Datasets:

License:
davzoku commited on
Commit
664f830
1 Parent(s): 784097b

Convert dataset to Parquet

Browse files
README.md CHANGED
@@ -109,12 +109,6 @@ language:
109
  - uz
110
  - vi
111
  - zh
112
- language_bcp47:
113
- - art-x-bork
114
- - fr-CA
115
- - pt-BR
116
- - zh-CN
117
- - zh-TW
118
  license:
119
  - cc-by-nc-nd-4.0
120
  multilinguality:
@@ -127,9 +121,72 @@ source_datasets:
127
  task_categories:
128
  - translation
129
  task_ids: []
130
- paperswithcode_id: null
131
  pretty_name: Web Inventory of Transcribed & Translated (WIT) Ted Talks
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
132
  dataset_info:
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
133
  - config_name: eu_ca_2014
134
  features:
135
  - name: translation
@@ -172,6 +229,48 @@ dataset_info:
172
  num_examples: 54
173
  download_size: 1666674366
174
  dataset_size: 19506
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
175
  - config_name: nl_en_2014
176
  features:
177
  - name: translation
@@ -256,106 +355,11 @@ dataset_info:
256
  num_examples: 496
257
  download_size: 1666674366
258
  dataset_size: 310922
 
259
  - config_name: de_ja_2014
260
- features:
261
- - name: translation
262
- dtype:
263
- translation:
264
- languages:
265
- - de
266
- - ja
267
- splits:
268
- - name: train
269
- num_bytes: 1074403
270
- num_examples: 2536
271
- download_size: 1666674366
272
- dataset_size: 1074403
273
- - config_name: de_ja_2015
274
- features:
275
- - name: translation
276
- dtype:
277
- translation:
278
- languages:
279
- - de
280
- - ja
281
- splits:
282
- - name: train
283
- num_bytes: 1442047
284
- num_examples: 3247
285
- download_size: 1666674366
286
- dataset_size: 1442047
287
- - config_name: de_ja_2016
288
- features:
289
- - name: translation
290
- dtype:
291
- translation:
292
- languages:
293
- - de
294
- - ja
295
- splits:
296
- - name: train
297
- num_bytes: 1630729
298
- num_examples: 3590
299
- download_size: 1666674366
300
- dataset_size: 1630729
301
- - config_name: fr-ca_hi_2014
302
- features:
303
- - name: translation
304
- dtype:
305
- translation:
306
- languages:
307
- - fr-ca
308
- - hi
309
- splits:
310
- - name: train
311
- num_bytes: 74472
312
- num_examples: 127
313
- download_size: 1666674366
314
- dataset_size: 74472
315
- - config_name: fr-ca_hi_2015
316
- features:
317
- - name: translation
318
- dtype:
319
- translation:
320
- languages:
321
- - fr-ca
322
- - hi
323
- splits:
324
- - name: train
325
- num_bytes: 82448
326
- num_examples: 141
327
- download_size: 1666674366
328
- dataset_size: 82448
329
- - config_name: fr-ca_hi_2016
330
- features:
331
- - name: translation
332
- dtype:
333
- translation:
334
- languages:
335
- - fr-ca
336
- - hi
337
- splits:
338
- - name: train
339
- num_bytes: 93425
340
- num_examples: 156
341
- download_size: 1666674366
342
- dataset_size: 93425
343
- config_names:
344
- - de_ja_2014
345
- - de_ja_2015
346
- - de_ja_2016
347
- - eu_ca_2014
348
- - eu_ca_2015
349
- - eu_ca_2016
350
- - fr-ca_hi_2014
351
- - fr-ca_hi_2015
352
- - fr-ca_hi_2016
353
- - nl_en_2014
354
- - nl_en_2015
355
- - nl_en_2016
356
- - nl_hi_2014
357
- - nl_hi_2015
358
- - nl_hi_2016
359
  ---
360
 
361
  # Dataset Card for Web Inventory of Transcribed & Translated(WIT) Ted Talks
 
109
  - uz
110
  - vi
111
  - zh
 
 
 
 
 
 
112
  license:
113
  - cc-by-nc-nd-4.0
114
  multilinguality:
 
121
  task_categories:
122
  - translation
123
  task_ids: []
 
124
  pretty_name: Web Inventory of Transcribed & Translated (WIT) Ted Talks
125
+ config_names:
126
+ - de_ja_2014
127
+ - de_ja_2015
128
+ - de_ja_2016
129
+ - eu_ca_2014
130
+ - eu_ca_2015
131
+ - eu_ca_2016
132
+ - fr-ca_hi_2014
133
+ - fr-ca_hi_2015
134
+ - fr-ca_hi_2016
135
