abhik1505040 commited on
Commit
d1b54be
1 Parent(s): 8b07084

Added necessary files

Browse files
Files changed (49) hide show
  1. .gitignore +64 -0
  2. README.md +337 -0
  3. dataset_infos.json +1 -0
  4. dummy/amharic/2.0.0/dummy_data.zip +3 -0
  5. dummy/arabic/2.0.0/dummy_data.zip +3 -0
  6. dummy/azerbaijani/2.0.0/dummy_data.zip +3 -0
  7. dummy/bengali/2.0.0/dummy_data.zip +3 -0
  8. dummy/burmese/2.0.0/dummy_data.zip +3 -0
  9. dummy/chinese_simplified/2.0.0/dummy_data.zip +3 -0
  10. dummy/chinese_traditional/2.0.0/dummy_data.zip +3 -0
  11. dummy/english/2.0.0/dummy_data.zip +3 -0
  12. dummy/french/2.0.0/dummy_data.zip +3 -0
  13. dummy/gujarati/2.0.0/dummy_data.zip +3 -0
  14. dummy/hausa/2.0.0/dummy_data.zip +3 -0
  15. dummy/hindi/2.0.0/dummy_data.zip +3 -0
  16. dummy/igbo/2.0.0/dummy_data.zip +3 -0
  17. dummy/indonesian/2.0.0/dummy_data.zip +3 -0
  18. dummy/japanese/2.0.0/dummy_data.zip +3 -0
  19. dummy/kirundi/2.0.0/dummy_data.zip +3 -0
  20. dummy/korean/2.0.0/dummy_data.zip +3 -0
  21. dummy/kyrgyz/2.0.0/dummy_data.zip +3 -0
  22. dummy/marathi/2.0.0/dummy_data.zip +3 -0
  23. dummy/nepali/2.0.0/dummy_data.zip +3 -0
  24. dummy/oromo/2.0.0/dummy_data.zip +3 -0
  25. dummy/pashto/2.0.0/dummy_data.zip +3 -0
  26. dummy/persian/2.0.0/dummy_data.zip +3 -0
  27. dummy/pidgin/2.0.0/dummy_data.zip +3 -0
  28. dummy/portuguese/2.0.0/dummy_data.zip +3 -0
  29. dummy/punjabi/2.0.0/dummy_data.zip +3 -0
  30. dummy/russian/2.0.0/dummy_data.zip +3 -0
  31. dummy/scottish_gaelic/2.0.0/dummy_data.zip +3 -0
  32. dummy/serbian_cyrillic/2.0.0/dummy_data.zip +3 -0
  33. dummy/serbian_latin/2.0.0/dummy_data.zip +3 -0
  34. dummy/sinhala/2.0.0/dummy_data.zip +3 -0
  35. dummy/somali/2.0.0/dummy_data.zip +3 -0
  36. dummy/spanish/2.0.0/dummy_data.zip +3 -0
  37. dummy/swahili/2.0.0/dummy_data.zip +3 -0
  38. dummy/tamil/2.0.0/dummy_data.zip +3 -0
  39. dummy/telugu/2.0.0/dummy_data.zip +3 -0
  40. dummy/thai/2.0.0/dummy_data.zip +3 -0
  41. dummy/tigrinya/2.0.0/dummy_data.zip +3 -0
  42. dummy/turkish/2.0.0/dummy_data.zip +3 -0
  43. dummy/ukrainian/2.0.0/dummy_data.zip +3 -0
  44. dummy/urdu/2.0.0/dummy_data.zip +3 -0
  45. dummy/uzbek/2.0.0/dummy_data.zip +3 -0
  46. dummy/vietnamese/2.0.0/dummy_data.zip +3 -0
  47. dummy/welsh/2.0.0/dummy_data.zip +3 -0
  48. dummy/yoruba/2.0.0/dummy_data.zip +3 -0
  49. xlsum.py +167 -0
.gitignore ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Locked files
2
+ *.lock
3
+ !dvc.lock
4
+
5
+ # Extracted dummy data
6
+ datasets/**/dummy_data-zip-extracted/
7
+
8
+ # Compiled python modules.
9
+ *.pyc
10
+
11
+ # Byte-compiled
12
+ _pycache__/
13
+ .cache/
14
+
15
+ # Python egg metadata, regenerated from source files by setuptools.
16
+ *.egg-info
17
+ .eggs/
18
+
19
+ # PyPI distribution artifacts.
20
+ build/
21
+ dist/
22
+
23
+ # Environments
24
+ .env
25
+ .venv
26
+ env/
27
+ venv/
28
+ ENV/
29
+ env.bak/
30
+ venv.bak/
31
+
32
+ # pyenv
33
+ .python-version
34
+
35
+ # Tests
36
+ .pytest_cache/
37
+
38
+ # Other
39
+ *.DS_Store
40
+
41
+ # PyCharm/vscode
42
+ .idea
43
+ .vscode
44
+
45
+ # keep only the empty datasets and metrics directory with it's __init__.py file
46
+ /src/*/datasets/*
47
+ !/src/*/datasets/__init__.py
48
+
49
+ /src/*/metrics/*
50
+ !/src/*/metrics/__init__.py
51
+
52
+ # Vim
53
+ .*.swp
54
+
55
+ # playground
56
+ /playground
57
+
58
+ # Sphinx documentation
59
+ docs/_build/
60
+ docs/source/_build/
61
+
62
+ # Benchmark results
63
+ report.json
64
+ report.md
README.md ADDED
@@ -0,0 +1,337 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ task_categories:
3
+ - conditional-text-generation
4
+ task_ids:
5
+ - summarization
6
+ languages:
7
+ - am
8
+ - ar
9
+ - az
10
+ - bn
11
+ - my
12
+ - zh
13
+ - en
14
+ - fr
15
+ - gu
16
+ - ha
17
+ - hi
18
+ - ig
19
+ - id
20
+ - ja
21
+ - rn
22
+ - ko
23
+ - ky
24
+ - mr
25
+ - ne
26
+ - om
27
+ - ps
28
+ - fa
29
+ - pcm
30
+ - pt
31
+ - pa
32
+ - ru
33
+ - gd
34
+ - sr
35
+ - si
36
+ - so
37
+ - es
38
+ - sw
39
+ - ta
40
+ - te
41
+ - th
42
+ - ti
43
+ - tr
44
+ - uk
45
+ - ur
46
+ - uz
47
+ - vi
48
+ - cy
49
+ - yo
50
+ size_categories:
51
+ - 1M<n<10M
52
+ licenses:
53
+ - cc-by-nc-sa-4.0
54
+ multilinguality:
55
+ - multilingual
56
+ source_datasets:
57
+ - original
58
+ paperswithcode_id: xl-sum
59
+ annotations_creators:
60
+ - found
61
+ language_creators:
62
+ - found
63
+ pretty_name: XL-Sum
64
+ ---
65
+
66
+ # Dataset Card for "XL-Sum"
67
+
68
+ ## Table of Contents
69
+ - [Dataset Card Creation Guide](#dataset-card-creation-guide)
70
+ - [Table of Contents](#table-of-contents)
71
+ - [Dataset Description](#dataset-description)
72
+ - [Dataset Summary](#dataset-summary)
73
+ - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
74
+ - [Languages](#languages)
75
+ - [Dataset Structure](#dataset-structure)
76
+ - [Data Instances](#data-instances)
77
+ - [Data Fields](#data-fields)
78
+ - [Data Splits](#data-splits)
79
+ - [Dataset Creation](#dataset-creation)
80
+ - [Curation Rationale](#curation-rationale)
81
+ - [Source Data](#source-data)
82
+ - [Initial Data Collection and Normalization](#initial-data-collection-and-normalization)
83
+ - [Who are the source language producers?](#who-are-the-source-language-producers)
84
+ - [Annotations](#annotations)
85
+ - [Annotation process](#annotation-process)
86
+ - [Who are the annotators?](#who-are-the-annotators)
87
+ - [Personal and Sensitive Information](#personal-and-sensitive-information)
88
+ - [Considerations for Using the Data](#considerations-for-using-the-data)
89
+ - [Social Impact of Dataset](#social-impact-of-dataset)
90
+ - [Discussion of Biases](#discussion-of-biases)
91
+ - [Other Known Limitations](#other-known-limitations)
92
+ - [Additional Information](#additional-information)
93
+ - [Dataset Curators](#dataset-curators)
94
+ - [Licensing Information](#licensing-information)
95
+ - [Citation Information](#citation-information)
96
+ - [Contributions](#contributions)
