lhoestq HF staff commited on
Commit
53d4fb1
1 Parent(s): cdad49d

add dataset_info in dataset metadata

Browse files
Files changed (1) hide show
  1. README.md +1388 -2
README.md CHANGED
@@ -21,10 +21,1396 @@ annotations_creators:
21
  multilinguality:
22
  - multilingual
23
  task_categories:
24
- - question-answering
25
  task_ids:
26
  - extractive-qa
27
  paperswithcode_id: mlqa
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
  ---
29
 
30
  # Dataset Card for "mlqa"
@@ -261,4 +1647,4 @@ The data fields are the same among all splits.
261
 
262
  ### Contributions
263
 
264
- Thanks to [@patrickvonplaten](https://github.com/patrickvonplaten), [@M-Salti](https://github.com/M-Salti), [@lewtun](https://github.com/lewtun), [@thomwolf](https://github.com/thomwolf), [@mariamabarham](https://github.com/mariamabarham), [@lhoestq](https://github.com/lhoestq) for adding this dataset.
 
21
  multilinguality:
22
  - multilingual
23
  task_categories:
24
+ - question-answering
25
  task_ids:
26
  - extractive-qa
27
  paperswithcode_id: mlqa
28
+ dataset_info:
29
+ - config_name: mlqa-translate-train.ar
30
+ features:
31
+ - name: context
32
+ dtype: string
33
+ - name: question
34
+ dtype: string
35
+ - name: answers
36
+ sequence:
37
+ - name: answer_start
38
+ dtype: int32
39
+ - name: text
40
+ dtype: string
41
+ - name: id
42
+ dtype: string
43
+ splits:
44
+ - name: train
45
+ num_bytes: 101227245
46
+ num_examples: 78058
47
+ - name: validation
48
+ num_bytes: 13144332
49
+ num_examples: 9512
50
+ download_size: 63364123
51
+ dataset_size: 114371577
52
+ - config_name: mlqa-translate-train.de
53
+ features:
54
+ - name: context
55
+ dtype: string
56
+ - name: question
57
+ dtype: string
58
+ - name: answers
59
+ sequence:
60
+ - name: answer_start
61
+ dtype: int32
62
+ - name: text
63
+ dtype: string
64
+ - name: id
65
+ dtype: string
66
+ splits:
67
+ - name: train
68
+ num_bytes: 77996825
69
+ num_examples: 80069
70
+ - name: validation
71
+ num_bytes: 10322113
72
+ num_examples: 9927
73
+ download_size: 63364123
74
+ dataset_size: 88318938
75
+ - config_name: mlqa-translate-train.vi
76
+ features:
77
+ - name: context
78
+ dtype: string
79
+ - name: question
80
+ dtype: string
81
+ - name: answers
82
+ sequence:
83
+ - name: answer_start
84
+ dtype: int32
85
+ - name: text
86
+ dtype: string
87
+ - name: id
88
+ dtype: string
89
+ splits:
90
+ - name: train
91
+ num_bytes: 97387431
92
+ num_examples: 84816
93
+ - name: validation
94
+ num_bytes: 12731112
95
+ num_examples: 10356
96
+ download_size: 63364123
97
+ dataset_size: 110118543
98
+ - config_name: mlqa-translate-train.zh
99
+ features:
100
+ - name: context
101
+ dtype: string
102
+ - name: question
103
+ dtype: string
104
+ - name: answers
105
+ sequence:
106
+ - name: answer_start
107
+ dtype: int32
108
+ - name: text
109
+ dtype: string
110
+ - name: id
111
+ dtype: string
112
+ splits:
113
+ - name: train
114
+ num_bytes: 55143547
115
+ num_examples: 76285
116
+ - name: validation
117
+ num_bytes: 7418070
118
+ num_examples: 9568
119
+ download_size: 63364123
120
+ dataset_size: 62561617
121
+ - config_name: mlqa-translate-train.es
122
+ features:
123
+ - name: context
124
+ dtype: string
125
+ - name: question
126
+ dtype: string
127
+ - name: answers
128
+ sequence:
129
+ - name: answer_start
130
+ dtype: int32
131
+ - name: text
132
+ dtype: string
133
+ - name: id
134
+ dtype: string
135
+ splits:
136
+ - name: train
137
+ num_bytes: 80789653
138
+ num_examples: 81810
139
+ - name: validation
140
+ num_bytes: 10718376
141
+ num_examples: 10123
142
+ download_size: 63364123
143
+ dataset_size: 91508029
144
+ - config_name: mlqa-translate-train.hi
145
+ features:
146
+ - name: context
147
+ dtype: string
148
+ - name: question
149
+ dtype: string
150
+ - name: answers
151
+ sequence:
152
+ - name: answer_start
153
+ dtype: int32
154
+ - name: text
155
+ dtype: string
156
+ - name: id
157
+ dtype: string
158
+ splits:
159
+ - name: train
160
+ num_bytes: 168117671
161
+ num_examples: 82451
162
+ - name: validation
163