+ - nl_en_2014
136
+ - nl_en_2015
137
+ - nl_en_2016
138
+ - nl_hi_2014
139
+ - nl_hi_2015
140
+ - nl_hi_2016
141
+ language_bcp47:
142
+ - art-x-bork
143
+ - fr-CA
144
+ - pt-BR
145
+ - zh-CN
146
+ - zh-TW
147
  dataset_info:
148
+ - config_name: de_ja_2014
149
+ features:
150
+ - name: translation
151
+ dtype:
152
+ translation:
153
+ languages:
154
+ - de
155
+ - ja
156
+ splits:
157
+ - name: train
158
+ num_bytes: 1074403
159
+ num_examples: 2536
160
+ download_size: 687561
161
+ dataset_size: 1074403
162
+ - config_name: de_ja_2015
163
+ features:
164
+ - name: translation
165
+ dtype:
166
+ translation:
167
+ languages:
168
+ - de
169
+ - ja
170
+ splits:
171
+ - name: train
172
+ num_bytes: 1442047
173
+ num_examples: 3247
174
+ download_size: 1666674366
175
+ dataset_size: 1442047
176
+ - config_name: de_ja_2016
177
+ features:
178
+ - name: translation
179
+ dtype:
180
+ translation:
181
+ languages:
182
+ - de
183
+ - ja
184
+ splits:
185
+ - name: train
186
+ num_bytes: 1630729
187
+ num_examples: 3590
188
+ download_size: 1666674366
189
+ dataset_size: 1630729
190
  - config_name: eu_ca_2014
191
  features:
192
  - name: translation
 
229
  num_examples: 54
230
  download_size: 1666674366
231
  dataset_size: 19506
232
+ - config_name: fr-ca_hi_2014
233
+ features:
234
+ - name: translation
235
+ dtype:
236
+ translation:
237
+ languages:
238
+ - fr-ca
239
+ - hi
240
+ splits:
241
+ - name: train
242
+ num_bytes: 74472
243
+ num_examples: 127
244
+ download_size: 1666674366
245
+ dataset_size: 74472
246
+ - config_name: fr-ca_hi_2015
247
+ features:
248
+ - name: translation
249
+ dtype:
250
+ translation:
251
+ languages:
252
+ - fr-ca
253
+ - hi
254
+ splits:
255
+ - name: train
256
+ num_bytes: 82448
257
+ num_examples: 141
258
+ download_size: 1666674366
259
+ dataset_size: 82448
260
+ - config_name: fr-ca_hi_2016
261
+ features:
262
+ - name: translation
263
+ dtype:
264
+ translation:
265
+ languages:
266
+ - fr-ca
267
+ - hi
268
+ splits:
269
+ - name: train
270
+ num_bytes: 93425
271
+ num_examples: 156
272
+ download_size: 1666674366
273
+ dataset_size: 93425
274
  - config_name: nl_en_2014
275
  features:
276
  - name: translation
 
355
  num_examples: 496
356
  download_size: 1666674366
357
  dataset_size: 310922
358
+ configs:
359
  - config_name: de_ja_2014
360
+ data_files:
361
+ - split: train
362
+ path: de_ja_2014/train-*
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
363
  ---
364
 
365
  # Dataset Card for Web Inventory of Transcribed & Translated(WIT) Ted Talks
dataset_infos.json CHANGED
@@ -1 +1,667 @@
1
- {"eu_ca_2014": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["eu", "ca"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "eu_ca_2014", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 15200, "num_examples": 44, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 15200, "size_in_bytes": 1666689566}, "eu_ca_2015": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["eu", "ca"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "eu_ca_2015", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 18776, "num_examples": 52, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 18776, "size_in_bytes": 1666693142}, "eu_ca_2016": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["eu", "ca"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "eu_ca_2016", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 19514, "num_examples": 54, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 19514, "size_in_bytes": 1666693880}, "nl_en_2014": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["nl", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "nl_en_2014", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1035553, "num_examples": 