Datasets:
parquet-converter
commited on
Commit
•
c9e7943
1
Parent(s):
4ed67dc
Update parquet files
Browse files- README.md +0 -495
- dataset_infos.json +0 -1
- de-zh/un_multi-train.parquet +3 -0
- un_multi.py +0 -109
README.md
DELETED
@@ -1,495 +0,0 @@
|
|
1 |
-
---
|
2 |
-
annotations_creators:
|
3 |
-
- found
|
4 |
-
language_creators:
|
5 |
-
- found
|
6 |
-
language:
|
7 |
-
- ar
|
8 |
-
- de
|
9 |
-
- en
|
10 |
-
- es
|
11 |
-
- fr
|
12 |
-
- ru
|
13 |
-
- zh
|
14 |
-
license:
|
15 |
-
- unknown
|
16 |
-
multilinguality:
|
17 |
-
- multilingual
|
18 |
-
size_categories:
|
19 |
-
- 100K<n<1M
|
20 |
-
source_datasets:
|
21 |
-
- original
|
22 |
-
task_categories:
|
23 |
-
- translation
|
24 |
-
task_ids: []
|
25 |
-
paperswithcode_id: multiun
|
26 |
-
pretty_name: Multilingual Corpus from United Nation Documents
|
27 |
-
configs:
|
28 |
-
- ar-de
|
29 |
-
- ar-en
|
30 |
-
- ar-es
|
31 |
-
- ar-fr
|
32 |
-
- ar-ru
|
33 |
-
- ar-zh
|
34 |
-
- de-en
|
35 |
-
- de-es
|
36 |
-
- de-fr
|
37 |
-
- de-ru
|
38 |
-
- de-zh
|
39 |
-
- en-es
|
40 |
-
- en-fr
|
41 |
-
- en-ru
|
42 |
-
- en-zh
|
43 |
-
- es-fr
|
44 |
-
- es-ru
|
45 |
-
- es-zh
|
46 |
-
- fr-ru
|
47 |
-
- fr-zh
|
48 |
-
- ru-zh
|
49 |
-
dataset_info:
|
50 |
-
- config_name: ar-de
|
51 |
-
features:
|
52 |
-
- name: translation
|
53 |
-
dtype:
|
54 |
-
translation:
|
55 |
-
languages:
|
56 |
-
- ar
|
57 |
-
- de
|
58 |
-
splits:
|
59 |
-
- name: train
|
60 |
-
num_bytes: 94466397
|
61 |
-
num_examples: 165090
|
62 |
-
download_size: 21869935
|
63 |
-
dataset_size: 94466397
|
64 |
-
- config_name: ar-en
|
65 |
-
features:
|
66 |
-
- name: translation
|
67 |
-
dtype:
|
68 |
-
translation:
|
69 |
-
languages:
|
70 |
-
- ar
|
71 |
-
- en
|
72 |
-
splits:
|
73 |
-
- name: train
|
74 |
-
num_bytes: 4189852369
|
75 |
-
num_examples: 9759125
|
76 |
-
download_size: 1036296368
|
77 |
-
dataset_size: 4189852369
|
78 |
-
- config_name: ar-es
|
79 |
-
features:
|
80 |
-
- name: translation
|
81 |
-
dtype:
|
82 |
-
translation:
|
83 |
-
languages:
|
84 |
-
- ar
|
85 |
-
- es
|
86 |
-
splits:
|
87 |
-
- name: train
|
88 |
-
num_bytes: 4509675284
|
89 |
-
num_examples: 10119379
|
90 |
-
download_size: 1101206667
|
91 |
-
dataset_size: 4509675284
|
92 |
-
- config_name: ar-fr
|
93 |
-
features:
|
94 |
-
- name: translation
|
95 |
-
dtype:
|
96 |
-
translation:
|
97 |
-
languages:
|
98 |
-
- ar
|
99 |
-
- fr
|
100 |
-
splits:
|
101 |
-
- name: train
|
102 |
-
num_bytes: 4516850009
|
103 |
-
num_examples: 9929567
|
104 |
-
download_size: 1109705925
|
105 |
-
dataset_size: 4516850009
|
106 |
-
- config_name: ar-ru
|
107 |
-
features:
|
108 |
-
- name: translation
|
109 |
-
dtype:
|
110 |
-
translation:
|
111 |
-
languages:
|
112 |
-
- ar
|
113 |
-
- ru
|
114 |
-
splits:
|
115 |
-
- name: train
|
116 |
-
num_bytes: 5932866867
|
117 |
-
num_examples: 10206243
|
118 |
-
download_size: 1261123878
|
119 |
-
dataset_size: 5932866867
|
120 |
-
- config_name: ar-zh
|
121 |
-
features:
|
122 |
-
- name: translation
|
123 |
-
dtype:
|
124 |
-
translation:
|
125 |
-
languages:
|
126 |
-
- ar
|
127 |
-
- zh
|
128 |
-
splits:
|
129 |
-
- name: train
|
130 |
-
num_bytes: 3781658413
|
131 |
-
num_examples: 9832293
|
132 |
-
download_size: 1009696775
|
133 |
-
dataset_size: 3781658413
|
134 |
-
- config_name: de-en
|
135 |
-
features:
|
136 |
-
- name: translation
|
137 |
-
dtype:
|
138 |
-
translation:
|
139 |
-
languages:
|
140 |
-
- de
|
141 |
-
- en
|
142 |
-
splits:
|
143 |
-
- name: train
|
144 |
-
num_bytes: 76684549
|
145 |
-
num_examples: 162981
|
146 |
-
download_size: 19468529
|
147 |
-
dataset_size: 76684549
|
148 |
-
- config_name: de-es
|
149 |
-
features:
|
150 |
-
- name: translation
|
151 |
-
dtype:
|
152 |
-
translation:
|
153 |
-
languages:
|
154 |
-
- de
|
155 |
-
- es
|
156 |
-
splits:
|
157 |
-
- name: train
|
158 |
-
num_bytes: 80936653
|
159 |