+ num_bytes: 22422152
164
+ num_examples: 10253
165
+ download_size: 63364123
166
+ dataset_size: 190539823
167
+ - config_name: mlqa-translate-test.ar
168
+ features:
169
+ - name: context
170
+ dtype: string
171
+ - name: question
172
+ dtype: string
173
+ - name: answers
174
+ sequence:
175
+ - name: answer_start
176
+ dtype: int32
177
+ - name: text
178
+ dtype: string
179
+ - name: id
180
+ dtype: string
181
+ splits:
182
+ - name: test
183
+ num_bytes: 5484467
184
+ num_examples: 5335
185
+ download_size: 10075488
186
+ dataset_size: 5484467
187
+ - config_name: mlqa-translate-test.de
188
+ features:
189
+ - name: context
190
+ dtype: string
191
+ - name: question
192
+ dtype: string
193
+ - name: answers
194
+ sequence:
195
+ - name: answer_start
196
+ dtype: int32
197
+ - name: text
198
+ dtype: string
199
+ - name: id
200
+ dtype: string
201
+ splits:
202
+ - name: test
203
+ num_bytes: 3884332
204
+ num_examples: 4517
205
+ download_size: 10075488
206
+ dataset_size: 3884332
207
+ - config_name: mlqa-translate-test.vi
208
+ features:
209
+ - name: context
210
+ dtype: string
211
+ - name: question
212
+ dtype: string
213
+ - name: answers
214
+ sequence:
215
+ - name: answer_start
216
+ dtype: int32
217
+ - name: text
218
+ dtype: string
219
+ - name: id
220
+ dtype: string
221
+ splits:
222
+ - name: test
223
+ num_bytes: 5998327
224
+ num_examples: 5495
225
+ download_size: 10075488
226
+ dataset_size: 5998327
227
+ - config_name: mlqa-translate-test.zh
228
+ features:
229
+ - name: context
230
+ dtype: string
231
+ - name: question
232
+ dtype: string
233
+ - name: answers
234
+ sequence:
235
+ - name: answer_start
236
+ dtype: int32
237
+ - name: text
238
+ dtype: string
239
+ - name: id
240
+ dtype: string
241
+ splits:
242
+ - name: test
243
+ num_bytes: 4831704
244
+ num_examples: 5137
245
+ download_size: 10075488
246
+ dataset_size: 4831704
247
+ - config_name: mlqa-translate-test.es
248
+ features:
249
+ - name: context
250
+ dtype: string
251
+ - name: question
252
+ dtype: string
253
+ - name: answers
254
+ sequence:
255
+ - name: answer_start
256
+ dtype: int32
257
+ - name: text
258
+ dtype: string
259
+ - name: id
260
+ dtype: string
261
+ splits:
262
+ - name: test
263
+ num_bytes: 3916758
264
+ num_examples: 5253
265
+ download_size: 10075488
266
+ dataset_size: 3916758
267
+ - config_name: mlqa-translate-test.hi
268
+ features:
269
+ - name: context
270
+ dtype: string
271
+ - name: question
272
+ dtype: string
273
+ - name: answers
274
+ sequence:
275
+ - name: answer_start
276
+ dtype: int32
277
+ - name: text
278
+ dtype: string
279
+ - name: id
280
+ dtype: string
281
+ splits:
282
+ - name: test
283
+ num_bytes: 4608811
284
+ num_examples: 4918
285
+ download_size: 10075488
286
+ dataset_size: 4608811
287
+ - config_name: mlqa.ar.ar
288
+ features:
289
+ - name: context
290
+ dtype: string
291
+ - name: question
292
+ dtype: string
293
+ - name: answers
294
+ sequence:
295
+ - name: answer_start
296
+ dtype: int32
297
+ - name: text
298
+ dtype: string
299
+ - name: id
300
+ dtype: string
301
+ splits:
302
+ - name: test
303
+ num_bytes: 8216837
304
+ num_examples: 5335
305
+ - name: validation
306
+ num_bytes: 808830
307
+ num_examples: 517
308
+ download_size: 75719050
309
+ dataset_size: 9025667
310
+ - config_name: mlqa.ar.de
311
+ features:
312
+ - name: context
313
+ dtype: string
314
+ - name: question
315
+ dtype: string
316
+ - name: answers
317
+ sequence:
318
+ - name: answer_start
319
+ dtype: int32
320
+ - name: text
321
+ dtype: string
322
+ - name: id
323
+ dtype: string
324
+ splits:
325
+ - name: test
326
+ num_bytes: 2132247
327
+ num_examples: 1649
328
+ - name: validation
329
+ num_bytes: 358554
330
+ num_examples: 207
331
+ download_size: 75719050
332
+ dataset_size: 2490801
333
+ - config_name: mlqa.ar.vi
334
+ features:
335
+ - name: context
336
+ dtype: string
337
+ - name: question
338
+ dtype: string
339
+ - name: answers
340
+ sequence:
341
+ - name: answer_start
342
+ dtype: int32
343
+ - name: text
344
+ dtype: string
345
+ - name: id
