speaker_id
314
+ dtype: string
315
+ - name: transcription
316
+ dtype: string
317
+ - name: language
318
+ dtype: string
319
+ - name: gender
320
+ dtype: string
321
+ - name: audio
322
+ dtype: audio
323
+ - config_name: dga_asr
324
+ features:
325
+ - name: id
326
+ dtype: string
327
+ - name: speaker_id
328
+ dtype: string
329
+ - name: transcription
330
+ dtype: string
331
+ - name: language
332
+ dtype: string
333
+ - name: gender
334
+ dtype: string
335
+ - name: audio
336
+ dtype: audio
337
+ - config_name: ewe_asr
338
+ features:
339
+ - name: id
340
+ dtype: string
341
+ - name: speaker_id
342
+ dtype: string
343
+ - name: transcription
344
+ dtype: string
345
+ - name: language
346
+ dtype: string
347
+ - name: gender
348
+ dtype: string
349
+ - name: audio
350
+ dtype: audio
351
+ - config_name: fat_tts
352
+ features:
353
+ - name: id
354
+ dtype: string
355
+ - name: speaker_id
356
+ dtype: string
357
+ - name: transcription
358
+ dtype: string
359
+ - name: locale
360
+ dtype: string
361
+ - name: gender
362
+ dtype: string
363
+ - name: audio
364
+ dtype: audio
365
+ - config_name: ful_asr
366
+ features:
367
+ - name: id
368
+ dtype: string
369
+ - name: speaker_id
370
+ dtype: string
371
+ - name: transcription
372
+ dtype: string
373
+ - name: language
374
+ dtype: string
375
+ - name: gender
376
+ dtype: string
377
+ - name: audio
378
+ dtype: audio
379
+ - config_name: ful_tts
380
+ features:
381
+ - name: id
382
+ dtype: string
383
+ - name: speaker_id
384
+ dtype: string
385
+ - name: transcription
386
+ dtype: string
387
+ - name: locale
388
+ dtype: string
389
+ - name: gender
390
+ dtype: string
391
+ - name: audio
392
+ dtype: audio
393
+ - config_name: hau_tts
394
+ features:
395
+ - name: id
396
+ dtype: string
397
+ - name: speaker_id
398
+ dtype: string
399
+ - name: transcription
400
+ dtype: string
401
+ - name: locale
402
+ dtype: string
403
+ - name: gender
404
+ dtype: string
405
+ - name: audio
406
+ dtype: audio
407
+ - config_name: ibo_tts
408
+ features:
409
+ - name: id
410
+ dtype: string
411
+ - name: speaker_id
412
+ dtype: string
413
+ - name: transcription
414
+ dtype: string
415
+ - name: locale
416
+ dtype: string
417
+ - name: gender
418
+ dtype: string
419
+ - name: audio
420
+ dtype: audio
421
+ - config_name: kpo_asr
422
+ features:
423
+ - name: id
424
+ dtype: string
425
+ - name: speaker_id
426
+ dtype: string
427
+ - name: transcription
428
+ dtype: string
429
+ - name: language
430
+ dtype: string
431
+ - name: gender
432
+ dtype: string
433
+ - name: audio
434
+ dtype: audio
435
+ - config_name: lin_asr
436
+ features:
437
+ - name: id
438
+ dtype: string
439
+ - name: speaker_id
440
+ dtype: string
441
+ - name: transcription
442
+ dtype: string
443
+ - name: language
444
+ dtype: string
445
+ - name: gender
446
+ dtype: string
447
+ - name: audio
448
+ dtype: audio
449
+ - config_name: lug_asr
450
+ features:
451
+ - name: id
452
+ dtype: string
453
+ - name: speaker_id
454
+ dtype: string
455
+ - name: transcription
456
+ dtype: string
457
+ - name: language
458
+ dtype: string
459
+ - name: gender
460
+ dtype: string
461
+ - name: audio
462
+ dtype: audio
463
+ - config_name: lug_tts
464
+ features:
465
+ - name: id
466
+ dtype: string
467
+ - name: speaker_id
468
+ dtype: string
469
+ - name: transcription
470
+ dtype: string
471
+ - name: locale
472
+ dtype: string
473
+ - name: gender
474
+ dtype: string
475
+ - name: audio
476
+ dtype: audio
477
+ - config_name: mas_asr
478
+ features:
479
+ - name: id