-
num_examples: 162078
|
160 |
-
download_size: 20266591
|
161 |
-
dataset_size: 80936653
|
162 |
-
- config_name: de-fr
|
163 |
-
features:
|
164 |
-
- name: translation
|
165 |
-
dtype:
|
166 |
-
translation:
|
167 |
-
languages:
|
168 |
-
- de
|
169 |
-
- fr
|
170 |
-
splits:
|
171 |
-
- name: train
|
172 |
-
num_bytes: 81888435
|
173 |
-
num_examples: 164025
|
174 |
-
download_size: 20692837
|
175 |
-
dataset_size: 81888435
|
176 |
-
- config_name: de-ru
|
177 |
-
features:
|
178 |
-
- name: translation
|
179 |
-
dtype:
|
180 |
-
translation:
|
181 |
-
languages:
|
182 |
-
- de
|
183 |
-
- ru
|
184 |
-
splits:
|
185 |
-
- name: train
|
186 |
-
num_bytes: 111517934
|
187 |
-
num_examples: 164792
|
188 |
-
download_size: 23507789
|
189 |
-
dataset_size: 111517934
|
190 |
-
- config_name: de-zh
|
191 |
-
features:
|
192 |
-
- name: translation
|
193 |
-
dtype:
|
194 |
-
translation:
|
195 |
-
languages:
|
196 |
-
- de
|
197 |
-
- zh
|
198 |
-
splits:
|
199 |
-
- name: train
|
200 |
-
num_bytes: 70534818
|
201 |
-
num_examples: 176933
|
202 |
-
download_size: 19927209
|
203 |
-
dataset_size: 70534818
|
204 |
-
- config_name: en-es
|
205 |
-
features:
|
206 |
-
- name: translation
|
207 |
-
dtype:
|
208 |
-
translation:
|
209 |
-
languages:
|
210 |
-
- en
|
211 |
-
- es
|
212 |
-
splits:
|
213 |
-
- name: train
|
214 |
-
num_bytes: 4128141663
|
215 |
-
num_examples: 11350967
|
216 |
-
download_size: 1123164180
|
217 |
-
dataset_size: 4128141663
|
218 |
-
- config_name: en-fr
|
219 |
-
features:
|
220 |
-
- name: translation
|
221 |
-
dtype:
|
222 |
-
translation:
|
223 |
-
languages:
|
224 |
-
- en
|
225 |
-
- fr
|
226 |
-
splits:
|
227 |
-
- name: train
|
228 |
-
num_bytes: 4678055160
|
229 |
-
num_examples: 13172019
|
230 |
-
download_size: 1355002731
|
231 |
-
dataset_size: 4678055160
|
232 |
-
- config_name: en-ru
|
233 |
-
features:
|
234 |
-
- name: translation
|
235 |
-
dtype:
|
236 |
-
translation:
|
237 |
-
languages:
|
238 |
-
- en
|
239 |
-
- ru
|
240 |
-
splits:
|
241 |
-
- name: train
|
242 |
-
num_bytes: 5632662839
|
243 |
-
num_examples: 11654416
|
244 |
-
download_size: 1285801078
|
245 |
-
dataset_size: 5632662839
|
246 |
-
- config_name: en-zh
|
247 |
-
features:
|
248 |
-
- name: translation
|
249 |
-
dtype:
|
250 |
-
translation:
|
251 |
-
languages:
|
252 |
-
- en
|
253 |
-
- zh
|
254 |
-
splits:
|
255 |
-
- name: train
|
256 |
-
num_bytes: 2960376046
|
257 |
-
num_examples: 9564315
|
258 |
-
download_size: 900076520
|
259 |
-
dataset_size: 2960376046
|
260 |
-
- config_name: es-fr
|
261 |
-
features:
|
262 |
-
- name: translation
|
263 |
-
dtype:
|
264 |
-
translation:
|
265 |
-
languages:
|
266 |
-
- es
|
267 |
-
- fr
|
268 |
-
splits:
|
269 |
-
- name: train
|
270 |
-
num_bytes: 4454712498
|
271 |
-
num_examples: 11441889
|
272 |
-
download_size: 1195733510
|
273 |
-
dataset_size: 4454712498
|
274 |
-
- config_name: es-ru
|
275 |
-
features:
|
276 |
-
- name: translation
|
277 |
-
dtype:
|
278 |
-
translation:
|
279 |
-
languages:
|
280 |
-
- es
|
281 |
-
- ru
|
282 |
-
splits:
|
283 |
-
- name: train
|
284 |
-
num_bytes: 5442655730
|
285 |
-
num_examples: 10605056
|
286 |
-
download_size: 1228045966
|
287 |
-
dataset_size: 5442655730
|
288 |
-
- config_name: es-zh
|
289 |
-
features:
|
290 |
-
- name: translation
|
291 |
-
dtype:
|
292 |
-
translation:
|
293 |
-
languages:
|
294 |
-
- es
|
295 |
-
- zh
|
296 |
-
splits:
|
297 |
-
- name: train
|
298 |
-
num_bytes: 3223871198
|
299 |
-
num_examples: 9847770
|
300 |
-
download_size: 953250084
|
301 |
-
dataset_size: 3223871198
|
302 |
-
- config_name: fr-ru
|
303 |
-
features:
|
304 |
-
- name: translation
|
305 |
-
dtype:
|
306 |
-
translation:
|
307 |
-
languages:
|
308 |
-
- fr
|
309 |
-
- ru
|
310 |
-
splits:
|
311 |
-
- name: train
|
312 |
-
num_bytes: 5979879089
|
313 |
-
num_examples: 11761738
|
314 |
-