346
+ dtype: string
347
+ splits:
348
+ - name: test
349
+ num_bytes: 3235363
350
+ num_examples: 2047
351
+ - name: validation
352
+ num_bytes: 283834
353
+ num_examples: 163
354
+ download_size: 75719050
355
+ dataset_size: 3519197
356
+ - config_name: mlqa.ar.zh
357
+ features:
358
+ - name: context
359
+ dtype: string
360
+ - name: question
361
+ dtype: string
362
+ - name: answers
363
+ sequence:
364
+ - name: answer_start
365
+ dtype: int32
366
+ - name: text
367
+ dtype: string
368
+ - name: id
369
+ dtype: string
370
+ splits:
371
+ - name: test
372
+ num_bytes: 3175660
373
+ num_examples: 1912
374
+ - name: validation
375
+ num_bytes: 334016
376
+ num_examples: 188
377
+ download_size: 75719050
378
+ dataset_size: 3509676
379
+ - config_name: mlqa.ar.en
380
+ features:
381
+ - name: context
382
+ dtype: string
383
+ - name: question
384
+ dtype: string
385
+ - name: answers
386
+ sequence:
387
+ - name: answer_start
388
+ dtype: int32
389
+ - name: text
390
+ dtype: string
391
+ - name: id
392
+ dtype: string
393
+ splits:
394
+ - name: test
395
+ num_bytes: 8074057
396
+ num_examples: 5335
397
+ - name: validation
398
+ num_bytes: 794775
399
+ num_examples: 517
400
+ download_size: 75719050
401
+ dataset_size: 8868832
402
+ - config_name: mlqa.ar.es
403
+ features:
404
+ - name: context
405
+ dtype: string
406
+ - name: question
407
+ dtype: string
408
+ - name: answers
409
+ sequence:
410
+ - name: answer_start
411
+ dtype: int32
412
+ - name: text
413
+ dtype: string
414
+ - name: id
415
+ dtype: string
416
+ splits:
417
+ - name: test
418
+ num_bytes: 2981237
419
+ num_examples: 1978
420
+ - name: validation
421
+ num_bytes: 223188
422
+ num_examples: 161
423
+ download_size: 75719050
424
+ dataset_size: 3204425
425
+ - config_name: mlqa.ar.hi
426
+ features:
427
+ - name: context
428
+ dtype: string
429
+ - name: question
430
+ dtype: string
431
+ - name: answers
432
+ sequence:
433
+ - name: answer_start
434
+ dtype: int32
435
+ - name: text
436
+ dtype: string
437
+ - name: id
438
+ dtype: string
439
+ splits:
440
+ - name: test
441
+ num_bytes: 2993225
442
+ num_examples: 1831
443
+ - name: validation
444
+ num_bytes: 276727
445
+ num_examples: 186
446
+ download_size: 75719050
447
+ dataset_size: 3269952
448
+ - config_name: mlqa.de.ar
449
+ features:
450
+ - name: context
451
+ dtype: string
452
+ - name: question
453
+ dtype: string
454
+ - name: answers
455
+ sequence:
456
+ - name: answer_start
457
+ dtype: int32
458
+ - name: text
459
+ dtype: string
460
+ - name: id
461
+ dtype: string
462
+ splits:
463
+ - name: test
464
+ num_bytes: 1587005
465
+ num_examples: 1649
466
+ - name: validation
467
+ num_bytes: 195822
468
+ num_examples: 207
469
+ download_size: 75719050
470
+ dataset_size: 1782827
471
+ - config_name: mlqa.de.de
472
+ features:
473
+ - name: context
474
+ dtype: string
475
+ - name: question
476
+ dtype: string
477
+ - name: answers
478
+ sequence:
479
+ - name: answer_start
480
+ dtype: int32
481
+ - name: text
482
+ dtype: string
483
+ - name: id
484
+ dtype: string
485
+ splits:
486
+ - name: test
487
+ num_bytes: 4274496
488
+ num_examples: 4517
489
+ - name: validation
490
+ num_bytes: 477366
491
+ num_examples: 512
492
+ download_size: 75719050
493
+ dataset_size: 4751862
494
+ - config_name: mlqa.de.vi
495
+ features:
496
+ - name: context
497
+ dtype: string
498
+ - name: question
499
+ dtype: string
500
+ - name: answers
501
+ sequence:
502
+ - name: answer_start
503
+ dtype: int32
504
+ - name: text
505
+ dtype: string
506
+ - name: id
507
+ dtype: string
508
+ splits:
509
+ - name: test
510
+ num_bytes: 1654540
511
+ num_examples: 1675
512
+ - name: validation
513
+ num_bytes: 211985
514
+ num_examples: 182
515
+ download_size: 75719050
516
+ dataset_size: 1866525
517
+ - config_name: mlqa.de.zh
518
+ features:
519
+ - name: context
520
+ dtype: string
521
+ - name: question
522
+ dtype: string
523
+ - name: answers
524
+ sequence:
525
+ - name: answer_start
526
+ dtype: int32
527
+ - name: text
528
+ dtype: string
529
+ - name: id
530
