---
task_categories:
- text-generation
language:
- ita
license: unknown
size_categories:
- 10K<n<100K
dataset_info:
  features:
    - name: text
      dtype: string
    - name: doc-id
      dtype: string
    - name: category
      dtype: string
    - name: data-source
      dtype: string
    - name: script
      dtype: string
    - name: age-estimate
      dtype: string
    - name: license
      dtype: string
    - name: misc
      dtype: string
    - name: num-tokens
      dtype: int64
    - name: language
      dtype: string
---

# BabyLM Dataset

## Dataset Description

This dataset is part of the BabyLM multilingual collection.   
More information at: [babylm.github.io/babybabellm](https://babylm.github.io/babybabellm/)

### Dataset Summary

- **Language:** ita
- **Script:** Latn
- **Tier:** 10M
- **Byte Premium Factor:** 1.066887
- **Size (MB):** 57.96
- **Expected Size (MB):** 57.93
- **Number of Documents:** 16,575
- **Total Tokens:** 9,467,687
- **Tokenizer:** separate by whitespace

### Tokens Per Category

- **child-available-speech:** 160,051 tokens
- **child-books:** 5,616,333 tokens
- **child-directed-speech:** 1,029,580 tokens
- **child-wiki:** 1,180,821 tokens
- **educational:** 990,522 tokens
- **padding-opensubtitles:** 471,404 tokens
- **simplified-text:** 18,976 tokens

### Tokens Per Group

- **Transcription:** 1,189,631 tokens
- **Education:** 990,522 tokens
- **Books, Wiki, News:** 6,797,154 tokens
- **Subtitles:** 0 tokens
- **Padding:** 490,380 tokens


### Data Fields

- `text`: The document text
- `doc-id`: Unique identifier for the document
- `category`: Type of content (e.g., child-directed-speech, educational, etc.)
- `data-source`: Original source of the data
- `script`: Writing system used (ISO 15924)
- `age-estimate`: Target age or age range
- `license`: Data license
- `misc`: Additional metadata (JSON string)
- `num-tokens`: Number of tokens per item (based on white-space split)
- `language`: Language code (ISO 639-3)

### Licensing Information

Please see license in individual documents

### Data Sources & Attribution

#### Books
- Biancoenero Edizioni
	 - source: https://www.biancoeneroedizioni.it/
	 - description: Collection of books from the independent Italian publishing house Biancoenero Edizioni
- Logos
	 - source: https://children.logoslibrary.eu/
	 - description: Collection of children's stories
- Fairy tales (Italian)
	 - source: https://aclanthology.org/2024.clicit-1.46/
	 - description: Collection offairy tales curated in the study by Fusco et al., 2024
- Project Gutenberg
	 - source: https://www.gutenberg.org/
	 - description: Collection of books for children and young adults

#### Educational
- INVALSI
	 - source: https://www.invalsi.it/invalsi/index.php
	 - description: Exam tests in Italian and Math at both primary and secondary levels
- Esami di Stato
	 - source: https://www.mim.gov.it/
	 - description: Archive of national high school final exams
- YouTube – Calimero
	 - source: https://www.youtube.com/@calimeroitaliano2815
	 - description: YouTube video transcripts from the animated cartoon Calimero
- Terence and Teacher
	 - source: https://aclanthology.org/2025.coling-main.281/
	 - description: Text simplification datasets curated by Brunato et al., 2015
- MLSP2024
	 - source: https://aclanthology.org/2024.readi-1.4.pdf
	 - description: Lexical simplification dataset developed for the MLSP2024 shared task by Shardlow et al., 2024

#### Educational (songs)
- Zecchino D’Oro
	 - source: https://arxiv.org/abs/2503.09481
	 - description: Collection of children’s songs from the Zecchino D’Oro archive

#### Transcription
- Children’s Acquisition of Nouns and Verbs in Italian: Contrasting The Roles of Frequency and Positional Salience in Maternal Language
	 - source: https://pmc.ncbi.nlm.nih.gov/articles/PMC5822718/
	 - description: Child-directed transcripts collected by Longobardi et al., 2015
- L’INSEGNAMENTO DELLA GRAMMATICA NELLA CLASSE MULTILINGUEUn esperimento di focus on form nella scuola primaria
	 - source: https://riviste.unimi.it/index.php/promoitals/article/view/5049
	 - description: Child-directed transcripts collected by Whittle and Nuzzo, 2015
- Editorial: Are there different types of child-directed speech? Dynamic variations according to individual and contextual factors
	 - source: https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2022.1056816/full
	 - description: Child-directed transcripts collected by Spinelli et al., 2023
- MagicHub
	 - source: https://magichub.com/datasets/
	 - description: Transcribed adult conversational speech
- VolIP
	 - source: https://aclanthology.org/L14-1693/
	 - description: Dataset of telephone conversations curated by Alfano et al., 2014

#### Wikis
- Wikibooks (Italian)
	 - source: https://it.wikibooks.org/wiki/Italiano
	 - description: Age-appropriate books for children
- Wikijunior (Italian)
	 - source: https://en.wikibooks.org/wiki/Wikijunior:Languages/Italian
	 - description: Online encyclopedia based on wiki, adapted for children

### Data Curators

* Francesca Padovani



### Comments 

None