Update README.md
Browse files
README.md
CHANGED
@@ -1,19 +1,17 @@
|
|
1 |
---
|
2 |
license: cc
|
3 |
-
task_categories:
|
4 |
-
- text-generation
|
5 |
language:
|
6 |
- eu
|
7 |
tags:
|
8 |
- Basque
|
9 |
- Euskara
|
10 |
-
pretty_name:
|
11 |
version: v1
|
12 |
release_date: 2024/02/27
|
13 |
size_categories:
|
14 |
- 100M<n<1B
|
15 |
---
|
16 |
-
# ZelaiHandi🌱: A Large
|
17 |
|
18 |
<img src="https://storage.googleapis.com/orai-nlp/datasets/Grass_Block.png" alt="grass" width="277" />
|
19 |
|
@@ -21,12 +19,12 @@ size_categories:
|
|
21 |
|
22 |
<!-- Provide a quick summary of the dataset. -->
|
23 |
ZelaiHandi🌱: A Large Open Dataset of Basque.
|
24 |
-
ZelaiHandi
|
25 |
-
|
26 |
|
27 |
The dataset will receive yearly updates.
|
28 |
|
29 |
-
|
30 |
|
31 |
## Dataset Details
|
32 |
|
@@ -43,7 +41,7 @@ The corpus is released with the aim of "feeding" LLMs like llama-2, Latxa, etc.
|
|
43 |
|
44 |
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
|
45 |
|
46 |
-
The corpus is released in
|
47 |
|
48 |
```json
|
49 |
{
|
@@ -53,7 +51,7 @@ The corpus is released in jsonl format. Each document contains the following pro
|
|
53 |
"url": "document url",
|
54 |
"title": "document title",
|
55 |
"author": "document author if known, null otherwise",
|
56 |
-
"date": "yyyy-mm-dd",
|
57 |
"text": "document content, ready to use raw text. Title text (for news and articles) is included as well.",
|
58 |
"domain": "document domain. one of ['news', 'administrative', 'wikipedia', 'science', 'subtitles', 'literature', 'videogames']"
|
59 |
}
|
@@ -75,17 +73,17 @@ The following table summarizes corpus statistics and license information.
|
|
75 |
| Berria | https://berria.eus | news | 178.77 | cc-by-sa 4.0 |
|
76 |
| Eusko Legebiltzarra | https://www.legebiltzarra.eus | administrative | 80 | public domain |
|
77 |
| Wikipedia | https://wikipedia.org | wikipedia | 63.8 | cc0 |
|
78 |
-
|
|
79 |
-
|
|
80 |
-
|
|
81 |
| GFA | https://www.bngipuzkoa.eus/ | administrative | 8.4 | Custom copyright license |
|
82 |
-
|
|
83 |
-
|
|
84 |
| BFA (+Nafarroa?) | https://jjggbizkaia.eus | administrative | 4.2 | Custom copyright license |
|
85 |
-
|
|
86 |
-
|
|
87 |
-
|
|
88 |
-
|
|
89 |
| **Total** | | | **566.36** |
|
90 |
|
91 |
\* Tokikom is a Basque local media network. The following websites are included in ZelaiHandi:
|
@@ -107,7 +105,7 @@ The corpus is comprised of documents from various sources. All documents include
|
|
107 |
## Bias, Risks, and Limitations
|
108 |
|
109 |
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
110 |
-
This corpus
|
111 |
|
112 |
<!--### Recommendations -->
|
113 |
|
|
|
1 |
---
|
2 |
license: cc
|
|
|
|
|
3 |
language:
|
4 |
- eu
|
5 |
tags:
|
6 |
- Basque
|
7 |
- Euskara
|
8 |
+
pretty_name: ZelaiHandi🌱
|
9 |
version: v1
|
10 |
release_date: 2024/02/27
|
11 |
size_categories:
|
12 |
- 100M<n<1B
|
13 |
---
|
14 |
+
# ZelaiHandi🌱: A Large Collection of Basque texts
|
15 |
|
16 |
<img src="https://storage.googleapis.com/orai-nlp/datasets/Grass_Block.png" alt="grass" width="277" />
|
17 |
|
|
|
19 |
|
20 |
<!-- Provide a quick summary of the dataset. -->
|
21 |
ZelaiHandi🌱: A Large Open Dataset of Basque.