+ dtype: string
531
+ splits:
532
+ - name: test
533
+ num_bytes: 1645937
534
+ num_examples: 1621
535
+ - name: validation
536
+ num_bytes: 180114
537
+ num_examples: 190
538
+ download_size: 75719050
539
+ dataset_size: 1826051
540
+ - config_name: mlqa.de.en
541
+ features:
542
+ - name: context
543
+ dtype: string
544
+ - name: question
545
+ dtype: string
546
+ - name: answers
547
+ sequence:
548
+ - name: answer_start
549
+ dtype: int32
550
+ - name: text
551
+ dtype: string
552
+ - name: id
553
+ dtype: string
554
+ splits:
555
+ - name: test
556
+ num_bytes: 4251153
557
+ num_examples: 4517
558
+ - name: validation
559
+ num_bytes: 474863
560
+ num_examples: 512
561
+ download_size: 75719050
562
+ dataset_size: 4726016
563
+ - config_name: mlqa.de.es
564
+ features:
565
+ - name: context
566
+ dtype: string
567
+ - name: question
568
+ dtype: string
569
+ - name: answers
570
+ sequence:
571
+ - name: answer_start
572
+ dtype: int32
573
+ - name: text
574
+ dtype: string
575
+ - name: id
576
+ dtype: string
577
+ splits:
578
+ - name: test
579
+ num_bytes: 1678176
580
+ num_examples: 1776
581
+ - name: validation
582
+ num_bytes: 166193
583
+ num_examples: 196
584
+ download_size: 75719050
585
+ dataset_size: 1844369
586
+ - config_name: mlqa.de.hi
587
+ features:
588
+ - name: context
589
+ dtype: string
590
+ - name: question
591
+ dtype: string
592
+ - name: answers
593
+ sequence:
594
+ - name: answer_start
595
+ dtype: int32
596
+ - name: text
597
+ dtype: string
598
+ - name: id
599
+ dtype: string
600
+ splits:
601
+ - name: test
602
+ num_bytes: 1343983
603
+ num_examples: 1430
604
+ - name: validation
605
+ num_bytes: 150679
606
+ num_examples: 163
607
+ download_size: 75719050
608
+ dataset_size: 1494662
609
+ - config_name: mlqa.vi.ar
610
+ features:
611
+ - name: context
612
+ dtype: string
613
+ - name: question
614
+ dtype: string
615
+ - name: answers
616
+ sequence:
617
+ - name: answer_start
618
+ dtype: int32
619
+ - name: text
620
+ dtype: string
621
+ - name: id
622
+ dtype: string
623
+ splits:
624
+ - name: test
625
+ num_bytes: 3164094
626
+ num_examples: 2047
627
+ - name: validation
628
+ num_bytes: 226724
629
+ num_examples: 163
630
+ download_size: 75719050
631
+ dataset_size: 3390818
632
+ - config_name: mlqa.vi.de
633
+ features:
634
+ - name: context
635
+ dtype: string
636
+ - name: question
637
+ dtype: string
638
+ - name: answers
639
+ sequence:
640
+ - name: answer_start
641
+ dtype: int32
642
+ - name: text
643
+ dtype: string
644
+ - name: id
645
+ dtype: string
646
+ splits:
647
+ - name: test
648
+ num_bytes: 2189315
649
+ num_examples: 1675
650
+ - name: validation
651
+ num_bytes: 272794
652
+ num_examples: 182
653
+ download_size: 75719050
654
+ dataset_size: 2462109
655
+ - config_name: mlqa.vi.vi
656
+ features:
657
+ - name: context
658
+ dtype: string
659
+ - name: question
660
+ dtype: string
661
+ - name: answers
662
+ sequence:
663
+ - name: answer_start
664
+ dtype: int32
665
+ - name: text
666
+ dtype: string
667
+ - name: id
668
+ dtype: string
669
+ splits:
670
+ - name: test
671
+ num_bytes: 7807045
672
+ num_examples: 5495
673
+ - name: validation
674
+ num_bytes: 715291
675
+ num_examples: 511
676
+ download_size: 75719050
677
+ dataset_size: 8522336
678
+ - config_name: mlqa.vi.zh
679
+ features:
680
+ - name: context
681
+ dtype: string
682
+ - name: question
683
+ dtype: string
684
+ - name: answers
685
+ sequence:
686
+ - name: answer_start
687
+ dtype: int32
688
+ - name: text
689
+ dtype: string
690
+ - name: id
691
+ dtype: string
692
+ splits:
693
+ - name: test
694
+ num_bytes: 2947458
695
+ num_examples: 1943
696
+ - name: validation
697
+ num_bytes: 265154
698
+ num_examples: 184
699
+ download_size: 75719050
700
+ dataset_size: 3212612
701
+ - config_name: mlqa.vi.en
702
+ features:
703
+ - name: context
704
+ dtype: string
705
+ - name: question
706
+ dtype: string
707
+ - name: answers
708
+ sequence:
709
+ - name: answer_start
710
+ dtype: int32
711
+ - name: text
712
+ dtype: string
713
+ - name: id
714
+ dtype: string