2966, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 1035553, "size_in_bytes": 1667709919}, "nl_en_2015": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["nl", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "nl_en_2015", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1292618, "num_examples": 3550, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 1292618, "size_in_bytes": 1667966984}, "nl_en_2016": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["nl", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "nl_en_2016", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1434215, "num_examples": 3852, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 1434215, "size_in_bytes": 1668108581}, "nl_hi_2014": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["nl", "hi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "nl_hi_2014", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 214878, "num_examples": 367, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 214878, "size_in_bytes": 1666889244}, "nl_hi_2015": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["nl", "hi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "nl_hi_2015", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 252200, "num_examples": 421, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 252200, "size_in_bytes": 1666926566}, "nl_hi_2016": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["nl", "hi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "nl_hi_2016", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 310930, "num_examples": 496, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 310930, "size_in_bytes": 1666985296}, "de_ja_2014": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["de", "ja"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "de_ja_2014", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1074411, "num_examples": 2536, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 1074411, "size_in_bytes": 1667748777}, "de_ja_2015": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["de", "ja"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "de_ja_2015", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1442055, "num_examples": 3247, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 1442055, "size_in_bytes": 1668116421}, "de_ja_2016": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["de", "ja"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "de_ja_2016", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1630737, "num_examples": 3590, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 1630737, "size_in_bytes": 1668305103}, "fr-ca_hi_2014": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["fr-ca", "hi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "fr-ca_hi_2014", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 74480, "num_examples": 127, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 74480, "size_in_bytes": 1666748846}, "fr-ca_hi_2015": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["fr-ca", "hi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "fr-ca_hi_2015", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 82456, "num_examples": 141, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 82456, "size_in_bytes": 1666756822}, "fr-ca_hi_2016": {"description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n", "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n", "homepage": "https://wit3.fbk.eu/", "license": "CC-BY-NC-4.0", "features": {"translation": {"languages": ["fr-ca", "hi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": null, "builder_name": "ted_talks_iwslt", "config_name": "fr-ca_hi_2016", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 93433, "num_examples": 156, "dataset_name": "ted_talks_iwslt"}}, "download_checksums": {"https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {"num_bytes": 1666674366, "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"}}, "download_size": 1666674366, "post_processing_size": null, "dataset_size": 93433, "size_in_bytes": 1666767799}}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "eu_ca_2014": {