download_size: 1364307157
|
315 |
-
dataset_size: 5979879089
|
316 |
-
- config_name: fr-zh
|
317 |
-
features:
|
318 |
-
- name: translation
|
319 |
-
dtype:
|
320 |
-
translation:
|
321 |
-
languages:
|
322 |
-
- fr
|
323 |
-
- zh
|
324 |
-
splits:
|
325 |
-
- name: train
|
326 |
-
num_bytes: 3241098333
|
327 |
-
num_examples: 9690914
|
328 |
-
download_size: 962824881
|
329 |
-
dataset_size: 3241098333
|
330 |
-
- config_name: ru-zh
|
331 |
-
features:
|
332 |
-
- name: translation
|
333 |
-
dtype:
|
334 |
-
translation:
|
335 |
-
languages:
|
336 |
-
- ru
|
337 |
-
- zh
|
338 |
-
splits:
|
339 |
-
- name: train
|
340 |
-
num_bytes: 4233875537
|
341 |
-
num_examples: 9557007
|
342 |
-
download_size: 1037881127
|
343 |
-
dataset_size: 4233875537
|
344 |
-
---
|
345 |
-
|
346 |
-
# Dataset Card for [Dataset Name]
|
347 |
-
|
348 |
-
## Table of Contents
|
349 |
-
- [Dataset Description](#dataset-description)
|
350 |
-
- [Dataset Summary](#dataset-summary)
|
351 |
-
- [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
|
352 |
-
- [Languages](#languages)
|
353 |
-
- [Dataset Structure](#dataset-structure)
|
354 |
-
- [Data Instances](#data-instances)
|
355 |
-
- [Data Fields](#data-fields)
|
356 |
-
- [Data Splits](#data-splits)
|
357 |
-
- [Dataset Creation](#dataset-creation)
|
358 |
-
- [Curation Rationale](#curation-rationale)
|
359 |
-
- [Source Data](#source-data)
|
360 |
-
- [Annotations](#annotations)
|
361 |
-
- [Personal and Sensitive Information](#personal-and-sensitive-information)
|
362 |
-
- [Considerations for Using the Data](#considerations-for-using-the-data)
|
363 |
-
- [Social Impact of Dataset](#social-impact-of-dataset)
|
364 |
-
- [Discussion of Biases](#discussion-of-biases)
|
365 |
-
- [Other Known Limitations](#other-known-limitations)
|
366 |
-
- [Additional Information](#additional-information)
|
367 |
-
- [Dataset Curators](#dataset-curators)
|
368 |
-
- [Licensing Information](#licensing-information)
|
369 |
-
- [Citation Information](#citation-information)
|
370 |
-
- [Contributions](#contributions)
|
371 |
-
|
372 |
-
## Dataset Description
|
373 |
-
|
374 |
-
- **Homepage:**[MultiUN](http://www.euromatrixplus.net/multi-unp)
|
375 |
-
- **Repository:**
|
376 |
-
- **Paper:**
|
377 |
-
- **Leaderboard:**
|
378 |
-
- **Point of Contact:**
|
379 |
-
|
380 |
-
### Dataset Summary
|
381 |
-
|
382 |
-
This is a collection of translated documents from the United Nations.
|
383 |
-
This corpus is available in all 6 official languages of the UN
|
384 |
-
consisting of around 300 million words per language
|
385 |
-
|
386 |
-
### Supported Tasks and Leaderboards
|
387 |
-
|
388 |
-
The underlying task is machine translation.
|
389 |
-
|
390 |
-
### Languages
|
391 |
-
|
392 |
-
[More Information Needed]
|
393 |
-
|
394 |
-
## Dataset Structure
|
395 |
-
|
396 |
-
### Data Instances
|
397 |
-
|
398 |
-
[More Information Needed]
|
399 |
-
|
400 |
-
### Data Fields
|
401 |
-
|
402 |
-
[More Information Needed]
|
403 |
-
|
404 |
-
### Data Splits
|
405 |
-
|
406 |
-
[More Information Needed]
|
407 |
-
|
408 |
-
## Dataset Creation
|
409 |
-
|
410 |
-
### Curation Rationale
|
411 |
-
|
412 |
-
[More Information Needed]
|
413 |
-
|
414 |
-
### Source Data
|
415 |
-
|
416 |
-
#### Initial Data Collection and Normalization
|
417 |
-
|
418 |
-
[More Information Needed]
|
419 |
-
|
420 |
-
#### Who are the source language producers?
|
421 |
-
|
422 |
-
[More Information Needed]
|
423 |
-
|
424 |
-
### Annotations
|
425 |
-
|
426 |
-
#### Annotation process
|
427 |
-
|
428 |
-
[More Information Needed]
|
429 |
-
|
430 |
-
#### Who are the annotators?
|
431 |
-
|
432 |
-
[More Information Needed]
|
433 |
-
|
434 |
-
### Personal and Sensitive Information
|
435 |
-
|
436 |
-
[More Information Needed]
|
437 |
-
|
438 |
-
## Considerations for Using the Data
|
439 |
-
|
440 |