715
+ splits:
716
+ - name: test
717
+ num_bytes: 7727204
718
+ num_examples: 5495
719
+ - name: validation
720
+ num_bytes: 707925
721
+ num_examples: 511
722
+ download_size: 75719050
723
+ dataset_size: 8435129
724
+ - config_name: mlqa.vi.es
725
+ features:
726
+ - name: context
727
+ dtype: string
728
+ - name: question
729
+ dtype: string
730
+ - name: answers
731
+ sequence:
732
+ - name: answer_start
733
+ dtype: int32
734
+ - name: text
735
+ dtype: string
736
+ - name: id
737
+ dtype: string
738
+ splits:
739
+ - name: test
740
+ num_bytes: 2822481
741
+ num_examples: 2018
742
+ - name: validation
743
+ num_bytes: 279235
744
+ num_examples: 189
745
+ download_size: 75719050
746
+ dataset_size: 3101716
747
+ - config_name: mlqa.vi.hi
748
+ features:
749
+ - name: context
750
+ dtype: string
751
+ - name: question
752
+ dtype: string
753
+ - name: answers
754
+ sequence:
755
+ - name: answer_start
756
+ dtype: int32
757
+ - name: text
758
+ dtype: string
759
+ - name: id
760
+ dtype: string
761
+ splits:
762
+ - name: test
763
+ num_bytes: 2738045
764
+ num_examples: 1947
765
+ - name: validation
766
+ num_bytes: 251470
767
+ num_examples: 177
768
+ download_size: 75719050
769
+ dataset_size: 2989515
770
+ - config_name: mlqa.zh.ar
771
+ features:
772
+ - name: context
773
+ dtype: string
774
+ - name: question
775
+ dtype: string
776
+ - name: answers
777
+ sequence:
778
+ - name: answer_start
779
+ dtype: int32
780
+ - name: text
781
+ dtype: string
782
+ - name: id
783
+ dtype: string
784
+ splits:
785
+ - name: test
786
+ num_bytes: 1697005
787
+ num_examples: 1912
788
+ - name: validation
789
+ num_bytes: 171743
790
+ num_examples: 188
791
+ download_size: 75719050
792
+ dataset_size: 1868748
793
+ - config_name: mlqa.zh.de
794
+ features:
795
+ - name: context
796
+ dtype: string
797
+ - name: question
798
+ dtype: string
799
+ - name: answers
800
+ sequence:
801
+ - name: answer_start
802
+ dtype: int32
803
+ - name: text
804
+ dtype: string
805
+ - name: id
806
+ dtype: string
807
+ splits:
808
+ - name: test
809
+ num_bytes: 1356268
810
+ num_examples: 1621
811
+ - name: validation
812
+ num_bytes: 170686
813
+ num_examples: 190
814
+ download_size: 75719050
815
+ dataset_size: 1526954
816
+ - config_name: mlqa.zh.vi
817
+ features:
818
+ - name: context
819
+ dtype: string
820
+ - name: question
821
+ dtype: string
822
+ - name: answers
823
+ sequence:
824
+ - name: answer_start
825
+ dtype: int32
826
+ - name: text
827
+ dtype: string
828
+ - name: id
829
+ dtype: string
830
+ splits:
831
+ - name: test
832
+ num_bytes: 1770535
833
+ num_examples: 1943
834
+ - name: validation
835
+ num_bytes: 169651
836
+ num_examples: 184
837
+ download_size: 75719050
838
+ dataset_size: 1940186
839
+ - config_name: mlqa.zh.zh
840
+ features:
841
+ - name: context
842
+ dtype: string
843
+ - name: question
844
+ dtype: string
845
+ - name: answers
846
+ sequence:
847
+ - name: answer_start
848
+ dtype: int32
849
+ - name: text
850
+ dtype: string
851
+ - name: id
852
+ dtype: string
853
+ splits:
854
+ - name: test
855
+ num_bytes: 4324740
856
+ num_examples: 5137
857
+ - name: validation
858
+ num_bytes: 433960
859
+ num_examples: 504
860
+ download_size: 75719050
861
+ dataset_size: 4758700
862
+ - config_name: mlqa.zh.en
863
+ features:
864
+ - name: context
865
+ dtype: string
866
+ - name: question
867
+ dtype: string
868
+ - name: answers
869
+ sequence:
870
+ - name: answer_start
871
+ dtype: int32
872
+ - name: text
873
+ dtype: string
874
+ - name: id
875
+ dtype: string
876
+ splits:
877
+ - name: test
878
+ num_bytes: 4353361
879
+ num_examples: 5137
880
+ - name: validation
881
+ num_bytes: 437016
882
+ num_examples: 504
883
+ download_size: 75719050
884
+ dataset_size: 4790377
885
+ - config_name: mlqa.zh.es
886
+ features:
887
+ - name: context
888
+ dtype: string
889
+ - name: question
890
+ dtype: string
891
+ - name: answers
892
+ sequence:
893
+ - name: answer_start
894
+ dtype: int32
895
+ - name: text
896
+ dtype: string
897
+ - name: id
898
+ dtype: string