3
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
4
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
5
+ "homepage": "https://wit3.fbk.eu/",
6
+ "license": "CC-BY-NC-4.0",
7
+ "features": {
8
+ "translation": {
9
+ "languages": [
10
+ "eu",
11
+ "ca"
12
+ ],
13
+ "id": null,
14
+ "_type": "Translation"
15
+ }
16
+ },
17
+ "post_processed": null,
18
+ "supervised_keys": null,
19
+ "builder_name": "ted_talks_iwslt",
20
+ "config_name": "eu_ca_2014",
21
+ "version": {
22
+ "version_str": "1.1.0",
23
+ "description": "",
24
+ "major": 1,
25
+ "minor": 1,
26
+ "patch": 0
27
+ },
28
+ "splits": {
29
+ "train": {
30
+ "name": "train",
31
+ "num_bytes": 15200,
32
+ "num_examples": 44,
33
+ "dataset_name": "ted_talks_iwslt"
34
+ }
35
+ },
36
+ "download_checksums": {
37
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
38
+ "num_bytes": 1666674366,
39
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
40
+ }
41
+ },
42
+ "download_size": 1666674366,
43
+ "post_processing_size": null,
44
+ "dataset_size": 15200,
45
+ "size_in_bytes": 1666689566
46
+ },
47
+ "eu_ca_2015": {
48
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
49
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
50
+ "homepage": "https://wit3.fbk.eu/",
51
+ "license": "CC-BY-NC-4.0",
52
+ "features": {
53
+ "translation": {
54
+ "languages": [
55
+ "eu",
56
+ "ca"
57
+ ],
58
+ "id": null,
59
+ "_type": "Translation"
60
+ }
61
+ },
62
+ "post_processed": null,
63
+ "supervised_keys": null,
64
+ "builder_name": "ted_talks_iwslt",
65
+ "config_name": "eu_ca_2015",
66
+ "version": {
67
+ "version_str": "1.1.0",
68
+ "description": "",
69
+ "major": 1,
70
+ "minor": 1,
71
+ "patch": 0
72
+ },
73
+ "splits": {
74
+ "train": {
75
+ "name": "train",
76
+ "num_bytes": 18776,
77
+ "num_examples": 52,
78
+ "dataset_name": "ted_talks_iwslt"
79
+ }
80
+ },
81
+ "download_checksums": {
82
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
83
+ "num_bytes": 1666674366,
84
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
85
+ }
86
+ },
87
+ "download_size": 1666674366,
88
+ "post_processing_size": null,
89
+ "dataset_size": 18776,
90
+ "size_in_bytes": 1666693142
91
+ },
92
+ "eu_ca_2016": {
93
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
94
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
95
+ "homepage": "https://wit3.fbk.eu/",
96
+ "license": "CC-BY-NC-4.0",
97
+ "features": {
98
+ "translation": {
99
+ "languages": [
100
+ "eu",
101
+ "ca"
102
+ ],
103
+ "id": null,
104
+ "_type": "Translation"
105
+ }
106
+ },
107
+ "post_processed": null,
108
+ "supervised_keys": null,
109
+ "builder_name": "ted_talks_iwslt",
110
+ "config_name": "eu_ca_2016",
111
+ "version": {
112
+ "version_str": "1.1.0",
113
+ "description": "",
114
+ "major": 1,
115
+ "minor": 1,
116
+ "patch": 0
117
+ },
118
+ "splits": {
119
+ "train": {
120
+ "name": "train",
121
+ "num_bytes": 19514,
122
+ "num_examples": 54,
123
+ "dataset_name": "ted_talks_iwslt"
124
+ }
125
+ },
126
+ "download_checksums": {
127
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
128
+ "num_bytes": 1666674366,
129
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
130
+ }
131
+ },
132
+ "download_size": 1666674366,
133
+ "post_processing_size": null,
134
+ "dataset_size": 19514,
135
+ "size_in_bytes": 1666693880
136
+ },
137
+ "nl_en_2014": {