-
### Social Impact of Dataset
|
441 |
-
|
442 |
-
[More Information Needed]
|
443 |
-
|
444 |
-
### Discussion of Biases
|
445 |
-
|
446 |
-
[More Information Needed]
|
447 |
-
|
448 |
-
### Other Known Limitations
|
449 |
-
|
450 |
-
[More Information Needed]
|
451 |
-
|
452 |
-
## Additional Information
|
453 |
-
|
454 |
-
### Dataset Curators
|
455 |
-
|
456 |
-
[More Information Needed]
|
457 |
-
|
458 |
-
### Licensing Information
|
459 |
-
|
460 |
-
[More Information Needed]
|
461 |
-
|
462 |
-
### Citation Information
|
463 |
-
|
464 |
-
```
|
465 |
-
@inproceedings{eisele-chen-2010-multiun,
|
466 |
-
title = "{M}ulti{UN}: A Multilingual Corpus from United Nation Documents",
|
467 |
-
author = "Eisele, Andreas and
|
468 |
-
Chen, Yu",
|
469 |
-
booktitle = "Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)",
|
470 |
-
month = may,
|
471 |
-
year = "2010",
|
472 |
-
address = "Valletta, Malta",
|
473 |
-
publisher = "European Language Resources Association (ELRA)",
|
474 |
-
url = "http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf",
|
475 |
-
abstract = "This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.",
|
476 |
-
}
|
477 |
-
```
|
478 |
-
|
479 |
-
```
|
480 |
-
@InProceedings{TIEDEMANN12.463,
|
481 |
-
author = {J�rg Tiedemann},
|
482 |
-
title = {Parallel Data, Tools and Interfaces in OPUS},
|
483 |
-
booktitle = {Proceedings of the Eight International Conference on Language Resources and Evaluation (LREC'12)},
|
484 |
-
year = {2012},
|
485 |
-
month = {may},
|
486 |
-
date = {23-25},
|
487 |
-
address = {Istanbul, Turkey},
|
488 |
-
editor = {Nicoletta Calzolari (Conference Chair) and Khalid Choukri and Thierry Declerck and Mehmet Ugur Dogan and Bente Maegaard and Joseph Mariani and Jan Odijk and Stelios Piperidis},
|
489 |
-
publisher = {European Language Resources Association (ELRA)},
|
490 |
-
isbn = {978-2-9517408-7-7},
|
491 |
-
}
|
492 |
-
```
|
493 |
-
### Contributions
|
494 |
-
|
495 |
-
Thanks to [@patil-suraj](https://github.com/patil-suraj) for adding this dataset.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
dataset_infos.json
DELETED
@@ -1 +0,0 @@
|
|
1 |
-
{"ar-de": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["ar", "de"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "ar-de", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 94466397, "num_examples": 165090, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/ar-de.txt.zip": {"num_bytes": 21869935, "checksum": "d56514f8dfa71c2265fb8b98b75f07e07eaa1c8add8a07aa5146edd4c16fa2c3"}}, "download_size": 21869935, "post_processing_size": null, "dataset_size": 94466397, "size_in_bytes": 116336332}, "ar-en": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["ar", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "ar-en", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4189852369, "num_examples": 9759125, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/ar-en.txt.zip": {"num_bytes": 1036296368, "checksum": "fc69155bcb8d8dc71f3466c62e20cca1d715837f8e1242396484f49264768e1d"}}, "download_size": 1036296368, "post_processing_size": null, "dataset_size": 4189852369, "size_in_bytes": 5226148737}, "ar-es": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["ar", "es"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "ar-es", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4509675284, "num_examples": 10119379, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/ar-es.txt.zip": {"num_bytes": 1101206667, "checksum": "ffe20bfc960550a2b98b7259f7483ad0825300ec28805117d08c41c039ae23e7"}}, "download_size": 1101206667, "post_processing_size": null, "dataset_size": 4509675284, "size_in_bytes": 5610881951}, "ar-fr": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["ar", "fr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "ar-fr", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4516850009, "num_examples": 