97
+
98
+ ## Dataset Description
99
+
100
+ - **Repository:** [https://github.com/csebuetnlp/xl-sum](https://github.com/csebuetnlp/xl-sum)
101
+ - **Paper:** [XL-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages](https://aclanthology.org/2021.findings-acl.413/)
102
+ - **Point of Contact:** [Tahmid Hasan](mailto:tahmidhasan@cse.buet.ac.bd)
103
+
104
+ ### Dataset Summary
105
+
106
+ We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally annotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics. The dataset covers 45 languages ranging from low to high-resource, for many of which no public dataset is currently available. XL-Sum is highly abstractive, concise, and of high quality, as indicated by human and intrinsic evaluation.
107
+
108
+
109
+ ### Supported Tasks and Leaderboards
110
+
111
+ **Tasks:** Summarization
112
+
113
+ **Leaderboards:** [ExplainaBoard](http://explainaboard.nlpedia.ai/leaderboard/task_xlsum/)
114
+
115
+ ### Languages
116
+
117
+ - `amharic`
118
+ - `arabic`
119
+ - `azerbaijani`
120
+ - `bengali`
121
+ - `burmese`
122
+ - `chinese_simplified`
123
+ - `chinese_traditional`
124
+ - `english`
125
+ - `french`
126
+ - `gujarati`
127
+ - `hausa`
128
+ - `hindi`
129
+ - `igbo`
130
+ - `indonesian`
131
+ - `japanese`
132
+ - `kirundi`
133
+ - `korean`
134
+ - `kyrgyz`
135
+ - `marathi`
136
+ - `nepali`
137
+ - `oromo`
138
+ - `pashto`
139
+ - `persian`
140
+ - `pidgin`
141
+ - `portuguese`
142
+ - `punjabi`
143
+ - `russian`
144
+ - `scottish_gaelic`
145
+ - `serbian_cyrillic`
146
+ - `serbian_latin`
147
+ - `sinhala`
148
+ - `somali`
149
+ - `spanish`
150
+ - `swahili`
151
+ - `tamil`
152
+ - `telugu`
153
+ - `thai`
154
+ - `tigrinya`
155
+ - `turkish`
156
+ - `ukrainian`
157
+ - `urdu`
158
+ - `uzbek`
159
+ - `vietnamese`
160
+ - `welsh`
161
+ - `yoruba`
162
+
163
+ ## Dataset Structure
164
+
165
+ ### Data Instances
166
+
167
+ One example from the `English` dataset is given below in JSON format.
168
+ ```
169
+ {
170
+ "gem_id": "GEM-xlsum_english-train-1589",
171
+ "url": "https://www.bbc.com/news/technology-17657859",
172
+ "title": "Yahoo files e-book advert system patent applications",
173
+ "summary": "Yahoo has signalled it is investigating e-book adverts as a way to stimulate its earnings.",
174
+ "text": "Yahoo's patents suggest users could weigh the type of ads against the sizes of discount before purchase. It says in two US patent applications that ads for digital book readers have been \"less than optimal\" to date. The filings suggest that users could be offered titles at a variety of prices depending on the ads' prominence They add that the products shown could be determined by the type of book being read, or even the contents of a specific chapter, phrase or word. The paperwork was published by the US Patent and Trademark Office late last week and relates to work carried out at the firm's headquarters in Sunnyvale, California. \"Greater levels of advertising, which may be more valuable to an advertiser and potentially more distracting to an e-book reader, may warrant higher discounts,\" it states. Free books It suggests users could be offered ads as hyperlinks based within the book's text, in-laid text or even \"dynamic content\" such as video. Another idea suggests boxes at the bottom of a page could trail later chapters or quotes saying \"brought to you by Company A\". It adds that the more willing the customer is to see the ads, the greater the potential discount. \"Higher frequencies... may even be great enough to allow the e-book to be obtained for free,\" it states. The authors write that the type of ad could influence the value of the discount, with \"lower class advertising... such as teeth whitener advertisements\" offering a cheaper price than \"high\" or \"middle class\" adverts, for things like pizza. The inventors also suggest that ads could be linked to the mood or emotional state the reader is in as a they progress through a title. For example, they say if characters fall in love or show affection during a chapter, then ads for flowers or entertainment could be triggered. The patents also suggest this could applied to children's books - giving the Tom Hanks animated film Polar Express as an example. It says a scene showing a waiter giving the protagonists hot drinks \"may be an excellent opportunity to show an advertisement for hot cocoa, or a branded chocolate bar\". Another example states: \"If the setting includes young characters, a Coke advertisement could be provided, inviting the reader to enjoy a glass of Coke with his book, and providing a graphic of a cool glass.\" It adds that such targeting could be further enhanced by taking account of previous titles the owner has bought. 'Advertising-free zone' At present, several Amazon and Kobo e-book readers offer full-screen adverts when the device is switched off and show smaller ads on their menu screens, but the main text of the titles remains free of marketing. Yahoo does not currently provide ads to these devices, and a move into the area could boost its shrinking revenues. However, Philip Jones, deputy editor of the Bookseller magazine, said that the internet firm might struggle to get some of its ideas adopted. \"This has been mooted before and was fairly well decried,\" he said. \"Perhaps in a limited context it could work if the merchandise was strongly related to the title and was kept away from the text. \"But readers - particularly parents - like the fact that reading is an advertising-free zone. Authors would also want something to say about ads interrupting their narrative flow.\""
175
+ }
176
+ ```
177
+
178
+ When downloading the dataset, the intended language name is required. For instance:
179
+
180
+ ```
181
+ from datasets import load_dataset
182
+ ds = load_dataset("GEM/xlsum", "english")
183
+ ```
184
+
185
+
186
+ ### Data Fields
187
+ - `gem_id`: A string representing the article ID.