899
+ splits:
900
+ - name: test
901
+ num_bytes: 1697983
902
+ num_examples: 1947
903
+ - name: validation
904
+ num_bytes: 134693
905
+ num_examples: 161
906
+ download_size: 75719050
907
+ dataset_size: 1832676
908
+ - config_name: mlqa.zh.hi
909
+ features:
910
+ - name: context
911
+ dtype: string
912
+ - name: question
913
+ dtype: string
914
+ - name: answers
915
+ sequence:
916
+ - name: answer_start
917
+ dtype: int32
918
+ - name: text
919
+ dtype: string
920
+ - name: id
921
+ dtype: string
922
+ splits:
923
+ - name: test
924
+ num_bytes: 1547159
925
+ num_examples: 1767
926
+ - name: validation
927
+ num_bytes: 180928
928
+ num_examples: 189
929
+ download_size: 75719050
930
+ dataset_size: 1728087
931
+ - config_name: mlqa.en.ar
932
+ features:
933
+ - name: context
934
+ dtype: string
935
+ - name: question
936
+ dtype: string
937
+ - name: answers
938
+ sequence:
939
+ - name: answer_start
940
+ dtype: int32
941
+ - name: text
942
+ dtype: string
943
+ - name: id
944
+ dtype: string
945
+ splits:
946
+ - name: test
947
+ num_bytes: 6641971
948
+ num_examples: 5335
949
+ - name: validation
950
+ num_bytes: 621075
951
+ num_examples: 517
952
+ download_size: 75719050
953
+ dataset_size: 7263046
954
+ - config_name: mlqa.en.de
955
+ features:
956
+ - name: context
957
+ dtype: string
958
+ - name: question
959
+ dtype: string
960
+ - name: answers
961
+ sequence:
962
+ - name: answer_start
963
+ dtype: int32
964
+ - name: text
965
+ dtype: string
966
+ - name: id
967
+ dtype: string
968
+ splits:
969
+ - name: test
970
+ num_bytes: 4966262
971
+ num_examples: 4517
972
+ - name: validation
973
+ num_bytes: 584725
974
+ num_examples: 512
975
+ download_size: 75719050
976
+ dataset_size: 5550987
977
+ - config_name: mlqa.en.vi
978
+ features:
979
+ - name: context
980
+ dtype: string
981
+ - name: question
982
+ dtype: string
983
+ - name: answers
984
+ sequence:
985
+ - name: answer_start
986
+ dtype: int32
987
+ - name: text
988
+ dtype: string
989
+ - name: id
990
+ dtype: string
991
+ splits:
992
+ - name: test
993
+ num_bytes: 6958087
994
+ num_examples: 5495
995
+ - name: validation
996
+ num_bytes: 631268
997
+ num_examples: 511
998
+ download_size: 75719050
999
+ dataset_size: 7589355
1000
+ - config_name: mlqa.en.zh
1001
+ features:
1002
+ - name: context
1003
+ dtype: string
1004
+ - name: question
1005
+ dtype: string
1006
+ - name: answers
1007
+ sequence:
1008
+ - name: answer_start
1009
+ dtype: int32
1010
+ - name: text
1011
+ dtype: string
1012
+ - name: id
1013
+ dtype: string
1014
+ splits:
1015
+ - name: test
1016
+ num_bytes: 6441614
1017
+ num_examples: 5137
1018
+ - name: validation
1019
+ num_bytes: 598772
1020
+ num_examples: 504
1021
+ download_size: 75719050
1022
+ dataset_size: 7040386
1023
+ - config_name: mlqa.en.en
1024
+ features:
1025
+ - name: context
1026
+ dtype: string
1027
+ - name: question
1028
+ dtype: string
1029
+ - name: answers
1030
+ sequence:
1031
+ - name: answer_start
1032
+ dtype: int32
1033
+ - name: text
1034
+ dtype: string
1035
+ - name: id
1036
+ dtype: string
1037
+ splits:
1038
+ - name: test
1039
+ num_bytes: 13787522
1040
+ num_examples: 11590
1041
+ - name: validation
1042
+ num_bytes: 1307399
1043
+ num_examples: 1148
1044
+ download_size: 75719050
1045
+ dataset_size: 15094921
1046
+ - config_name: mlqa.en.es
1047
+ features:
1048
+ - name: context
1049
+ dtype: string
1050
+ - name: question
1051
+ dtype: string
1052
+ - name: answers
1053
+ sequence:
1054
+ - name: answer_start
1055
+ dtype: int32
1056
+ - name: text
1057
+ dtype: string
1058
+ - name: id
1059
+ dtype: string
1060
+ splits:
1061
+ - name: test
1062
+ num_bytes: 6074990
1063
+ num_examples: 5253
1064
+ - name: validation
1065
+ num_bytes: 545657
1066
+ num_examples: 500
1067
+ download_size: 75719050
1068
+ dataset_size: 6620647
1069
+ - config_name: mlqa.en.hi
1070
+ features:
1071
+ - name: context
1072
+ dtype: string
1073
+ - name: question
1074
+ dtype: string
1075
+ - name: answers
1076
+ sequence:
1077
+ - name: answer_start
1078
+ dtype: int32