9929567, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/ar-fr.txt.zip": {"num_bytes": 1109705925, "checksum": "ff2d7327be095d92ecf65f143f108269e3766a91ceb911dafee5de212a692bbd"}}, "download_size": 1109705925, "post_processing_size": null, "dataset_size": 4516850009, "size_in_bytes": 5626555934}, "ar-ru": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["ar", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "ar-ru", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 5932866867, "num_examples": 10206243, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/ar-ru.txt.zip": {"num_bytes": 1261123878, "checksum": "eca37e8e1fe1da067be3c3b085d5cd134cce4bfb49c6c7d9fcb262e503109a47"}}, "download_size": 1261123878, "post_processing_size": null, "dataset_size": 5932866867, "size_in_bytes": 7193990745}, "ar-zh": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["ar", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "ar-zh", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3781658413, "num_examples": 9832293, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/ar-zh.txt.zip": {"num_bytes": 1009696775, "checksum": "5c4f4f02f3d69c0fc87e4c0e1c2bbb6dffcd65303051c91a39f396643671eafb"}}, "download_size": 1009696775, "post_processing_size": null, "dataset_size": 3781658413, "size_in_bytes": 4791355188}, "de-en": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["de", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "de-en", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 76684549, "num_examples": 162981, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/de-en.txt.zip": {"num_bytes": 19468529, "checksum": "3f4f3867dc690c0b27359474c9eae9c239020ab05d74b8c1b0023ce500d4b57b"}}, "download_size": 19468529, "post_processing_size": null, "dataset_size": 76684549, "size_in_bytes": 96153078}, "de-es": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["de", "es"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "de-es", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 80936653, "num_examples": 162078, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/de-es.txt.zip": {"num_bytes": 20266591, "checksum": "87e772d6d4b3b324d2f30381a55896614ab23a93ac5d089ad2a4bcdfea20c14b"}}, "download_size": 20266591, "post_processing_size": null, "dataset_size": 80936653, "size_in_bytes": 101203244}, "de-fr": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["de", "fr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "de-fr", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 81888435, "num_examples": 164025, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/de-fr.txt.zip": {"num_bytes": 20692837, "checksum": "887eb6bcb2c003cb8bb61e9c479c546827e12185042fa5d0de6a41aa9c28f3af"}}, "download_size": 20692837, "post_processing_size": null, "dataset_size": 81888435, "size_in_bytes": 102581272}, "de-ru": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["de", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "de-ru", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 111517934, "num_examples": 164792, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/de-ru.txt.zip": {"num_bytes": 23507789, "checksum": "b78de162ff223ae026403341db72b9ef23ed8a40ae9a3d015cde1997ec9feb4b"}}, "download_size": 23507789, "post_processing_size": null, "dataset_size": 111517934, "size_in_bytes": 135025723}, "de-zh": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["de", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "de-zh", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 70534818, "num_examples": 176933, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/de-zh.txt.zip": {"num_bytes": 19927209, "checksum": "5dc2f4574343628943c523640add722c4d431bf5b1701011f6302900d95c9cc3"}}, "download_size": 19927209, "post_processing_size": null, "dataset_size": 