138
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
139
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
140
+ "homepage": "https://wit3.fbk.eu/",
141
+ "license": "CC-BY-NC-4.0",
142
+ "features": {
143
+ "translation": {
144
+ "languages": [
145
+ "nl",
146
+ "en"
147
+ ],
148
+ "id": null,
149
+ "_type": "Translation"
150
+ }
151
+ },
152
+ "post_processed": null,
153
+ "supervised_keys": null,
154
+ "builder_name": "ted_talks_iwslt",
155
+ "config_name": "nl_en_2014",
156
+ "version": {
157
+ "version_str": "1.1.0",
158
+ "description": "",
159
+ "major": 1,
160
+ "minor": 1,
161
+ "patch": 0
162
+ },
163
+ "splits": {
164
+ "train": {
165
+ "name": "train",
166
+ "num_bytes": 1035553,
167
+ "num_examples": 2966,
168
+ "dataset_name": "ted_talks_iwslt"
169
+ }
170
+ },
171
+ "download_checksums": {
172
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
173
+ "num_bytes": 1666674366,
174
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
175
+ }
176
+ },
177
+ "download_size": 1666674366,
178
+ "post_processing_size": null,
179
+ "dataset_size": 1035553,
180
+ "size_in_bytes": 1667709919
181
+ },
182
+ "nl_en_2015": {
183
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
184
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
185
+ "homepage": "https://wit3.fbk.eu/",
186
+ "license": "CC-BY-NC-4.0",
187
+ "features": {
188
+ "translation": {
189
+ "languages": [
190
+ "nl",
191
+ "en"
192
+ ],
193
+ "id": null,
194
+ "_type": "Translation"
195
+ }
196
+ },
197
+ "post_processed": null,
198
+ "supervised_keys": null,
199
+ "builder_name": "ted_talks_iwslt",
200
+ "config_name": "nl_en_2015",
201
+ "version": {
202
+ "version_str": "1.1.0",
203
+ "description": "",
204
+ "major": 1,
205
+ "minor": 1,
206
+ "patch": 0
207
+ },
208
+ "splits": {
209
+ "train": {
210
+ "name": "train",
211
+ "num_bytes": 1292618,
212
+ "num_examples": 3550,
213
+ "dataset_name": "ted_talks_iwslt"
214
+ }
215
+ },
216
+ "download_checksums": {
217
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
218
+ "num_bytes": 1666674366,
219
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
220
+ }
221
+ },
222
+ "download_size": 1666674366,
223
+ "post_processing_size": null,
224
+ "dataset_size": 1292618,
225
+ "size_in_bytes": 1667966984
226
+ },
227
+ "nl_en_2016": {
228
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
229
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
230
+ "homepage": "https://wit3.fbk.eu/",
231
+ "license": "CC-BY-NC-4.0",
232
+ "features": {
233
+ "translation": {
234
+ "languages": [
235
+ "nl",
236
+ "en"
237
+ ],
238
+ "id": null,
239
+ "_type": "Translation"
240
+ }
241
+ },
242
+ "post_processed": null,
243
+ "supervised_keys": null,
244
+ "builder_name": "ted_talks_iwslt",
245
+ "config_name": "nl_en_2016",
246
+ "version": {
247
+ "version_str": "1.1.0",
248
+ "description": "",
249
+ "major": 1,
250
+ "minor": 1,
251
+ "patch": 0
252
+ },
253
+ "splits": {
254
+ "train": {
255
+ "name": "train",
256
+ "num_bytes": 1434215,
257
+ "num_examples": 3852,
258
+ "dataset_name": "ted_talks_iwslt"
259
+ }
260
+ },
261
+ "download_checksums": {
262
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
263
+ "num_bytes": 1666674366,
264
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
265
+ }
266
+ },
267
+ "download_size": 1666674366,
268
+ "post_processing_size": null,
269
+ "dataset_size": 1434215,
270
+ "size_in_bytes": 1668108581
271
+ },
272
+ "nl_hi_2014": {
273
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
274
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
275
+ "homepage": "https://wit3.fbk.eu/",
276
+ "license": "CC-BY-NC-4.0",
277
+ "features": {
278
+ "translation": {
279
+ "languages": [
280
+ "nl",
281
