Hoshikuzu commited on
Commit
b7a43fe
·
verified ·
1 Parent(s): fa63613

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +90 -0
README.md CHANGED
@@ -29,3 +29,93 @@ configs:
29
  - split: validation
30
  path: data/validation-*
31
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
29
  - split: validation
30
  path: data/validation-*
31
  ---
32
+
33
+ # Dataset Card for OPUS-100-en-ja
34
+
35
+ ### Dataset Summary
36
+
37
+ This corpus is extracted from **[Helsinki-NLP/opus-100](https://huggingface.co/datasets/Helsinki-NLP/opus-100)**, with Japanese and English pairs.
38
+
39
+ ### How to use
40
+
41
+ It is used in much the same way as **[Helsinki-NLP/opus-100](https://huggingface.co/datasets/Helsinki-NLP/opus-100)**. The only difference is that you do not have to specify the language.
42
+
43
+ ```
44
+ from datasets import load_dataset
45
+ dataset = load_dataset("Hoshikuzu/opus-100-en-ja")
46
+ ```
47
+ If data loading times are too long and boring, use Streaming.
48
+
49
+ ```
50
+ from datasets import load_dataset
51
+ dataset = load_dataset("Hoshikuzu/opus-100-en-ja", streaming=True)
52
+ ```
53
+
54
+ ## Dataset Structure
55
+
56
+ ### Data Instances
57
+
58
+ ```
59
+ {
60
+ 'translation': {
61
+ 'en': 'Yeah, Vincent Hanna.',
62
+ 'ja': '- ラウール - ラウールに ヴィンセント・ハンナだ'
63
+ }
64
+ }
65
+ ```
66
+
67
+ ### Data Fields
68
+ Translation dictionaries containing texts from languages 1 and 2.
69
+
70
+ ### Data Splits
71
+ Only a `train` split is provided.
72
+
73
+ ### Citation Information
74
+
75
+ Follow the instructions described in the Helsinki-NLP/opus-100 readme. The following is taken from Helsinki-NLP/opus-100:
76
+
77
+ If you use this corpus, please cite the paper:
78
+ ```bibtex
79
+ @inproceedings{zhang-etal-2020-improving,
80
+ title = "Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation",
81
+ author = "Zhang, Biao and
82
+ Williams, Philip and
83
+ Titov, Ivan and
84
+ Sennrich, Rico",
85
+ editor = "Jurafsky, Dan and
86
+ Chai, Joyce and
87
+ Schluter, Natalie and
88
+ Tetreault, Joel",
89
+ booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
90
+ month = jul,
91
+ year = "2020",
92
+ address = "Online",
93
+ publisher = "Association for Computational Linguistics",
94
+ url = "https://aclanthology.org/2020.acl-main.148",
95
+ doi = "10.18653/v1/2020.acl-main.148",
96
+ pages = "1628--1639",
97
+ }
98
+ ```
99
+ and, please, also acknowledge OPUS:
100
+ ```bibtex
101
+ @inproceedings{tiedemann-2012-parallel,
102
+ title = "Parallel Data, Tools and Interfaces in {OPUS}",
103
+ author = {Tiedemann, J{\"o}rg},
104
+ editor = "Calzolari, Nicoletta and
105
+ Choukri, Khalid and
106
+ Declerck, Thierry and
107
+ Do{\u{g}}an, Mehmet U{\u{g}}ur and
108
+ Maegaard, Bente and
109
+ Mariani, Joseph and
110
+ Moreno, Asuncion and
111
+ Odijk, Jan and
112
+ Piperidis, Stelios",
113
+ booktitle = "Proceedings of the Eighth International Conference on Language Resources and Evaluation ({LREC}'12)",
114
+ month = may,
115
+ year = "2012",
116
+ address = "Istanbul, Turkey",
117
+ publisher = "European Language Resources Association (ELRA)",
118
+ url = "http://www.lrec-conf.org/proceedings/lrec2012/pdf/463_Paper.pdf",
119
+ pages = "2214--2218",
120
+ }
121
+ ```