70534818, "size_in_bytes": 90462027}, "en-es": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["en", "es"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "en-es", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4128141663, "num_examples": 11350967, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/en-es.txt.zip": {"num_bytes": 1123164180, "checksum": "078aa771bf0db8cd0e0c5ebef1a4af7c357dac4e93dea2790d3da8442bbf9d2f"}}, "download_size": 1123164180, "post_processing_size": null, "dataset_size": 4128141663, "size_in_bytes": 5251305843}, "en-fr": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["en", "fr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "en-fr", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4678055160, "num_examples": 13172019, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/en-fr.txt.zip": {"num_bytes": 1355002731, "checksum": "7c7902d88f113562442b00fc98d885ae81296b4557b75a12246421ba45fa6f67"}}, "download_size": 1355002731, "post_processing_size": null, "dataset_size": 4678055160, "size_in_bytes": 6033057891}, "en-ru": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["en", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "en-ru", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 5632662839, "num_examples": 11654416, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/en-ru.txt.zip": {"num_bytes": 1285801078, "checksum": "641468670b0bcc254bc7e92a185bf56747eb855b8ab5150606a2e12efa61cbe6"}}, "download_size": 1285801078, "post_processing_size": null, "dataset_size": 5632662839, "size_in_bytes": 6918463917}, "en-zh": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["en", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "en-zh", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2960376046, "num_examples": 9564315, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/en-zh.txt.zip": {"num_bytes": 900076520, "checksum": "3cb1ff6c85cc0232715c4a4ecdb05115d05a400c4f3ec8aeabbeaaaf33c9c77d"}}, "download_size": 900076520, "post_processing_size": null, "dataset_size": 2960376046, "size_in_bytes": 3860452566}, "es-fr": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["es", "fr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "es-fr", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4454712498, "num_examples": 11441889, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/es-fr.txt.zip": {"num_bytes": 1195733510, "checksum": "756c3d6f9a4dbb2f95cb4fe68f26668a816cd8371e656cf6acb6345de53d22fb"}}, "download_size": 1195733510, "post_processing_size": null, "dataset_size": 4454712498, "size_in_bytes": 5650446008}, "es-ru": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["es", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "es-ru", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 5442655730, "num_examples": 10605056, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/es-ru.txt.zip": {"num_bytes": 1228045966, "checksum": "7b5c7ef88cdaebd92426f619fde4ecd5f6f3d72084a0c12fdd07f7f108c189a1"}}, "download_size": 1228045966, "post_processing_size": null, "dataset_size": 5442655730, "size_in_bytes": 6670701696}, "es-zh": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["es", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "es-zh", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3223871198, "num_examples": 9847770, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/es-zh.txt.zip": {"num_bytes": 953250084, "checksum": "6095851f7818b48936f1e86ea02072830d337ee02961366476af1bec23dc6d28"}}, "download_size": 953250084, "post_processing_size": null, "dataset_size": 3223871198, "size_in_bytes": 4177121282}, "fr-ru": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["fr", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "fr-ru", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 5979879089, "num_examples": 