1079
+ - name: text
1080
+ dtype: string
1081
+ - name: id
1082
+ dtype: string
1083
+ splits:
1084
+ - name: test
1085
+ num_bytes: 6293785
1086
+ num_examples: 4918
1087
+ - name: validation
1088
+ num_bytes: 614223
1089
+ num_examples: 507
1090
+ download_size: 75719050
1091
+ dataset_size: 6908008
1092
+ - config_name: mlqa.es.ar
1093
+ features:
1094
+ - name: context
1095
+ dtype: string
1096
+ - name: question
1097
+ dtype: string
1098
+ - name: answers
1099
+ sequence:
1100
+ - name: answer_start
1101
+ dtype: int32
1102
+ - name: text
1103
+ dtype: string
1104
+ - name: id
1105
+ dtype: string
1106
+ splits:
1107
+ - name: test
1108
+ num_bytes: 1696778
1109
+ num_examples: 1978
1110
+ - name: validation
1111
+ num_bytes: 145105
1112
+ num_examples: 161
1113
+ download_size: 75719050
1114
+ dataset_size: 1841883
1115
+ - config_name: mlqa.es.de
1116
+ features:
1117
+ - name: context
1118
+ dtype: string
1119
+ - name: question
1120
+ dtype: string
1121
+ - name: answers
1122
+ sequence:
1123
+ - name: answer_start
1124
+ dtype: int32
1125
+ - name: text
1126
+ dtype: string
1127
+ - name: id
1128
+ dtype: string
1129
+ splits:
1130
+ - name: test
1131
+ num_bytes: 1361983
1132
+ num_examples: 1776
1133
+ - name: validation
1134
+ num_bytes: 139968
1135
+ num_examples: 196
1136
+ download_size: 75719050
1137
+ dataset_size: 1501951
1138
+ - config_name: mlqa.es.vi
1139
+ features:
1140
+ - name: context
1141
+ dtype: string
1142
+ - name: question
1143
+ dtype: string
1144
+ - name: answers
1145
+ sequence:
1146
+ - name: answer_start
1147
+ dtype: int32
1148
+ - name: text
1149
+ dtype: string
1150
+ - name: id
1151
+ dtype: string
1152
+ splits:
1153
+ - name: test
1154
+ num_bytes: 1707141
1155
+ num_examples: 2018
1156
+ - name: validation
1157
+ num_bytes: 172801
1158
+ num_examples: 189
1159
+ download_size: 75719050
1160
+ dataset_size: 1879942
1161
+ - config_name: mlqa.es.zh
1162
+ features:
1163
+ - name: context
1164
+ dtype: string
1165
+ - name: question
1166
+ dtype: string
1167
+ - name: answers
1168
+ sequence:
1169
+ - name: answer_start
1170
+ dtype: int32
1171
+ - name: text
1172
+ dtype: string
1173
+ - name: id
1174
+ dtype: string
1175
+ splits:
1176
+ - name: test
1177
+ num_bytes: 1635294
1178
+ num_examples: 1947
1179
+ - name: validation
1180
+ num_bytes: 122829
1181
+ num_examples: 161
1182
+ download_size: 75719050
1183
+ dataset_size: 1758123
1184
+ - config_name: mlqa.es.en
1185
+ features:
1186
+ - name: context
1187
+ dtype: string
1188
+ - name: question
1189
+ dtype: string
1190
+ - name: answers
1191
+ sequence:
1192
+ - name: answer_start
1193
+ dtype: int32
1194
+ - name: text
1195
+ dtype: string
1196
+ - name: id
1197
+ dtype: string
1198
+ splits:
1199
+ - name: test
1200
+ num_bytes: 4249431
1201
+ num_examples: 5253
1202
+ - name: validation
1203
+ num_bytes: 408169
1204
+ num_examples: 500
1205
+ download_size: 75719050
1206
+ dataset_size: 4657600
1207
+ - config_name: mlqa.es.es
1208
+ features:
1209
+ - name: context
1210
+ dtype: string
1211
+ - name: question
1212
+ dtype: string
1213
+ - name: answers
1214
+ sequence:
1215
+ - name: answer_start
1216
+ dtype: int32
1217
+ - name: text
1218
+ dtype: string
1219
+ - name: id
1220
+ dtype: string
1221
+ splits:
1222
+ - name: test
1223
+ num_bytes: 4281273
1224
+ num_examples: 5253
1225
+ - name: validation
1226
+ num_bytes: 411196
1227
+ num_examples: 500
1228
+ download_size: 75719050
1229
+ dataset_size: 4692469
1230
+ - config_name: mlqa.es.hi
1231
+ features:
1232
+ - name: context
1233
+ dtype: string
1234
+ - name: question
1235
+ dtype: string
1236
+ - name: answers
1237
+ sequence:
1238
+ - name: answer_start
1239
+ dtype: int32
1240
+ - name: text
1241
+ dtype: string
1242
+ - name: id
1243
+ dtype: string
1244
+ splits:
1245
+ - name: test
1246
+ num_bytes: 1489611
1247
+ num_examples: 1723
1248
+ - name: validation
1249
+ num_bytes: 178003
1250
+ num_examples: 187
1251
+ download_size: 75719050
1252
+ dataset_size: 1667614
1253
+ - config_name: mlqa.hi.ar
1254
+ features:
1255