188
+ - `url`: A string representing the article URL.
189
+ - `title`: A string containing the article title.
190
+ - `summary`: A string containing the article summary.
191
+ - `text` : A string containing the article text.
192
+
193
+
194
+ ### Data Splits
195
+
196
+ We used a 80%-10%-10% split for all languages with a few exceptions. `English` was split 93%-3.5%-3.5% for the evaluation set size to resemble that of `CNN/DM` and `XSum`; `Scottish Gaelic`, `Kyrgyz` and `Sinhala` had relatively fewer samples, their evaluation sets were increased to 500 samples for more reliable evaluation. Same articles were used for evaluation in the two variants of Chinese and Serbian to prevent data leakage in multilingual training. Individual dataset download links with train-dev-test example counts are given below:
197
+
198
+ Language | ISO 639-1 Code | BBC subdomain(s) | Train | Dev | Test | Total |
199
+ --------------|----------------|------------------|-------|-----|------|-------|
200
+ Amharic | am | https://www.bbc.com/amharic | 5761 | 719 | 719 | 7199 |
201
+ Arabic | ar | https://www.bbc.com/arabic | 37519 | 4689 | 4689 | 46897 |
202
+ Azerbaijani | az | https://www.bbc.com/azeri | 6478 | 809 | 809 | 8096 |
203
+ Bengali | bn | https://www.bbc.com/bengali | 8102 | 1012 | 1012 | 10126 |
204
+ Burmese | my | https://www.bbc.com/burmese | 4569 | 570 | 570 | 5709 |
205
+ Chinese (Simplified) | zh-CN | https://www.bbc.com/ukchina/simp, https://www.bbc.com/zhongwen/simp | 37362 | 4670 | 4670 | 46702 |
206
+ Chinese (Traditional) | zh-TW | https://www.bbc.com/ukchina/trad, https://www.bbc.com/zhongwen/trad | 37373 | 4670 | 4670 | 46713 |
207
+ English | en | https://www.bbc.com/english, https://www.bbc.com/sinhala `*` | 306522 | 11535 | 11535 | 329592 |
208
+ French | fr | https://www.bbc.com/afrique | 8697 | 1086 | 1086 | 10869 |
209
+ Gujarati | gu | https://www.bbc.com/gujarati | 9119 | 1139 | 1139 | 11397 |
210
+ Hausa | ha | https://www.bbc.com/hausa | 6418 | 802 | 802 | 8022 |
211
+ Hindi | hi | https://www.bbc.com/hindi | 70778 | 8847 | 8847 | 88472 |
212
+ Igbo | ig | https://www.bbc.com/igbo | 4183 | 522 | 522 | 5227 |
213
+ Indonesian | id | https://www.bbc.com/indonesia | 38242 | 4780 | 4780 | 47802 |
214
+ Japanese | ja | https://www.bbc.com/japanese | 7113 | 889 | 889 | 8891 |
215
+ Kirundi | rn | https://www.bbc.com/gahuza | 5746 | 718 | 718 | 7182 |
216
+ Korean | ko | https://www.bbc.com/korean | 4407 | 550 | 550 | 5507 |
217
+ Kyrgyz | ky | https://www.bbc.com/kyrgyz | 2266 | 500 | 500 | 3266 |
218
+ Marathi | mr | https://www.bbc.com/marathi | 10903 | 1362 | 1362 | 13627 |
219
+ Nepali | np | https://www.bbc.com/nepali | 5808 | 725 | 725 | 7258 |
220
+ Oromo | om | https://www.bbc.com/afaanoromoo | 6063 | 757 | 757 | 7577 |
221
+ Pashto | ps | https://www.bbc.com/pashto | 14353 | 1794 | 1794 | 17941 |
222
+ Persian | fa | https://www.bbc.com/persian | 47251 | 5906 | 5906 | 59063 |
223
+ Pidgin`**` | pcm | https://www.bbc.com/pidgin | 9208 | 1151 | 1151 | 11510 |
224
+ Portuguese | pt | https://www.bbc.com/portuguese | 57402 | 7175 | 7175 | 71752 |
225
+ Punjabi | pa | https://www.bbc.com/punjabi | 8215 | 1026 | 1026 | 10267 |
226
+ Russian | ru | https://www.bbc.com/russian, https://www.bbc.com/ukrainian `*` | 62243 | 7780 | 7780 | 77803 |
227
+ Scottish Gaelic | gd | https://www.bbc.com/naidheachdan | 1313 | 500 | 500 | 2313 |
228
+ Serbian (Cyrillic) | sr | https://www.bbc.com/serbian/cyr | 7275 | 909 | 909 | 9093 |
229
+ Serbian (Latin) | sr | https://www.bbc.com/serbian/lat | 7276 | 909 | 909 | 9094 |
230
+ Sinhala | si | https://www.bbc.com/sinhala | 3249 | 500 | 500 | 4249 |
231
+ Somali | so | https://www.bbc.com/somali | 5962 | 745 | 745 | 7452 |
232
+ Spanish | es | https://www.bbc.com/mundo | 38110 | 4763 | 4763 | 47636 |
233
+ Swahili | sw | https://www.bbc.com/swahili | 7898 | 987 | 987 | 9872 |
234
+ Tamil | ta | https://www.bbc.com/tamil | 16222 | 2027 | 2027 | 20276 |
235
+ Telugu | te | https://www.bbc.com/telugu | 10421 | 1302 | 1302 | 13025 |
236
+ Thai | th | https://www.bbc.com/thai | 6616 | 826 | 826 | 8268 |
237
+ Tigrinya | ti | https://www.bbc.com/tigrinya | 5451 | 681 | 681 | 6813 |
238
+ Turkish | tr | https://www.bbc.com/turkce | 27176 | 3397 | 3397 | 33970 |
239
+ Ukrainian | uk | https://www.bbc.com/ukrainian | 43201 | 5399 | 5399 | 53999 |
240
+ Urdu | ur | https://www.bbc.com/urdu | 67665 | 8458 | 8458 | 84581 |
241
+ Uzbek | uz | https://www.bbc.com/uzbek | 4728 | 590 | 590 | 5908 |
242
+ Vietnamese | vi | https://www.bbc.com/vietnamese | 32111 | 4013 | 4013 | 40137 |
243
+ Welsh | cy | https://www.bbc.com/cymrufyw | 9732 | 1216 | 1216 | 12164 |
244
+ Yoruba | yo | https://www.bbc.com/yoruba | 6350 | 793 | 793 | 7936 |
245
+
246
+ `*` A lot of articles in BBC Sinhala and BBC Ukrainian were written in English and Russian respectively. They were identified using [Fasttext](https://arxiv.org/abs/1607.01759) and moved accordingly.