11761738, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/fr-ru.txt.zip": {"num_bytes": 1364307157, "checksum": "ecc5a2aeeef1924a70a218d370182ad233eaa916e5d7f5db386b1478ea9a642d"}}, "download_size": 1364307157, "post_processing_size": null, "dataset_size": 5979879089, "size_in_bytes": 7344186246}, "fr-zh": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["fr", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "fr-zh", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3241098333, "num_examples": 9690914, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/fr-zh.txt.zip": {"num_bytes": 962824881, "checksum": "51b2c15c19caabec8853747c57bee8951b376faea23652bc73abda05e1e9314a"}}, "download_size": 962824881, "post_processing_size": null, "dataset_size": 3241098333, "size_in_bytes": 4203923214}, "ru-zh": {"description": "This is a collection of translated documents from the United Nations. This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language\n", "citation": "@inproceedings{eisele-chen-2010-multiun,\n title = \"{M}ulti{UN}: A Multilingual Corpus from United Nation Documents\",\n author = \"Eisele, Andreas and\n Chen, Yu\",\n booktitle = \"Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)\",\n month = may,\n year = \"2010\",\n address = \"Valletta, Malta\",\n publisher = \"European Language Resources Association (ELRA)\",\n url = \"http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf\",\n abstract = \"This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.\",\n}\n\nJ. Tiedemann, 2012, Parallel Data, Tools and Interfaces in OPUS. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012)\n", "homepage": "http://www.euromatrixplus.net/multi-un/", "license": "", "features": {"translation": {"languages": ["ru", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "un_multi", "config_name": "ru-zh", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4233875537, "num_examples": 9557007, "dataset_name": "un_multi"}}, "download_checksums": {"http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses/ru-zh.txt.zip": {"num_bytes": 1037881127, "checksum": "c49a35cd9c698af1a0be7223604da661ab3805e4d3389aac41823d3c28ea0593"}}, "download_size": 1037881127, "post_processing_size": null, "dataset_size": 4233875537, "size_in_bytes": 5271756664}}
|
|
|
|
de-zh/un_multi-train.parquet
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8199e073d924d1bb0a44eac04ad2c648fb7b7aae6eb8df1dff0e9dc3962b20c6
|
3 |
+
size 34964647
|
un_multi.py
DELETED
@@ -1,109 +0,0 @@
|
|
1 |
-
# coding=utf-8
|
2 |
-
# Copyright 2020 The HuggingFace Datasets Authors.
|
3 |
-
#
|
4 |
-
# Licensed under the Apache License, Version 2.0 (the "License");
|
5 |
-
# you may not use this file except in compliance with the License.
|
6 |
-
# You may obtain a copy of the License at
|
7 |
-
#
|
8 |
-
# http://www.apache.org/licenses/LICENSE-2.0
|
9 |
-
#
|
10 |
-
# Unless required by applicable law or agreed to in writing, software
|
11 |
-
# distributed under the License is distributed on an "AS IS" BASIS,
|
12 |
-
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
13 |
-
# See the License for the specific language governing permissions and
|
14 |
-
# limitations under the License.
|
15 |
-
"""MultiUN: Multilingual UN Parallel Text 2000—2009"""
|
16 |
-
|
17 |
-
|
18 |
-
import itertools
|
19 |
-
import os
|
20 |
-
|
21 |
-
import datasets
|
22 |
-
|
23 |
-
|
24 |
-
_CITATION = """\
|
25 |
-
@inproceedings{eisele-chen-2010-multiun,
|
26 |
-
title = "{M}ulti{UN}: A Multilingual Corpus from United Nation Documents",
|
27 |
-
author = "Eisele, Andreas and
|
28 |
-
Chen, Yu",
|
29 |
-
booktitle = "Proceedings of the Seventh International Conference on Language Resources and Evaluation ({LREC}'10)",
|
30 |
-
month = may,
|
31 |
-
year = "2010",
|
32 |
-
address = "Valletta, Malta",
|
33 |
-