+ "hi"
282
+ ],
283
+ "id": null,
284
+ "_type": "Translation"
285
+ }
286
+ },
287
+ "post_processed": null,
288
+ "supervised_keys": null,
289
+ "builder_name": "ted_talks_iwslt",
290
+ "config_name": "nl_hi_2014",
291
+ "version": {
292
+ "version_str": "1.1.0",
293
+ "description": "",
294
+ "major": 1,
295
+ "minor": 1,
296
+ "patch": 0
297
+ },
298
+ "splits": {
299
+ "train": {
300
+ "name": "train",
301
+ "num_bytes": 214878,
302
+ "num_examples": 367,
303
+ "dataset_name": "ted_talks_iwslt"
304
+ }
305
+ },
306
+ "download_checksums": {
307
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
308
+ "num_bytes": 1666674366,
309
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
310
+ }
311
+ },
312
+ "download_size": 1666674366,
313
+ "post_processing_size": null,
314
+ "dataset_size": 214878,
315
+ "size_in_bytes": 1666889244
316
+ },
317
+ "nl_hi_2015": {
318
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
319
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
320
+ "homepage": "https://wit3.fbk.eu/",
321
+ "license": "CC-BY-NC-4.0",
322
+ "features": {
323
+ "translation": {
324
+ "languages": [
325
+ "nl",
326
+ "hi"
327
+ ],
328
+ "id": null,
329
+ "_type": "Translation"
330
+ }
331
+ },
332
+ "post_processed": null,
333
+ "supervised_keys": null,
334
+ "builder_name": "ted_talks_iwslt",
335
+ "config_name": "nl_hi_2015",
336
+ "version": {
337
+ "version_str": "1.1.0",
338
+ "description": "",
339
+ "major": 1,
340
+ "minor": 1,
341
+ "patch": 0
342
+ },
343
+ "splits": {
344
+ "train": {
345
+ "name": "train",
346
+ "num_bytes": 252200,
347
+ "num_examples": 421,
348
+ "dataset_name": "ted_talks_iwslt"
349
+ }
350
+ },
351
+ "download_checksums": {
352
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
353
+ "num_bytes": 1666674366,
354
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
355
+ }
356
+ },
357
+ "download_size": 1666674366,
358
+ "post_processing_size": null,
359
+ "dataset_size": 252200,
360
+ "size_in_bytes": 1666926566
361
+ },
362
+ "nl_hi_2016": {
363
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
364
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
365
+ "homepage": "https://wit3.fbk.eu/",
366
+ "license": "CC-BY-NC-4.0",
367
+ "features": {
368
+ "translation": {
369
+ "languages": [
370
+ "nl",
371
+ "hi"
372
+ ],
373
+ "id": null,
374
+ "_type": "Translation"
375
+ }
376
+ },
377
+ "post_processed": null,
378
+ "supervised_keys": null,
379
+ "builder_name": "ted_talks_iwslt",
380
+ "config_name": "nl_hi_2016",
381
+ "version": {
382
+ "version_str": "1.1.0",
383
+ "description": "",
384
+ "major": 1,
385
+ "minor": 1,
386
+ "patch": 0
387
+ },
388
+ "splits": {
389
+ "train": {
390
+ "name": "train",
391
+ "num_bytes": 310930,
392
+ "num_examples": 496,
393
+ "dataset_name": "ted_talks_iwslt"
394
+ }
395
+ },
396
+ "download_checksums": {
397
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
398
+ "num_bytes": 1666674366,
399
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
400
+ }
401
+ },
402
+ "download_size": 1666674366,
403
+ "post_processing_size": null,
404
+ "dataset_size": 310930,
405
+ "size_in_bytes": 1666985296
406
+ },
407
+ "de_ja_2014": {
408
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
409
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
410
+ "homepage": "https://wit3.fbk.eu/",
411
+ "license": "CC-BY-NC-4.0",
412
+ "features": {
413
+ "translation": {
414
+ "languages": [
415
+ "de",
416
+ "ja"
417
+ ],
418
+ "_type": "Translation"
419
+ }
420
+ },
421
+ "builder_name": "parquet",
422
+ "dataset_name": "ted_talks_iwslt",
423
+ "config_name": "de_ja_2014",
424
+ "version": {
425
+ "version_str": "1.1.0",
426
+ "major": 1,
427
+ "minor": 1,
428
+ "patch": 0
429
+ },