247
+
248
+ `**` West African Pidgin English
249
+
250
+ ## Dataset Creation
251
+
252
+ ### Curation Rationale
253
+
254
+ [More information needed](https://github.com/csebuetnlp/xl-sum)
255
+
256
+ ### Source Data
257
+
258
+ [BBC News](https://www.bbc.co.uk/ws/languages)
259
+
260
+ #### Initial Data Collection and Normalization
261
+
262
+ [Detailed in the paper](https://aclanthology.org/2021.findings-acl.413/)
263
+
264
+
265
+ #### Who are the source language producers?
266
+
267
+ [Detailed in the paper](https://aclanthology.org/2021.findings-acl.413/)
268
+
269
+
270
+ ### Annotations
271
+
272
+ [Detailed in the paper](https://aclanthology.org/2021.findings-acl.413/)
273
+
274
+
275
+ #### Annotation process
276
+
277
+ [Detailed in the paper](https://aclanthology.org/2021.findings-acl.413/)
278
+
279
+ #### Who are the annotators?
280
+
281
+ [Detailed in the paper](https://aclanthology.org/2021.findings-acl.413/)
282
+
283
+ ### Personal and Sensitive Information
284
+
285
+ [More information needed](https://github.com/csebuetnlp/xl-sum)
286
+
287
+ ## Considerations for Using the Data
288
+
289
+ ### Social Impact of Dataset
290
+
291
+ [More information needed](https://github.com/csebuetnlp/xl-sum)
292
+
293
+ ### Discussion of Biases
294
+
295
+ [More information needed](https://github.com/csebuetnlp/xl-sum)
296
+
297
+ ### Other Known Limitations
298
+
299
+ [More information needed](https://github.com/csebuetnlp/xl-sum)
300
+
301
+ ## Additional Information
302
+
303
+ ### Dataset Curators
304
+
305
+ [More information needed](https://github.com/csebuetnlp/xl-sum)
306
+
307
+ ### Licensing Information
308
+
309
+ Contents of this repository are restricted to only non-commercial research purposes under the [Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)](https://creativecommons.org/licenses/by-nc-sa/4.0/). Copyright of the dataset contents belongs to the original copyright holders.
310
+ ### Citation Information
311
+
312
+ If you use any of the datasets, models or code modules, please cite the following paper:
313
+ ```
314
+ @inproceedings{hasan-etal-2021-xl,
315
+ title = "{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages",
316
+ author = "Hasan, Tahmid and
317
+ Bhattacharjee, Abhik and
318
+ Islam, Md. Saiful and
319
+ Mubasshir, Kazi and
320
+ Li, Yuan-Fang and
321
+ Kang, Yong-Bin and
322
+ Rahman, M. Sohel and
323
+ Shahriyar, Rifat",
324
+ booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021",
325
+ month = aug,
326
+ year = "2021",
327
+ address = "Online",
328
+ publisher = "Association for Computational Linguistics",
329
+ url = "https://aclanthology.org/2021.findings-acl.413",
330
+ pages = "4693--4703",
331
+ }
332
+ ```
333
+
334
+
335
+ ### Contributions
336
+
337
+ Thanks to [@abhik1505040](https://github.com/abhik1505040) and [@Tahmid](https://github.com/Tahmid04) for adding this dataset.
dataset_infos.json ADDED
@@ -0,0 +1 @@
 
1
+ {"oromo": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "oromo", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 22123227, "num_examples": 6063, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 2472494, "num_examples": 757, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 2481104, "num_examples": 757, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/oromo_XLSum_v2.0.tar.bz2": {"num_bytes": 6986286, "checksum": "7fe25f0c15170769ba686d6c5d2ebb5b7a987c97a04f63b55d11f1654eaf6164"}}, "download_size": 6986286, "post_processing_size": null, "dataset_size": 27076825, "size_in_bytes": 34063111}, "french": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "french", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 28811870, "num_examples": 8697, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 2827033, "num_examples": 1086, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 2902552, "num_examples": 1086, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/french_XLSum_v2.0.tar.bz2": {"num_bytes": 8970255, "checksum": "8d3b0734508a8612598af8e47e023d6fd4e52546543cddb0e3246d720135adf2"}}, "download_size": 8970255, "post_processing_size": null, "dataset_size": 34541455, "size_in_bytes": 43511710}, "amharic": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "amharic", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 36262002, "num_examples": 5761, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 3463171, "num_examples": 719, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 3608789, "num_examples": 719, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/amharic_XLSum_v2.0.tar.bz2": {"num_bytes": 7777009, "checksum": "83c1263b6408d873159dd32d8be52255d39dba22d62cf91d8f01e4fd8d11280e"}}, "download_size": 7777009, "post_processing_size": null, "dataset_size": 43333962, "size_in_bytes": 51110971}, "arabic": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "arabic", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 190771449, "num_examples": 37519, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 21300892, "num_examples": 4689, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 21376004, "num_examples": 4689, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/arabic_XLSum_v2.0.tar.bz2": {"num_bytes": 44918575, "checksum": "371a07df0f89d83c2ca330f56ba33cc4cd39c1b532d6886910c97dfbacfdf2b7"}}, "download_size": 44918575, "post_processing_size": null, "dataset_size": 233448345, "size_in_bytes": 278366920}, "azerbaijani": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "azerbaijani", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 29859787, "num_examples": 6478, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 3179258, "num_examples": 809, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 3120124, "num_examples": 809, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/azerbaijani_XLSum_v2.0.tar.bz2": {"num_bytes": 8751865, "checksum": "7760a8d55e86231f8730f8983e3cb71b1d399d9aaa2a723b201b96ab96e25b3e"}}, "download_size": 8751865, "post_processing_size": null, "dataset_size": 36159169, "size_in_bytes": 44911034}, "bengali": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "bengali", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 90517942, "num_examples": 8102, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 9479839, "num_examples": 1012, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 9237184, "num_examples": 1012, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/bengali_XLSum_v2.0.tar.bz2": {"num_bytes": 12991421, "checksum": "1be42652f92e2427cf0082c153b7a8dc0e38ac0c82821dec3fba732aa011cdc3"}}, "download_size": 12991421, "post_processing_size": null, "dataset_size": 109234965, "size_in_bytes": 122226386}, "burmese": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "burmese", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 37593234, "num_examples": 4569, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 4719056, "num_examples": 570, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 4804369, "num_examples": 570, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/burmese_XLSum_v2.0.tar.bz2": {"num_bytes": 4791972, "checksum": "89669d91ef6fbad55254af5758b081e039f79a5654fb53280a73de4e495d719b"}}, "download_size": 4791972, "post_processing_size": null, "dataset_size": 47116659, "size_in_bytes": 51908631}, "chinese_simplified": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "chinese_simplified", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 137309185, "num_examples": 37362, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 15488723, "num_examples": 4670, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 15403619, "num_examples": 4670, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/chinese_simplified_XLSum_v2.0.tar.bz2": {"num_bytes": 54528291, "checksum": "eaa26651ca379a6f689cecd459eede68c3a67b0de64ce9e9bea62889c252fb2e"}}, "download_size": 54528291, "post_processing_size": null, "dataset_size": 