publisher = "European Language Resources Association (ELRA)",
|
34 |
-
url = "http://www.lrec-conf.org/proceedings/lrec2010/pdf/686_Paper.pdf",
|
35 |
-
abstract = "This paper describes the acquisition, preparation and properties of a corpus extracted from the official documents of the United Nations (UN). This corpus is available in all 6 official languages of the UN, consisting of around 300 million words per language. We describe the methods we used for crawling, document formatting, and sentence alignment. This corpus also includes a common test set for machine translation. We present the results of a French-Chinese machine translation experiment performed on this corpus.",
|
36 |
-
}
|
37 |
-
|
38 |
-
@InProceedings{TIEDEMANN12.463,
|
39 |
-
author = {J�rg Tiedemann},
|
40 |
-
title = {Parallel Data, Tools and Interfaces in OPUS},
|
41 |
-
booktitle = {Proceedings of the Eight International Conference on Language Resources and Evaluation (LREC'12)},
|
42 |
-
year = {2012},
|
43 |
-
month = {may},
|
44 |
-
date = {23-25},
|
45 |
-
address = {Istanbul, Turkey},
|
46 |
-
editor = {Nicoletta Calzolari (Conference Chair) and Khalid Choukri and Thierry Declerck and Mehmet Ugur Dogan and Bente Maegaard and Joseph Mariani and Jan Odijk and Stelios Piperidis},
|
47 |
-
publisher = {European Language Resources Association (ELRA)},
|
48 |
-
isbn = {978-2-9517408-7-7},
|
49 |
-
}
|
50 |
-
"""
|
51 |
-
|
52 |
-
|
53 |
-
_DESCRIPTION = """\
|
54 |
-
This is a collection of translated documents from the United Nations. \
|
55 |
-
This corpus is available in all 6 official languages of the UN, \
|
56 |
-
consisting of around 300 million words per language
|
57 |
-
"""
|
58 |
-
|
59 |
-
|
60 |
-
_HOMEPAGE = "http://www.euromatrixplus.net/multi-un/"
|
61 |
-
|
62 |
-
_LANGUAGES = ["ar", "de", "en", "es", "fr", "ru", "zh"]
|
63 |
-
_LANGUAGE_PAIRS = list(itertools.combinations(_LANGUAGES, 2))
|
64 |
-
|
65 |
-
_BASE_URL = "http://opus.nlpl.eu/download.php?f=MultiUN/v1/moses"
|
66 |
-
_URLS = {f"{l1}-{l2}": f"{_BASE_URL}/{l1}-{l2}.txt.zip" for l1, l2 in _LANGUAGE_PAIRS}
|
67 |
-
|
68 |
-
|
69 |
-
class UnMulti(datasets.GeneratorBasedBuilder):
|
70 |
-
"""MultiUN: Multilingual UN Parallel Text 2000—2009"""
|
71 |
-
|
72 |
-
VERSION = datasets.Version("1.0.0")
|
73 |
-
|
74 |
-
BUILDER_CONFIGS = [
|
75 |
-
datasets.BuilderConfig(name=f"{l1}-{l2}", version=datasets.Version("1.0.0"), description=f"MultiUN {l1}-{l2}")
|
76 |
-
for l1, l2 in _LANGUAGE_PAIRS
|
77 |
-
]
|
78 |
-
|
79 |
-
def _info(self):
|
80 |
-
return datasets.DatasetInfo(
|
81 |
-
description=_DESCRIPTION,
|
82 |
-
features=datasets.Features(
|
83 |
-
{"translation": datasets.features.Translation(languages=tuple(self.config.name.split("-")))}
|
84 |
-
),
|
85 |
-
supervised_keys=None,
|
86 |
-
homepage=_HOMEPAGE,
|
87 |
-
citation=_CITATION,
|
88 |
-
)
|
89 |
-
|
90 |
-
def _split_generators(self, dl_manager):
|
91 |
-
"""Returns SplitGenerators."""
|
92 |
-
lang_pair = self.config.name.split("-")
|
93 |
-
data_dir = dl_manager.download_and_extract(_URLS[self.config.name])
|
94 |
-
return [
|
95 |
-
datasets.SplitGenerator(
|
96 |
-
name=datasets.Split.TRAIN,
|
97 |
-
gen_kwargs={
|
98 |
-
"source_file": os.path.join(data_dir, f"MultiUN.{self.config.name}.{lang_pair[0]}"),
|
99 |
-
"target_file": os.path.join(data_dir, f"MultiUN.{self.config.name}.{lang_pair[1]}"),
|
100 |
-
},
|
101 |
-
),
|
102 |
-
]
|
103 |
-
|
104 |
-
def _generate_examples(self, source_file, target_file):
|
105 |
-
source, target = tuple(self.config.name.split("-"))
|
106 |
-
with open(source_file, encoding="utf-8") as src_f, open(target_file, encoding="utf-8") as tgt_f:
|
107 |
-
for idx, (l1, l2) in enumerate(zip(src_f, tgt_f)):
|
108 |
-
result = {"translation": {source: l1.strip(), target: l2.strip()}}
|
109 |
-
yield idx, result
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|