Datasets:

Languages:
English
Multilinguality:
monolingual
Size Categories:
100K<n<1M
Language Creators:
crowdsourced
Annotations Creators:
crowdsourced
Source Datasets:
original
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
3ffb306
1 Parent(s): 80596c4

Update dataset card (#8)

Browse files

- Update dataset card (d7df8e6afcfb0896687d1f6093694eeb7f7d9263)

Files changed (1) hide show
  1. README.md +50 -23
README.md CHANGED
@@ -72,7 +72,7 @@ train-eval-index:
72
  name: SQuAD v2
73
  ---
74
 
75
- # Dataset Card for "squad_v2"
76
 
77
  ## Table of Contents
78
  - [Dataset Card for "squad_v2"](#dataset-card-for-squad_v2)
@@ -108,27 +108,26 @@ train-eval-index:
108
 
109
  ## Dataset Description
110
 
111
- - **Homepage:** [https://rajpurkar.github.io/SQuAD-explorer/](https://rajpurkar.github.io/SQuAD-explorer/)
112
  - **Repository:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
113
- - **Paper:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
114
  - **Point of Contact:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
115
- - **Size of downloaded dataset files:** 46.49 MB
116
- - **Size of the generated dataset:** 128.52 MB
117
- - **Total amount of disk used:** 175.02 MB
118
 
119
  ### Dataset Summary
120
 
121
- combines the 100,000 questions in SQuAD1.1 with over 50,000 unanswerable questions written adversarially by crowdworkers
122
- to look similar to answerable ones. To do well on SQuAD2.0, systems must not only answer questions when possible, but
123
- also determine when no answer is supported by the paragraph and abstain from answering.
 
 
124
 
125
  ### Supported Tasks and Leaderboards
126
 
127
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
128
 
129
  ### Languages
130
 
131
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
132
 
133
  ## Dataset Structure
134
 
@@ -227,23 +226,51 @@ The data fields are the same among all splits.
227
 
228
  ### Licensing Information
229
 
230
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
231
 
232
  ### Citation Information
233
 
234
  ```
235
- @article{2016arXiv160605250R,
236
- author = {{Rajpurkar}, Pranav and {Zhang}, Jian and {Lopyrev},
237
- Konstantin and {Liang}, Percy},
238
- title = "{SQuAD: 100,000+ Questions for Machine Comprehension of Text}",
239
- journal = {arXiv e-prints},
240
- year = 2016,
241
- eid = {arXiv:1606.05250},
242
- pages = {arXiv:1606.05250},
243
- archivePrefix = {arXiv},
244
- eprint = {1606.05250},
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
245
  }
246
-
247
  ```
248
 
249
 
72
  name: SQuAD v2
73
  ---
74
 
75
+ # Dataset Card for SQuAD 2.0
76
 
77
  ## Table of Contents
78
  - [Dataset Card for "squad_v2"](#dataset-card-for-squad_v2)
108
 
109
  ## Dataset Description
110
 
111
+ - **Homepage:** https://rajpurkar.github.io/SQuAD-explorer/
112
  - **Repository:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
113
+ - **Paper:** https://arxiv.org/abs/1806.03822
114
  - **Point of Contact:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
 
 
 
115
 
116
  ### Dataset Summary
117
 
118
+ Stanford Question Answering Dataset (SQuAD) is a reading comprehension dataset, consisting of questions posed by crowdworkers on a set of Wikipedia articles, where the answer to every question is a segment of text, or span, from the corresponding reading passage, or the question might be unanswerable.
119
+
120
+ SQuAD 2.0 combines the 100,000 questions in SQuAD1.1 with over 50,000 unanswerable questions written adversarially by crowdworkers
121
+ to look similar to answerable ones. To do well on SQuAD2.0, systems must not only answer questions when possible, but
122
+ also determine when no answer is supported by the paragraph and abstain from answering.
123
 
124
  ### Supported Tasks and Leaderboards
125
 
126
+ Question Answering.
127
 
128
  ### Languages
129
 
130
+ English (`en`).
131
 
132
  ## Dataset Structure
133
 
226
 
227
  ### Licensing Information
228
 
229
+ The dataset is distributed under the CC BY-SA 4.0 license.
230
 
231
  ### Citation Information
232
 
233
  ```
234
+ @inproceedings{rajpurkar-etal-2018-know,
235
+ title = "Know What You Don{'}t Know: Unanswerable Questions for {SQ}u{AD}",
236
+ author = "Rajpurkar, Pranav and
237
+ Jia, Robin and
238
+ Liang, Percy",
239
+ editor = "Gurevych, Iryna and
240
+ Miyao, Yusuke",
241
+ booktitle = "Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
242
+ month = jul,
243
+ year = "2018",
244
+ address = "Melbourne, Australia",
245
+ publisher = "Association for Computational Linguistics",
246
+ url = "https://aclanthology.org/P18-2124",
247
+ doi = "10.18653/v1/P18-2124",
248
+ pages = "784--789",
249
+ eprint={1806.03822},
250
+ archivePrefix={arXiv},
251
+ primaryClass={cs.CL}
252
+ }
253
+ @inproceedings{rajpurkar-etal-2016-squad,
254
+ title = "{SQ}u{AD}: 100,000+ Questions for Machine Comprehension of Text",
255
+ author = "Rajpurkar, Pranav and
256
+ Zhang, Jian and
257
+ Lopyrev, Konstantin and
258
+ Liang, Percy",
259
+ editor = "Su, Jian and
260
+ Duh, Kevin and
261
+ Carreras, Xavier",
262
+ booktitle = "Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing",
263
+ month = nov,
264
+ year = "2016",
265
+ address = "Austin, Texas",
266
+ publisher = "Association for Computational Linguistics",
267
+ url = "https://aclanthology.org/D16-1264",
268
+ doi = "10.18653/v1/D16-1264",
269
+ pages = "2383--2392",
270
+ eprint={1606.05250},
271
+ archivePrefix={arXiv},
272
+ primaryClass={cs.CL},
273
  }
 
274
  ```
275
 
276