|
22 |
+
ZelaiHandi, which means "large pasture" in Basque, is the largest collection of freely licensed and clean Basque texts to date (February 27th, 2024), gathered from selected web sources.
|
23 |
+
This collection comprises approximately 560 million tokens.
|
24 |
|
25 |
The dataset will receive yearly updates.
|
26 |
|
27 |
+
TThe corpus has been released with the objective of “feeding” Large Language Models. Naturally, models that are not strictly "herbivorous" can also ingest ZelaiHandi.
|
28 |
|
29 |
## Dataset Details
|
30 |
|
|
|
41 |
|
42 |
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
|
43 |
|
44 |
+
The corpus is released in JSONL format. Each document contains the following attributes:
|
45 |
|
46 |
```json
|
47 |
{
|
|
|
51 |
"url": "document url",
|
52 |
"title": "document title",
|
53 |
"author": "document author if known, null otherwise",
|
54 |
+
"date": "publication date in yyyy-mm-dd format",
|
55 |
"text": "document content, ready to use raw text. Title text (for news and articles) is included as well.",
|
56 |
"domain": "document domain. one of ['news', 'administrative', 'wikipedia', 'science', 'subtitles', 'literature', 'videogames']"
|
57 |
}
|
|
|
73 |
| Berria | https://berria.eus | news | 178.77 | cc-by-sa 4.0 |
|
74 |
| Eusko Legebiltzarra | https://www.legebiltzarra.eus | administrative | 80 | public domain |
|
75 |
| Wikipedia | https://wikipedia.org | wikipedia | 63.8 | cc0 |
|
76 |
+
| Addi | https://addi.ehu.es/ | science | 17 | cc-by-nc-sa / cc-by-nc-nd |
|
77 |
+
| Argia | https://www.argia.eus | news | 17.44 | cc-by-sa 4.0 |
|
78 |
+
| Opendata Euskadi subtiitles | https://opendata.euskadi.eus/catalogo/-/euskarazko-azpitituluak-filmak-eta-telesailak-euskaraz/ | subtitles | 11 | cc-by-sa |
|
79 |
| GFA | https://www.bngipuzkoa.eus/ | administrative | 8.4 | Custom copyright license |
|
80 |
+
| Zientzia.eus | https://zientzia.eus | science | 8.3 | cc-by-sa |
|
81 |
+
| Susa | https://www.susa-literatura.eus/ | literature | 5.8 | cc-by |
|
82 |
| BFA (+Nafarroa?) | https://jjggbizkaia.eus | administrative | 4.2 | Custom copyright license |
|
83 |
+
| Zientzia Kaiera | https://zientziakaiera.eus/ | science | 2 | cc-by-sa 3.0 |
|
84 |
+
| Ekaia | https://ojs.ehu.eus/index.php/ekaia | science | 1.8 | cc-by-nc-nd |
|
85 |
+
| Ikergazte | https://www.buruxkak.eus/bilatu?bilaketa=ikergazte | science | 1.7 | cc-by-sa 3.0 |
|
86 |
+
| Game-erauntsia | https://gamerauntsia.eus/ | videogames | 0.40 | cc-by-sa |
|
87 |
| **Total** | | | **566.36** |
|
88 |
|
89 |
\* Tokikom is a Basque local media network. The following websites are included in ZelaiHandi:
|
|
|
105 |
## Bias, Risks, and Limitations
|
106 |
|
107 |
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
108 |
+
This corpus may reflect societal stereotypes and biases. Additionally, it could contain explicit language, slurs, and hate speech, particularly in fictional texts from literature.
|
109 |
|
110 |
<!--### Recommendations -->
|
111 |
|