Datasets:

Modalities:
Text
Formats:
parquet
Languages:
Japanese
ArXiv:
Libraries:
Datasets
pandas
License:
babylm-jpn / README.md
nikitastheo's picture
Upload README.md with huggingface_hub
3ba6c77 verified
metadata
task_categories:
  - text-generation
language:
  - jpn
license: unknown
size_categories:
  - 1K<n<10K
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

Dataset Summary

  • Language: jpn
  • Script: Hira, Jpan, Kana
  • Tier: 10M
  • Byte Premium Factor: 1.321974
  • Size (MB): 71.78
  • Expected Size (MB): 71.78
  • Number of Documents: 2,043
  • Total Tokens: 16,524,324
  • Tokenizer: tohoku-nlp/bert-base-japanese

Tokens Per Category

  • child-books: 9,712,521 tokens
  • educational: 291,053 tokens
  • padding-opensubtitles: 6,520,750 tokens

Tokens Per Group

  • Transcription: 0 tokens
  • Education: 291,053 tokens
  • Books, Wiki, News: 9,712,521 tokens
  • Subtitles: 0 tokens
  • Padding: 6,520,750 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

Wikis

Data Curators

n/a

Comments

None