168201527, "size_in_bytes": 222729818}, "chinese_traditional": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "chinese_traditional", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 137485507, "num_examples": 37373, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 15446336, "num_examples": 4670, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 15096396, "num_examples": 4670, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/chinese_traditional_XLSum_v2.0.tar.bz2": {"num_bytes": 54414844, "checksum": "b6955ac6dd659d2c15d8d4a5531c25c774597111aa654ae2861b54460b1ab471"}}, "download_size": 54414844, "post_processing_size": null, "dataset_size": 168028239, "size_in_bytes": 222443083}, "welsh": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "welsh", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26244194, "num_examples": 9732, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 3245290, "num_examples": 1216, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 3190757, "num_examples": 1216, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/welsh_XLSum_v2.0.tar.bz2": {"num_bytes": 8726880, "checksum": "a0041055fd1c324cff3d78b7d03265ba7c2f0db536d6ed8733490b5fddab8284"}}, "download_size": 8726880, "post_processing_size": null, "dataset_size": 32680241, "size_in_bytes": 41407121}, "english": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "english", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 925663615, "num_examples": 306522, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 33125078, "num_examples": 11535, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 33391372, "num_examples": 11535, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/english_XLSum_v2.0.tar.bz2": {"num_bytes": 282122304, "checksum": "ce2e965b9e4b8b337a801fda8b55e673fe2b17c4f3755d7fff886944633fec29"}}, "download_size": 282122304, "post_processing_size": null, "dataset_size": 992180065, "size_in_bytes": 1274302369}, "kirundi": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "kirundi", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 13800280, "num_examples": 5746, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 1753902, "num_examples": 718, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 1732930, "num_examples": 718, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/kirundi_XLSum_v2.0.tar.bz2": {"num_bytes": 4541685, "checksum": "0c60993dffee6b536500872a9f42118081a3a74d5361b6f7d776b644af29e8f5"}}, "download_size": 4541685, "post_processing_size": null, "dataset_size": 17287112, "size_in_bytes": 21828797}, "gujarati": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "gujarati", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 113868106, "num_examples": 9119, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 10134388, "num_examples": 1139, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 10327726, "num_examples": 1139, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/gujarati_XLSum_v2.0.tar.bz2": {"num_bytes": 16617867, "checksum": "901378aadcc57f495aee1b7de2e8778832635fec128ed1ccf3eab8286f157f02"}}, "download_size": 16617867, "post_processing_size": null, "dataset_size": 134330220, "size_in_bytes": 150948087}, "hausa": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "hausa", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 14697774, "num_examples": 6418, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 1828543, "num_examples": 802, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 1865713, "num_examples": 802, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/hausa_XLSum_v2.0.tar.bz2": {"num_bytes": 4706193, "checksum": "b47ce8b01ee540a863147e122308de63274e989e4bde9f045a6a91f6ef8ca12a"}}, "download_size": 4706193, "post_processing_size": null, "dataset_size": 18392030, "size_in_bytes": 23098223}, "hindi": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "hindi", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 554968859, "num_examples": 70778, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 57326682, "num_examples": 8847, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 57069754, "num_examples": 8847, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/hindi_XLSum_v2.0.tar.bz2": {"num_bytes": 82265197, "checksum": "23e625cfdd596ae59bab6e1756756c4d68bf50f355437943b3e57f0593662193"}}, "download_size": 82265197, "post_processing_size": null, "dataset_size": 669365295, "size_in_bytes": 751630492}, "igbo": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "igbo", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 9722856, "num_examples": 4183, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 1280421, "num_examples": 522, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 1273998, "num_examples": 522, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/igbo_XLSum_v2.0.tar.bz2": {"num_bytes": 2883995, "checksum": "cf3e5d6653a4730d31623e80eb3fd785a0febdeaf13502f0daf84d3598abeb23"}}, "download_size": 2883995, "post_processing_size": null, "dataset_size": 12277275, "size_in_bytes": 15161270}, "indonesian": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "indonesian", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 129899335, "num_examples": 38242, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 14090766, "num_examples": 4780, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 13903128, "num_examples": 4780, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/indonesian_XLSum_v2.0.tar.bz2": {"num_bytes": 41083471, "checksum": "a2a058e8753a0541121817801a6bc73ac60728d88b31a7d2a0fc92f1a1a68845"}}, "download_size": 41083471, "post_processing_size": null, "dataset_size": 157893229, "size_in_bytes": 198976700}, "japanese": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "japanese", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 37104034, "num_examples": 7113, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 4468817, "num_examples": 889, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 4386558, "num_examples": 889, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/japanese_XLSum_v2.0.tar.bz2": {"num_bytes": 10559850, "checksum": "d277164a97aaa8a4bab0ea219075255db1074b5aaccedb202156db3bd02529ff"}}, "download_size": 10559850, "post_processing_size": null, "dataset_size": 45959409, "size_in_bytes": 56519259}, "korean": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "korean", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 19502723, "num_examples": 4407, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 2268575, "num_examples": 550, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 2275111, "num_examples": 550, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/korean_XLSum_v2.0.tar.bz2": {"num_bytes": 5894987, "checksum": "b6ba096c57ac088305c6bd1cd5c147420ce7ae004d330db111cce412ed818daf"}}, "download_size": 5894987, "post_processing_size": null, "dataset_size": 24046409, "size_in_bytes": 29941396}, "kyrgyz": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "kyrgyz", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 17399605, "num_examples": 2266, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 3345256, "num_examples": 500, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 3283663, "num_examples": 500, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/kyrgyz_XLSum_v2.0.tar.bz2": {"num_bytes": 3740941, "checksum": "250cb05da41156614e55b58193b50ff33e87cd40ed6435b6c2621910c8b70f07"}}, "download_size": 3740941, "post_processing_size": null, "dataset_size": 24028524, "size_in_bytes": 27769465}, "marathi": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "marathi", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 138789757, "num_examples": 10903, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 14225851, "num_examples": 1362, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 13985234, "num_examples": 