480
+ dtype: string
481
+ - name: speaker_id
482
+ dtype: string
483
+ - name: transcription
484
+ dtype: string
485
+ - name: language
486
+ dtype: string
487
+ - name: gender
488
+ dtype: string
489
+ - name: audio
490
+ dtype: audio
491
+ - config_name: mlg_asr
492
+ features:
493
+ - name: id
494
+ dtype: string
495
+ - name: speaker_id
496
+ dtype: string
497
+ - name: transcription
498
+ dtype: string
499
+ - name: language
500
+ dtype: string
501
+ - name: gender
502
+ dtype: string
503
+ - name: audio
504
+ dtype: audio
505
+ - config_name: nyn_asr
506
+ features:
507
+ - name: id
508
+ dtype: string
509
+ - name: speaker_id
510
+ dtype: string
511
+ - name: transcription
512
+ dtype: string
513
+ - name: language
514
+ dtype: string
515
+ - name: gender
516
+ dtype: string
517
+ - name: audio
518
+ dtype: audio
519
+ - config_name: nyn_tts
520
+ features:
521
+ - name: id
522
+ dtype: string
523
+ - name: speaker_id
524
+ dtype: string
525
+ - name: transcription
526
+ dtype: string
527
+ - name: locale
528
+ dtype: string
529
+ - name: gender
530
+ dtype: string
531
+ - name: audio
532
+ dtype: audio
533
+ - config_name: sna_asr
534
+ features:
535
+ - name: id
536
+ dtype: string
537
+ - name: speaker_id
538
+ dtype: string
539
+ - name: transcription
540
+ dtype: string
541
+ - name: language
542
+ dtype: string
543
+ - name: gender
544
+ dtype: string
545
+ - name: audio
546
+ dtype: audio
547
+ - config_name: sog_asr
548
+ features:
549
+ - name: id
550
+ dtype: string
551
+ - name: speaker_id
552
+ dtype: string
553
+ - name: transcription
554
+ dtype: string
555
+ - name: language
556
+ dtype: string
557
+ - name: gender
558
+ dtype: string
559
+ - name: audio
560
+ dtype: audio
561
+ - config_name: swa_tts
562
+ features:
563
+ - name: id
564
+ dtype: string
565
+ - name: speaker_id
566
+ dtype: string
567
+ - name: transcription
568
+ dtype: string
569
+ - name: locale
570
+ dtype: string
571
+ - name: gender
572
+ dtype: string
573
+ - name: audio
574
+ dtype: audio
575
+ - config_name: twi_tts
576
+ features:
577
+ - name: id
578
+ dtype: string
579
+ - name: speaker_id
580
+ dtype: string
581
+ - name: transcription
582
+ dtype: string
583
+ - name: locale
584
+ dtype: string
585
+ - name: gender
586
+ dtype: string
587
+ - name: audio
588
+ dtype: audio
589
+ - config_name: yor_tts
590
  features:
591
  - name: id
592
  dtype: string
 
618
  - [Considerations for Using the Data](#considerations-for-using-the-data)
619
  - [Additional Information](#additional-information)
620
 
621
+ ## Dataset Description
622
 
623
  The Waxal project provides datasets for both Automated Speech Recognition (ASR)
624
  and Text-to-Speech (TTS) for African languages. The goal of this dataset's
 
670
 
671
  **Loading ASR Data**
672
 
673
+ To load ASR data for a specific language, specify the configuration name, e.g.
674
+ `sna_asr` for Shona ASR data.
675
 
676
  ```python
677
  from datasets import load_dataset, Audio
678
 
679
  # Load Shona (sna) ASR dataset
680
+ asr_data = load_dataset("google/WaxalNLP", "sna_asr")
681
 
682
  # Access splits
683
  train = asr_data['train']
 
694
 
695
  **Loading TTS Data**
696
 
697
+ To load TTS data for a specific language, specify the configuration name, e.g.
698
+ `swa_tts` for Swahili TTS data.
699
 
700
  ```python
701
  from datasets import load_dataset
702
 
703
  # Load Swahili (swa) TTS dataset
704
+ tts_data = load_dataset("google/WaxalNLP", "swa_tts")
705
 
706
  # Access splits
707
  train = tts_data['train']