430
+ "splits": {
431
+ "train": {
432
+ "name": "train",
433
+ "num_bytes": 1074403,
434
+ "num_examples": 2536,
435
+ "dataset_name": null
436
+ }
437
+ },
438
+ "download_size": 687561,
439
+ "dataset_size": 1074403,
440
+ "size_in_bytes": 1761964
441
+ },
442
+ "de_ja_2015": {
443
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
444
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
445
+ "homepage": "https://wit3.fbk.eu/",
446
+ "license": "CC-BY-NC-4.0",
447
+ "features": {
448
+ "translation": {
449
+ "languages": [
450
+ "de",
451
+ "ja"
452
+ ],
453
+ "id": null,
454
+ "_type": "Translation"
455
+ }
456
+ },
457
+ "post_processed": null,
458
+ "supervised_keys": null,
459
+ "builder_name": "ted_talks_iwslt",
460
+ "config_name": "de_ja_2015",
461
+ "version": {
462
+ "version_str": "1.1.0",
463
+ "description": "",
464
+ "major": 1,
465
+ "minor": 1,
466
+ "patch": 0
467
+ },
468
+ "splits": {
469
+ "train": {
470
+ "name": "train",
471
+ "num_bytes": 1442055,
472
+ "num_examples": 3247,
473
+ "dataset_name": "ted_talks_iwslt"
474
+ }
475
+ },
476
+ "download_checksums": {
477
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
478
+ "num_bytes": 1666674366,
479
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
480
+ }
481
+ },
482
+ "download_size": 1666674366,
483
+ "post_processing_size": null,
484
+ "dataset_size": 1442055,
485
+ "size_in_bytes": 1668116421
486
+ },
487
+ "de_ja_2016": {
488
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
489
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
490
+ "homepage": "https://wit3.fbk.eu/",
491
+ "license": "CC-BY-NC-4.0",
492
+ "features": {
493
+ "translation": {
494
+ "languages": [
495
+ "de",
496
+ "ja"
497
+ ],
498
+ "id": null,
499
+ "_type": "Translation"
500
+ }
501
+ },
502
+ "post_processed": null,
503
+ "supervised_keys": null,
504
+ "builder_name": "ted_talks_iwslt",
505
+ "config_name": "de_ja_2016",
506
+ "version": {
507
+ "version_str": "1.1.0",
508
+ "description": "",
509
+ "major": 1,
510
+ "minor": 1,
511
+ "patch": 0
512
+ },
513
+ "splits": {
514
+ "train": {
515
+ "name": "train",
516
+ "num_bytes": 1630737,
517
+ "num_examples": 3590,
518
+ "dataset_name": "ted_talks_iwslt"
519
+ }
520
+ },
521
+ "download_checksums": {
522
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
523
+ "num_bytes": 1666674366,
524
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
525
+ }
526
+ },
527
+ "download_size": 1666674366,
528
+ "post_processing_size": null,
529
+ "dataset_size": 1630737,
530
+ "size_in_bytes": 1668305103
531
+ },
532
+ "fr-ca_hi_2014": {
533
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
534
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
535
+ "homepage": "https://wit3.fbk.eu/",
536
+ "license": "CC-BY-NC-4.0",
537
+ "features": {
538
+ "translation": {
539
+ "languages": [
540
+ "fr-ca",
541
+ "hi"
542
+ ],
543
+ "id": null,
544
+ "_type": "Translation"
545
+ }
546
+ },
547
+ "post_processed": null,
548
+ "supervised_keys": null,
549
+ "builder_name": "ted_talks_iwslt",
550
+ "config_name": "fr-ca_hi_2014",
551
+ "version": {
552
+ "version_str": "1.1.0",
553
+ "description": "",
554
+ "major": 1,
555
+ "minor": 1,
556
+ "patch": 0
557
+ },
558
+ "splits": {
559
+ "train": {
560
+ "name": "train",
561
+ "num_bytes": 74480,
562
+ "num_examples": 127,
563
+ "dataset_name": "ted_talks_iwslt"
564
+ }
565
+ },
566
+ "download_checksums": {
567
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
568
+ "num_bytes": 1666674366,
569
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
570
+ }
571
+ },
572
+ "download_size": 1666674366,
573
+ "post_processing_size": null,
574
+ "dataset_size": 74480,
575
+ "size_in_bytes": 1666748846
576
+ },
577