+ - name: context
1256
+ dtype: string
1257
+ - name: question
1258
+ dtype: string
1259
+ - name: answers
1260
+ sequence:
1261
+ - name: answer_start
1262
+ dtype: int32
1263
+ - name: text
1264
+ dtype: string
1265
+ - name: id
1266
+ dtype: string
1267
+ splits:
1268
+ - name: test
1269
+ num_bytes: 4374373
1270
+ num_examples: 1831
1271
+ - name: validation
1272
+ num_bytes: 402817
1273
+ num_examples: 186
1274
+ download_size: 75719050
1275
+ dataset_size: 4777190
1276
+ - config_name: mlqa.hi.de
1277
+ features:
1278
+ - name: context
1279
+ dtype: string
1280
+ - name: question
1281
+ dtype: string
1282
+ - name: answers
1283
+ sequence:
1284
+ - name: answer_start
1285
+ dtype: int32
1286
+ - name: text
1287
+ dtype: string
1288
+ - name: id
1289
+ dtype: string
1290
+ splits:
1291
+ - name: test
1292
+ num_bytes: 2961556
1293
+ num_examples: 1430
1294
+ - name: validation
1295
+ num_bytes: 294325
1296
+ num_examples: 163
1297
+ download_size: 75719050
1298
+ dataset_size: 3255881
1299
+ - config_name: mlqa.hi.vi
1300
+ features:
1301
+ - name: context
1302
+ dtype: string
1303
+ - name: question
1304
+ dtype: string
1305
+ - name: answers
1306
+ sequence:
1307
+ - name: answer_start
1308
+ dtype: int32
1309
+ - name: text
1310
+ dtype: string
1311
+ - name: id
1312
+ dtype: string
1313
+ splits:
1314
+ - name: test
1315
+ num_bytes: 4664436
1316
+ num_examples: 1947
1317
+ - name: validation
1318
+ num_bytes: 411654
1319
+ num_examples: 177
1320
+ download_size: 75719050
1321
+ dataset_size: 5076090
1322
+ - config_name: mlqa.hi.zh
1323
+ features:
1324
+ - name: context
1325
+ dtype: string
1326
+ - name: question
1327
+ dtype: string
1328
+ - name: answers
1329
+ sequence:
1330
+ - name: answer_start
1331
+ dtype: int32
1332
+ - name: text
1333
+ dtype: string
1334
+ - name: id
1335
+ dtype: string
1336
+ splits:
1337
+ - name: test
1338
+ num_bytes: 4281309
1339
+ num_examples: 1767
1340
+ - name: validation
1341
+ num_bytes: 416192
1342
+ num_examples: 189
1343
+ download_size: 75719050
1344
+ dataset_size: 4697501
1345
+ - config_name: mlqa.hi.en
1346
+ features:
1347
+ - name: context
1348
+ dtype: string
1349
+ - name: question
1350
+ dtype: string
1351
+ - name: answers
1352
+ sequence:
1353
+ - name: answer_start
1354
+ dtype: int32
1355
+ - name: text
1356
+ dtype: string
1357
+ - name: id
1358
+ dtype: string
1359
+ splits:
1360
+ - name: test
1361
+ num_bytes: 11245629
1362
+ num_examples: 4918
1363
+ - name: validation
1364
+ num_bytes: 1076115
1365
+ num_examples: 507
1366
+ download_size: 75719050
1367
+ dataset_size: 12321744
1368
+ - config_name: mlqa.hi.es
1369
+ features:
1370
+ - name: context
1371
+ dtype: string
1372
+ - name: question
1373
+ dtype: string
1374
+ - name: answers
1375
+ sequence:
1376
+ - name: answer_start
1377
+ dtype: int32
1378
+ - name: text
1379
+ dtype: string
1380
+ - name: id
1381
+ dtype: string
1382
+ splits:
1383
+ - name: test
1384
+ num_bytes: 3789337
1385
+ num_examples: 1723
1386
+ - name: validation
1387
+ num_bytes: 412469
1388
+ num_examples: 187
1389
+ download_size: 75719050
1390
+ dataset_size: 4201806
1391
+ - config_name: mlqa.hi.hi
1392
+ features:
1393
+ - name: context
1394
+ dtype: string
1395
+ - name: question
1396
+ dtype: string
1397
+ - name: answers
1398
+ sequence:
1399
+ - name: answer_start
1400
+ dtype: int32
1401
+ - name: text
1402
+ dtype: string
1403
+ - name: id
1404
+ dtype: string
1405
+ splits:
1406
+ - name: test
1407
+ num_bytes: 11606982
1408
+ num_examples: 4918
1409
+ - name: validation
1410
+ num_bytes: 1115055
1411
+ num_examples: 507
1412
+ download_size: 75719050
1413
+ dataset_size: 12722037
1414
  ---
1415
 
1416
  # Dataset Card for "mlqa"
 
1647
 
1648
  ### Contributions
1649
 
1650
+ Thanks to [@patrickvonplaten](https://github.com/patrickvonplaten), [@M-Salti](https://github.com/M-Salti), [@lewtun](https://github.com/lewtun), [@thomwolf](https://github.com/thomwolf), [@mariamabarham](https://github.com/mariamabarham), [@lhoestq](https://github.com/lhoestq) for adding this dataset.