1362, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/marathi_XLSum_v2.0.tar.bz2": {"num_bytes": 19617263, "checksum": "e0021b4eebe27efe0c7736c73d6ffb0a59234d9b1adcdffee6d692522bfa24b1"}}, "download_size": 19617263, "post_processing_size": null, "dataset_size": 167000842, "size_in_bytes": 186618105}, "spanish": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "spanish", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 209679507, "num_examples": 38110, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 24224627, "num_examples": 4763, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 24320027, "num_examples": 4763, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/spanish_XLSum_v2.0.tar.bz2": {"num_bytes": 69377384, "checksum": "70499154fe1d1c8df3b4667921d2c8c7b508da5473aa9387c4330b3b22288360"}}, "download_size": 69377384, "post_processing_size": null, "dataset_size": 258224161, "size_in_bytes": 327601545}, "scottish_gaelic": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "scottish_gaelic", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2175305, "num_examples": 1313, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 946869, "num_examples": 500, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 949138, "num_examples": 500, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/scottish_gaelic_XLSum_v2.0.tar.bz2": {"num_bytes": 867345, "checksum": "868b6cb1d711fd15ad13665bb949552722affec1644adf3783fc27dd060e8209"}}, "download_size": 867345, "post_processing_size": null, "dataset_size": 4071312, "size_in_bytes": 4938657}, "nepali": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "nepali", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 44298011, "num_examples": 5808, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 5367550, "num_examples": 725, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 5438476, "num_examples": 725, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/nepali_XLSum_v2.0.tar.bz2": {"num_bytes": 6311786, "checksum": "a2d0101abdfe83bca059e6e47af40ee2e389c362da0e798db9dc1606bd1201cd"}}, "download_size": 6311786, "post_processing_size": null, "dataset_size": 55104037, "size_in_bytes": 61415823}, "pashto": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "pashto", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 59609964, "num_examples": 14353, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 6932317, "num_examples": 1794, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 7130916, "num_examples": 1794, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/pashto_XLSum_v2.0.tar.bz2": {"num_bytes": 13070774, "checksum": "6793398f12a8205ccf72806de69dc2dc752469d9f697b977d0d43ee4268b4bc2"}}, "download_size": 13070774, "post_processing_size": null, "dataset_size": 73673197, "size_in_bytes": 86743971}, "persian": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "persian", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 301907581, "num_examples": 47251, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 30070685, "num_examples": 5906, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 29755056, "num_examples": 5906, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/persian_XLSum_v2.0.tar.bz2": {"num_bytes": 66138331, "checksum": "b7f1d3d2050e097f6dbd880d239f47cd320bd34a99db06b560007e79eef0a0d9"}}, "download_size": 66138331, "post_processing_size": null, "dataset_size": 361733322, "size_in_bytes": 427871653}, "pidgin": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "pidgin", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 17844227, "num_examples": 9208, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 2369029, "num_examples": 1151, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 2378570, "num_examples": 1151, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/pidgin_XLSum_v2.0.tar.bz2": {"num_bytes": 6284961, "checksum": "33f280d203900bd404782a4333bde5729ea6db42fb6f9978b55f5468daa4d5f4"}}, "download_size": 6284961, "post_processing_size": null, "dataset_size": 22591826, "size_in_bytes": 28876787}, "portuguese": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "portuguese", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 245742156, "num_examples": 57402, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 23631130, "num_examples": 7175, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 23822520, "num_examples": 7175, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/portuguese_XLSum_v2.0.tar.bz2": {"num_bytes": 79871451, "checksum": "ba6f99f8865872b0d7b348b77b4edfa8ae97fd3ebd986e2820a87cd2b548c760"}}, "download_size": 79871451, "post_processing_size": null, "dataset_size": 293195806, "size_in_bytes": 373067257}, "punjabi": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "punjabi", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 95946232, "num_examples": 8215, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 7897965, "num_examples": 1026, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 8021785, "num_examples": 1026, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/punjabi_XLSum_v2.0.tar.bz2": {"num_bytes": 13554146, "checksum": "998263da7e6e60dc8418ad65bf138c443b06bef2087cf339a1aee00855986e84"}}, "download_size": 13554146, "post_processing_size": null, "dataset_size": 111865982, "size_in_bytes": 125420128}, "russian": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "russian", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 493572785, "num_examples": 62243, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 51724894, "num_examples": 7780, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 51888219, "num_examples": 7780, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/russian_XLSum_v2.0.tar.bz2": {"num_bytes": 105139577, "checksum": "5a582ea371c981295da82b6d4c3a8f44d8e266837a86aeae199e2a49b04af3bf"}}, "download_size": 105139577, "post_processing_size": null, "dataset_size": 597185898, "size_in_bytes": 702325475}, "serbian_cyrillic": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "serbian_cyrillic", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 71924762, "num_examples": 7275, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 5996969, "num_examples": 909, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 6160614, "num_examples": 909, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/serbian_cyrillic_XLSum_v2.0.tar.bz2": {"num_bytes": 15410451, "checksum": "66d76eede7b1e36510a6c18ba5585a46dd22f2a33c8c0667a05ffbb114e94b52"}}, "download_size": 15410451, "post_processing_size": null, "dataset_size": 84082345, "size_in_bytes": 99492796}, "serbian_latin": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "serbian_latin", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 41804900, "num_examples": 7276, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 3625928, "num_examples": 909, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 3523888, "num_examples": 909, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/serbian_latin_XLSum_v2.0.tar.bz2": {"num_bytes": 14810788, "checksum": "88413e18996dcb4df178dcd8a604f47d25f73c68d3fca98ad26008235212def9"}}, "download_size": 14810788, "post_processing_size": null, "dataset_size": 48954716, "size_in_bytes": 63765504}, "sinhala": {"description": "We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally \nannotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.\nThe dataset covers 45 languages ranging from low to high-resource, for many of which no\npublic dataset is currently available. XL-Sum is highly abstractive, concise, \nand of high quality, as indicated by human and intrinsic evaluation. \n", "citation": "@inproceedings{hasan-etal-2021-xl,\n title = \"{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages\",\n author = \"Hasan, Tahmid and\n Bhattacharjee, Abhik and\n Islam, Md. Saiful and\n Mubasshir, Kazi and\n Li, Yuan-Fang and\n Kang, Yong-Bin and\n Rahman, M. Sohel and\n Shahriyar, Rifat\",\n booktitle = \"Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021\",\n month = aug,\n year = \"2021\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://aclanthology.org/2021.findings-acl.413\",\n pages = \"4693--4703\",\n}\n", "homepage": "https://github.com/csebuetnlp/xl-sum", "license": "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "xlsum", "config_name": "sinhala", "version": {"version_str": "2.0.0", "description": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 20809891, "num_examples": 3249, "dataset_name": "xlsum"}, "test": {"name": "test", "num_bytes": 2953071, "num_examples": 500, "dataset_name": "xlsum"}, "validation": {"name": "validation", "num_bytes": 2914695, "num_examples": 500, "dataset_name": "xlsum"}}, "download_checksums": {"https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/sinhala_XLSum_v2.0.tar.bz2": {"num_bytes": 3087807, "checksum": "23dee7024e1feb6475e5a807e46438b3a22443c7e34daa22ad9973fed71f9245"}}, "download_size": 3087807, "post_processing_size": null, "dataset_size": 26677657, "size_in_bytes": 29765464}}
