Datasets:
Commit
•
0852495
1
Parent(s):
7572873
Fix loading example in opus dataset cards (#4813)
Browse files* Fix loading example in opus dataset cards
* Fix dataset cards
Commit from https://github.com/huggingface/datasets/commit/18287ed524ac79f538abf77183f906e91d0de3de
README.md
CHANGED
@@ -46,7 +46,7 @@ configs:
|
|
46 |
- en-vi
|
47 |
---
|
48 |
|
49 |
-
# Dataset Card
|
50 |
|
51 |
## Table of Contents
|
52 |
- [Dataset Description](#dataset-description)
|
@@ -82,12 +82,18 @@ configs:
|
|
82 |
|
83 |
### Dataset Summary
|
84 |
|
|
|
85 |
|
86 |
-
|
87 |
-
|
88 |
-
|
|
|
89 |
|
90 |
-
|
|
|
|
|
|
|
|
|
91 |
|
92 |
|
93 |
### Supported Tasks and Leaderboards
|
@@ -96,21 +102,50 @@ E.g.
|
|
96 |
|
97 |
### Languages
|
98 |
|
99 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
100 |
|
101 |
## Dataset Structure
|
102 |
|
103 |
### Data Instances
|
104 |
|
105 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
106 |
|
107 |
### Data Fields
|
108 |
|
109 |
-
|
|
|
110 |
|
111 |
### Data Splits
|
112 |
|
113 |
-
|
114 |
|
115 |
## Dataset Creation
|
116 |
|
@@ -172,6 +207,23 @@ E.g.
|
|
172 |
|
173 |
### Citation Information
|
174 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
175 |
@InProceedings{TIEDEMANN12.463,
|
176 |
author = {J{\"o}rg Tiedemann},
|
177 |
title = {Parallel Data, Tools and Interfaces in OPUS},
|
@@ -184,7 +236,8 @@ E.g.
|
|
184 |
publisher = {European Language Resources Association (ELRA)},
|
185 |
isbn = {978-2-9517408-7-7},
|
186 |
language = {english}
|
187 |
-
|
|
|
188 |
|
189 |
### Contributions
|
190 |
|
|
|
46 |
- en-vi
|
47 |
---
|
48 |
|
49 |
+
# Dataset Card for OpusWikipedia
|
50 |
|
51 |
## Table of Contents
|
52 |
- [Dataset Description](#dataset-description)
|
|
|
82 |
|
83 |
### Dataset Summary
|
84 |
|
85 |
+
This is a corpus of parallel sentences extracted from Wikipedia by Krzysztof Wołk and Krzysztof Marasek.
|
86 |
|
87 |
+
Tha dataset contains 20 languages and 36 bitexts.
|
88 |
+
|
89 |
+
To load a language pair which isn't part of the config, all you need to do is specify the language code as pairs,
|
90 |
+
e.g.
|
91 |
|
92 |
+
```python
|
93 |
+
dataset = load_dataset("opus_wikipedia", lang1="it", lang2="pl")
|
94 |
+
```
|
95 |
+
|
96 |
+
You can find the valid pairs in Homepage section of Dataset Description: http://opus.nlpl.eu/Wikipedia.php
|
97 |
|
98 |
|
99 |
### Supported Tasks and Leaderboards
|
|
|
102 |
|
103 |
### Languages
|
104 |
|
105 |
+
The languages in the dataset are:
|
106 |
+
- ar
|
107 |
+
- bg
|
108 |
+
- cs
|
109 |
+
- de
|
110 |
+
- el
|
111 |
+
- en
|
112 |
+
- es
|
113 |
+
- fa
|
114 |
+
- fr
|
115 |
+
- he
|
116 |
+
- hu
|
117 |
+
- it
|
118 |
+
- nl
|
119 |
+
- pl
|
120 |
+
- pt
|
121 |
+
- ro
|
122 |
+
- ru
|
123 |
+
- sl
|
124 |
+
- tr
|
125 |
+
- vi
|
126 |
|
127 |
## Dataset Structure
|
128 |
|
129 |
### Data Instances
|
130 |
|
131 |
+
```
|
132 |
+
{
|
133 |
+
'id': '0',
|
134 |
+
'translation': {
|
135 |
+
"ar": "* Encyclopaedia of Mathematics online encyclopaedia from Springer, Graduate-level reference work with over 8,000 entries, illuminating nearly 50,000 notions in mathematics.",
|
136 |
+
"en": "*Encyclopaedia of Mathematics online encyclopaedia from Springer, Graduate-level reference work with over 8,000 entries, illuminating nearly 50,000 notions in mathematics."
|
137 |
+
}
|
138 |
+
}
|
139 |
+
```
|
140 |
|
141 |
### Data Fields
|
142 |
|
143 |
+
- `id` (`str`): Unique identifier of the parallel sentence for the pair of languages.
|
144 |
+
- `translation` (`dict`): Parallel sentences for the pair of languages.
|
145 |
|
146 |
### Data Splits
|
147 |
|
148 |
+
The dataset contains a single `train` split.
|
149 |
|
150 |
## Dataset Creation
|
151 |
|
|
|
207 |
|
208 |
### Citation Information
|
209 |
|
210 |
+
```bibtex
|
211 |
+
@article{WOLK2014126,
|
212 |
+
title = {Building Subject-aligned Comparable Corpora and Mining it for Truly Parallel Sentence Pairs},
|
213 |
+
journal = {Procedia Technology},
|
214 |
+
volume = {18},
|
215 |
+
pages = {126-132},
|
216 |
+
year = {2014},
|
217 |
+
note = {International workshop on Innovations in Information and Communication Science and Technology, IICST 2014, 3-5 September 2014, Warsaw, Poland},
|
218 |
+
issn = {2212-0173},
|
219 |
+
doi = {https://doi.org/10.1016/j.protcy.2014.11.024},
|
220 |
+
url = {https://www.sciencedirect.com/science/article/pii/S2212017314005453},
|
221 |
+
author = {Krzysztof Wołk and Krzysztof Marasek},
|
222 |
+
keywords = {Comparable corpora, machine translation, NLP},
|
223 |
+
}
|
224 |
+
```
|
225 |
+
|
226 |
+
```bibtex
|
227 |
@InProceedings{TIEDEMANN12.463,
|
228 |
author = {J{\"o}rg Tiedemann},
|
229 |
title = {Parallel Data, Tools and Interfaces in OPUS},
|
|
|
236 |
publisher = {European Language Resources Association (ELRA)},
|
237 |
isbn = {978-2-9517408-7-7},
|
238 |
language = {english}
|
239 |
+
}
|
240 |
+
```
|
241 |
|
242 |
### Contributions
|
243 |
|