---
license: cc
task_categories:
- text-generation
language:
- eu
tags:
- Basque
- Euskara
pretty_name: Zelai Handi🌱
version: v1
release_date: 2024/02/27
size_categories:
- 100M<n<1B
---
# Dataset Card for ZelaiHandi🌱🟩

<!-- Provide a quick summary of the dataset. -->
ZelaiHandi🌱: A Large Open Dataset of Basque. 
ZelaiHandi (large pasture in Basque) is the largest open corpus publicly available in Basque (as of 2024/02/27 ). 

The corpus is released with the aim of "feeding" LLMs like llama-2, Latxa, etc. Of course other non "herbivore" models can also use  ZelaiHandi.

## Dataset Details

### Dataset Description

<!-- Provide a longer summary of what this dataset is. -->

- **Size:** 559M words.
- **Language(s) (NLP):** EU
- **License:** CC
- **Curated by:** Orai NLP Technologies

<!--## Uses -->

<!-- Address questions around how the dataset is intended to be used. -->

<!--The main aim of this dataset is to be used to pretrain fundational LLMs for Basque and their tokenizers. -->

<!--### Direct Use -->

<!-- This section describes suitable use cases for the dataset. -->

<!-- [More Information Needed] -->

<!-- ### Out-of-Scope Use -->

<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->

<!--  [More Information Needed] -->

## Dataset Structure

<!-- 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. -->

The corpus is released in jsonl format. Each document contains the following properties:


```json
{
        "source": "source", 
        "license": "document license", 
        "lang": "eu", 
        "url": "document url", 
        "title": "document title", 
        "author": "document author if known, source otherwise", 
        "date": "yyyy-mm-dd", 
        "text": "document content, ready to use raw text. Title text (for news and articles) is included as well.",
        "domain": "document domain. one of ['news', 'administrative', 'wikipedia', 'science', 'subtitles', 'literature', 'videogames']"
}
```


## Dataset Creation

### Source Data

<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
This is a corpus of Basque language. It contains 559.380.531 raw tokens. It was gathered from various web sources and contains data up to 2024/02/23. 
The corpus is comprised of documents from various sources. All documents included have a free license, but each source has its own license. The following table summarizes corpus statistics and license information. 

TODO: domain sartu eta admin banatu

| **source**                  | **base url**                                                                                    | **domain**     | **tokens (M)** | **license**                 |
|----------------------------:|------------------------------------------------------------------------------------------------:|---------------:|---------------:|----------------------------:|
| Tokikom                     | https://tokikom.eus/bazkideak/                                                                  | news           | 165,75         | cc-by / cc-by-sa            |
| Berria                      | https://berria.eus                                                                              | news           | 178,77         | cc-by-sa 4.0                |
| Eusko Legebiltzarra         | https://www.legebiltzarra.eus                                                                   | administrative | 80             | public domain               |
| Wiki                        | https://wikipedia.org                                                                           | wikipedia      | 63,8           | cc0                         |
| addi                        | https://addi.ehu.es/                                                                            | science        | 17             | cc-by-nc-sa / cc-by-nc-nd   |
| argia                       | https://www.argia.eus                                                                           | science        | 17,44          | cc-by-sa 4.0                |
| opendata euskadi subtiitles | https://opendata.euskadi.eus/catalogo/-/euskarazko-azpitituluak-filmak-eta-telesailak-euskaraz/ | subtitles      | 11             | cc-by-sa                    |
| GFA                         | https://www.bngipuzkoa.eus/                                                                     | administrative | 8,4            | Custom copyright license    |
| zientzia.eus                | https://zientzia.eus                                                                            | science        | 8,3            | cc-by-sa                    |
| susa                        | https://www.susa-literatura.eus/                                                                | literature     | 5,8            | cc-by                       |
| BFA (+Nafarroa?)            | https://jjggbizkaia.eus                                                                         | administrative | 4,2            | Custom copyright license    |
| zientzia kaiera             | https://zientziakaiera.eus/                                                                     | science        | 2              | cc-by-sa 3.0                |
| ekaia                       | https://ojs.ehu.eus/index.php/ekaia                                                             | science        | 1,8            | cc-by-nc-nd                 |
| ikergazte                   | https://www.buruxkak.eus/bilatu?bilaketa=ikergazte                                              | science        | 1,7            | cc-by-sa 3.0                |
| game-erauntsia              | https://gamerauntsia.eus/                                                                       | videogames     | 0,40           | cc-by-sa                    |
| **Total**                   |                                                                                                 |                | **566,36**     |

\* Custom license, allows reproducibility. License text here: https://www.bngipuzkoa.eus/WAS/CORP/DJGPortalWEB/aviso_legal.jsp

\*\* Custom license, allows reproducibility License text here: https://jjggbizkaia.eus/eu/pribadutasun-politikea

Data collection efforts were performed between 2024/01/01-2024/02/23.

## Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->
This corpus inherits the Basque society's stereotypes and biases. Moreover, it might include explicit language, slurs and hate speech, (mostly on the fictional text from literature and subtitles).

<!--### Recommendations -->

<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->

<!--Users should be made aware of the risks, biases and limitations of the dataset. 
We suggest filtering sources as literature and subtitles that might contain hate speech, slurs or explicit language before pretraining an LLM for end tasks like interactive chat.-->

## Citation

<!--We plan to release a paper soon, but, for now, please, cite as follows:

**BibTeX:**

@misc{ZelaiHaundi,
 title     = "{Zelai-Haundi: A Big Open Dataset to Feed the next Gen of Basque LLMs}",
 author    = "{Iñaki, Gorka, Xabi, Ander}",
 url       = "https://huggingface.co/datasets/orai-nlp/ZelaiHaundi",    
 year = 2024 }
-->

## Dataset Card Contact

- Iñaki San Vicente (i.sanvicente@orai.eus)
- Gorka Urbizu (g.urbizu@orai.eus)