dummy/amharic/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a0fb5edc57c777c588ecd0866ee387263fde8dfb42f32a784ddaf8005b95dd6
3
+ size 32031
dummy/arabic/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fde69eab35d1e14af7bcfbba55b0fb022925468a39e1e3a98517659c15b07e06
3
+ size 26628
dummy/azerbaijani/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d1696061c16c16d5d8d6506f98a2ea90e8d6144c6d0a64c81b3c442cb2f0088
3
+ size 29683
dummy/bengali/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1a03035a20942adbfcfdc955604ecf2f05f0ed079434ca48da4254707b961bc
3
+ size 37245
dummy/burmese/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab3f88fe3e93a3d101b3ef7e9985d603d62de1c1b8c9630512064e6a9c4bffdb
3
+ size 18088
dummy/chinese_simplified/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83b15d88f10a2509d78ada614a3693ca6e943e3062bf307ab7b56c18d4947d74
3
+ size 32654
dummy/chinese_traditional/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30f41acc6624340f3a5364f7a39e7f982b340e30a5a164c1343156e2d5c11a66
3
+ size 36527
dummy/english/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7630d021c3b2684c69c39b566e5ac2c0b47863a37091080518a166a90ec558f3
3
+ size 18812
dummy/french/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:561a8fe6e769065fcc4421592cae7f225f484c91488ab370d08bf1961811d113
3
+ size 28986
dummy/gujarati/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0aa09e0bff4bf20c69281896236505d5faa78bd5d1967c78891a043a1e14ec3
3
+ size 30119
dummy/hausa/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:667a190b5d0dde6bd9ef7c2de99bca84a36aa26d1d5f89dd5d4c2e5aedc14751
3
+ size 16125
dummy/hindi/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d8b453ccd5f8130b792954ee20b27a3d26809336316afb0b617704728e8f952
3
+ size 38763
dummy/igbo/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b1d6a528b080fe75c791b3a1660f252bba7cdeec46247c6d151547bb20a67c5
3
+ size 11720
dummy/indonesian/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52618b72dd3421377cc61ed372f1297dfc0d18abdf74d1d3df4accacfc110ee8
3
+ size 19065
dummy/japanese/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f5ebf5d2d7fce6c4734ae7daf22d65a202f144041969b6035e7b6791ee20972
3
+ size 36656
dummy/kirundi/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:74dcb0db009a45678b38a0e80f7e079faa9d322a3fe1a464313c84095e26c922
3
+ size 12633
dummy/korean/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95d8c6c64b079075d7165958c94c70059701be68af9baa2c7d66be17db23f7ff
3
+ size 19075
dummy/kyrgyz/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c74bef8162900eb203e6a4ff435f4fb0e63fdfd9bdf9d5763fc99a3c671116cb
3
+ size 32342
dummy/marathi/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7ed754778df105c9f9ffeeb2a5986300c4636d1d4b2cbf348a32424076a46c0
3
+ size 31366
dummy/nepali/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cc3f05e8efb0cf8518067fb65cd632112676a349a1952d963ded9404a0afdfa
3
+ size 35058
dummy/oromo/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab8fb5da02236a9b3275cb33fa9dac79147125576c36c0fe6d551c1a6e23ae14
3
+ size 27841
dummy/pashto/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e6d051e8273c44cb71b3dc50bd1c7627c4f086d250cc15715fd6447682039ff
3
+ size 20331
dummy/persian/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:98bb075c8ddab2dfdd7ab32aa7a040cc3ffdb73ad1c01e3457d546361fa8e4bf
3
+ size 40850
dummy/pidgin/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a06d32ea73d918e000a34fdbd7508469e0c61aac66f59ea4961ae0d3a8d6b03
3
+ size 12275
dummy/portuguese/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:325a7342d5cb6e405ad762bf0e0cf4ebc50caab6c80275bb13da4b5e26f28aa3
3
+ size 26744
dummy/punjabi/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff417195cf72ed65fc98f019f1166231897464d0c65e3c94f219e3a97cc36883
3
+ size 33011
dummy/russian/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5b818921cee75e961e8fd647abfd1c18f9fc794908f12a524446d1a836e4b72
3
+ size 31622
dummy/scottish_gaelic/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f2aab8c8e8aa51490a7d42bd2e55c9f2425df87184843d005bd4710d4bd2ca6
3
+ size 10871
dummy/serbian_cyrillic/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f8c8d8f2bcdc89182a778aef28b385a06cf24abbd9b25406be2efde7be66439
3
+ size 47399
dummy/serbian_latin/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7a1c03d572b8d25e6f419fdac092be53fc0e54bf6084f8fd5e1ef89a51cf225
3
+ size 42772
dummy/sinhala/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e3939d570cdb7287262705f384ebc2316aa5c970e2052d2c39b37ed8aa0f8734
3
+ size 12355
dummy/somali/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d3703074c96af2ba7f08ae7ef975028c3cd2169003af38c59296a4669fb8186
3
+ size 17472
dummy/spanish/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cdb4f68cf4152b48f97f9c03248350cb97a92586f8cf34bf6b94b5fbf9a0d241
3
+ size 27270
dummy/swahili/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c0eedd49a4c7a3bff5567a44da2fce7cf78306ef0cd4534271b8867dbf8a961
3
+ size 15130
dummy/tamil/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3431ba7b5650d6036eafc7d7cc95e9e05fca3c048b8603bfe3d21b1f6717f9f8
3
+ size 36906
dummy/telugu/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77422584003c14ac7ad57d01ffc297749e984cde932191dd7cbe36c73fbc0007
3
+ size 31510
dummy/thai/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2bb642021652a73f0c1c3d6a78480fa4c42686c8b20415ee39f612ddb8abdcff
3
+ size 32318
dummy/tigrinya/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f22f33a7dbb7c53b7b1ff74f0ef1402e424b033e27511c882a28a4bfa346d1ac
3
+ size 27173
dummy/turkish/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5484ecb480512f3b416443a649e573131828719da807f6d7cb2fcaa876c8f891
3
+ size 30610
dummy/ukrainian/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f888ff5be25e9b21b2411ad35f6d517f784bd23189bb6b01224804cf4542d9b0
3
+ size 26369
dummy/urdu/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63b2505b5a4bb28c56a933c15b8ecd33164edad4f1d88672ca7bdad281e9b0c2
3
+ size 21933
dummy/uzbek/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e44471b3bc3cd5097acbb0c27ef5dd05b30a2aa13df4c28af214053bdf14f476
3
+ size 41258
dummy/vietnamese/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:575eb2a14bb8b77a984e39166682eb9f739784cf79696f35edf0d1f1b46907e6
3
+ size 43882
dummy/welsh/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d4ebd3e74c2ce1d64ca9b5345746178875a5d337202f07572a9954837df7d97
3
+ size 17625
dummy/yoruba/2.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:111fe3615ecd3344995325cabc7895a2eb45411db1939dc760135dd621ae43f5
3
+ size 18455
xlsum.py ADDED
@@ -0,0 +1,167 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """XL-Sum abstractive summarization dataset."""