+ "fr-ca_hi_2015": {
578
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
579
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
580
+ "homepage": "https://wit3.fbk.eu/",
581
+ "license": "CC-BY-NC-4.0",
582
+ "features": {
583
+ "translation": {
584
+ "languages": [
585
+ "fr-ca",
586
+ "hi"
587
+ ],
588
+ "id": null,
589
+ "_type": "Translation"
590
+ }
591
+ },
592
+ "post_processed": null,
593
+ "supervised_keys": null,
594
+ "builder_name": "ted_talks_iwslt",
595
+ "config_name": "fr-ca_hi_2015",
596
+ "version": {
597
+ "version_str": "1.1.0",
598
+ "description": "",
599
+ "major": 1,
600
+ "minor": 1,
601
+ "patch": 0
602
+ },
603
+ "splits": {
604
+ "train": {
605
+ "name": "train",
606
+ "num_bytes": 82456,
607
+ "num_examples": 141,
608
+ "dataset_name": "ted_talks_iwslt"
609
+ }
610
+ },
611
+ "download_checksums": {
612
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
613
+ "num_bytes": 1666674366,
614
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
615
+ }
616
+ },
617
+ "download_size": 1666674366,
618
+ "post_processing_size": null,
619
+ "dataset_size": 82456,
620
+ "size_in_bytes": 1666756822
621
+ },
622
+ "fr-ca_hi_2016": {
623
+ "description": "The core of WIT3 is the TED Talks corpus, that basically redistributes the original content published by the TED Conference website (http://www.ted.com). Since 2007,\nthe TED Conference, based in California, has been posting all video recordings of its talks together with subtitles in English\nand their translations in more than 80 languages. Aside from its cultural and social relevance, this content, which is published under the Creative Commons BYNC-ND license, also represents a precious\nlanguage resource for the machine translation research community, thanks to its size, variety of topics, and covered languages.\nThis effort repurposes the original content in a way which is more convenient for machine translation researchers.\n",
624
+ "citation": "@inproceedings{cettolo-etal-2012-wit3,\n title = \"{WIT}3: Web Inventory of Transcribed and Translated Talks\",\n author = \"Cettolo, Mauro and\n Girardi, Christian and\n Federico, Marcello\",\n booktitle = \"Proceedings of the 16th Annual conference of the European Association for Machine Translation\",\n month = may # \" 28{--}30\",\n year = \"2012\",\n address = \"Trento, Italy\",\n publisher = \"European Association for Machine Translation\",\n url = \"https://www.aclweb.org/anthology/2012.eamt-1.60\",\n pages = \"261--268\",\n}\n",
625
+ "homepage": "https://wit3.fbk.eu/",
626
+ "license": "CC-BY-NC-4.0",
627
+ "features": {
628
+ "translation": {
629
+ "languages": [
630
+ "fr-ca",
631
+ "hi"
632
+ ],
633
+ "id": null,
634
+ "_type": "Translation"
635
+ }
636
+ },
637
+ "post_processed": null,
638
+ "supervised_keys": null,
639
+ "builder_name": "ted_talks_iwslt",
640
+ "config_name": "fr-ca_hi_2016",
641
+ "version": {
642
+ "version_str": "1.1.0",
643
+ "description": "",
644
+ "major": 1,
645
+ "minor": 1,
646
+ "patch": 0
647
+ },
648
+ "splits": {
649
+ "train": {
650
+ "name": "train",
651
+ "num_bytes": 93433,
652
+ "num_examples": 156,
653
+ "dataset_name": "ted_talks_iwslt"
654
+ }
655
+ },
656
+ "download_checksums": {
657
+ "https://drive.google.com/u/0/uc?id=1Cz1Un9p8Xn9IpEMMrg2kXSDt0dnjxc4z&export=download": {
658
+ "num_bytes": 1666674366,
659
+ "checksum": "4f3d977f743e330d566024420ea53ca4db50649191fd350f8e9a60d070b13dff"
660
+ }
661
+ },
662
+ "download_size": 1666674366,
663
+ "post_processing_size": null,
664
+ "dataset_size": 93433,
665
+ "size_in_bytes": 1666767799
666
+ }
667
+ }
de_ja_2014/train-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3956abfbdea6148b8d10b7d667b5b32bcfa652e3f8ff5e0a7be15c693fb7295
3
+ size 687561