2
+
3
+
4
+ import json
5
+ import os
6
+
7
+ import datasets
8
+
9
+
10
+ _CITATION = """\
11
+ @inproceedings{hasan-etal-2021-xl,
12
+ title = "{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages",
13
+ author = "Hasan, Tahmid and
14
+ Bhattacharjee, Abhik and
15
+ Islam, Md. Saiful and
16
+ Mubasshir, Kazi and
17
+ Li, Yuan-Fang and
18
+ Kang, Yong-Bin and
19
+ Rahman, M. Sohel and
20
+ Shahriyar, Rifat",
21
+ booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021",
22
+ month = aug,
23
+ year = "2021",
24
+ address = "Online",
25
+ publisher = "Association for Computational Linguistics",
26
+ url = "https://aclanthology.org/2021.findings-acl.413",
27
+ pages = "4693--4703",
28
+ }
29
+ """
30
+
31
+
32
+ _DESCRIPTION = """\
33
+ We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally
34
+ annotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.
35
+ The dataset covers 45 languages ranging from low to high-resource, for many of which no
36
+ public dataset is currently available. XL-Sum is highly abstractive, concise,
37
+ and of high quality, as indicated by human and intrinsic evaluation.
38
+ """
39
+
40
+ _HOMEPAGE = "https://github.com/csebuetnlp/xl-sum"
41
+
42
+ _LICENSE = "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)"
43
+
44
+ _URL = "https://huggingface.co/datasets/csebuetnlp/xlsum/resolve/main/data/{}_XLSum_v{}.tar.bz2"
45
+
46
+ _LANGUAGES = [
47
+ "oromo",
48
+ "french",
49
+ "amharic",
50
+ "arabic",
51
+ "azerbaijani",
52
+ "bengali",
53
+ "burmese",
54
+ "chinese_simplified",
55
+ "chinese_traditional",
56
+ "welsh",
57
+ "english",
58
+ "kirundi",
59
+ "gujarati",
60
+ "hausa",
61
+ "hindi",
62
+ "igbo",
63
+ "indonesian",
64
+ "japanese",
65
+ "korean",
66
+ "kyrgyz",
67
+ "marathi",
68
+ "spanish",
69
+ "scottish_gaelic",
70
+ "nepali",
71
+ "pashto",
72
+ "persian",
73
+ "pidgin",
74
+ "portuguese",
75
+ "punjabi",
76
+ "russian",
77
+ "serbian_cyrillic",
78
+ "serbian_latin",
79
+ "sinhala",
80
+ "somali",
81
+ "swahili",
82
+ "tamil",
83
+ "telugu",
84
+ "thai",
85
+ "tigrinya",
86
+ "turkish",
87
+ "ukrainian",
88
+ "urdu",
89
+ "uzbek",
90
+ "vietnamese",
91
+ "yoruba",
92
+ ]
93
+
94
+
95
+ class Xlsum(datasets.GeneratorBasedBuilder):
96
+ VERSION = datasets.Version("2.0.0")
97
+
98
+ BUILDER_CONFIGS = [
99
+ datasets.BuilderConfig(
100
+ name="{}".format(lang),
101
+ version=datasets.Version("2.0.0")
102
+ )
103
+ for lang in _LANGUAGES
104
+ ]
105
+
106
+ def _info(self):
107
+ return datasets.DatasetInfo(
108
+ description=_DESCRIPTION,
109
+ features=datasets.Features(
110
+ {
111
+ "gem_id": datasets.Value("string"),
112
+ "url": datasets.Value("string"),
113
+ "title": datasets.Value("string"),
114
+ "summary": datasets.Value("string"),
115
+ "text": datasets.Value("string"),
116
+ }
117
+ ),
118
+ supervised_keys=None,
119
+ homepage=_HOMEPAGE,
120
+ citation=_CITATION,
121
+ license=_LICENSE,
122
+ version=self.VERSION,
123
+ )
124
+
125
+ def _split_generators(self, dl_manager):
126
+ """Returns SplitGenerators."""
127
+ lang = str(self.config.name)
128
+ url = _URL.format(lang, self.VERSION.version_str[:-2])
129
+
130
+ data_dir = dl_manager.download_and_extract(url)
131
+ return [
132
+ datasets.SplitGenerator(
133
+ name=datasets.Split.TRAIN,
134
+ gen_kwargs={
135
+ "filepath": os.path.join(data_dir, lang + "_train.jsonl"),
136
+ "split": "train"
137
+ },
138
+ ),
139
+ datasets.SplitGenerator(
140
+ name=datasets.Split.TEST,
141
+ gen_kwargs={
142
+ "filepath": os.path.join(data_dir, lang + "_test.jsonl"),
143
+ "split": "test"
144
+ },
145
+ ),
146
+ datasets.SplitGenerator(
147
+ name=datasets.Split.VALIDATION,
148
+ gen_kwargs={
149
+ "filepath": os.path.join(data_dir, lang + "_val.jsonl"),
150
+ "split": "validation"
151
+ },
152
+ ),
153
+ ]
154
+
155
+ def _generate_examples(self, filepath, split):
156
+ """Yields examples as (key, example) tuples."""
157
+
158
+ with open(filepath, encoding="utf-8") as f:
159
+ for idx_, row in enumerate(f, 1):
160
+ data = json.loads(row)
161
+ yield idx_, {
162
+ "gem_id": f"GEM-xlsum_{self.config.name}-{split}-{idx_}",
163
+ "url": data["url"],
164
+ "title": data["title"],
165
+ "summary": data["summary"],
166
+ "text": data["text"],
167
+ }