Hugging Face's logo

Datasets:
SlayerLab
/
polish-dynaword

Tasks:
Text Generation
Languages:
Polish
Size:
1M<n<10M
ArXiv:
Tags:
polish
pretraining
dynaword
License:
Dataset card Files
xet
 Community
9
polish-dynaword / data /wikivoyage /wikivoyage.md
kacperwikiel's picture
kacperwikiel
Squash: collapse LFS history to reclaim private storage
cfe406c
preview code
|
Raw
History Blame
1.14 kB

wikivoyage

Polish Wikivoyage (travel guides)

Dataset description

  • Source (upstream): https://pl.wikivoyage.org/
  • Domain: travel
  • Language: Polish (pl)
  • License: CC-BY-SA-3.0
  • Created (range): 2012-01-01, 2025-12-31
  • Added: 2026-06-15

Licensing — traceable basis

Wikivoyage content is released under CC-BY-SA 3.0; attribution + share-alike preserved.

Provenance

Fetched directly from the official Wikimedia dump (plwikivoyage-latest-pages-articles) by src/fetch_wikimedia.py; mainspace only, wikitext stripped via mwparserfromhell.

Statistics

documents characters tokens (tiktoken proxy)
13,645 45,184,300 17,128,200

Filters applied (build_dynaword.py)

Minimal, per Dynaword guidelines (heavy filtering left to downstream use):

  • drop documents < 200 chars: 0
  • drop non-Polish (diacritic ratio): 9
  • exact cross-source dedup (sha1): 1
  • OCR alpha-ratio < 0.70 (OCR sources only): 0
  • read 13,655 → kept 13,645

Token counts are a fast tiktoken (cl100k) proxy (~1% off Llama-3); the canonical Llama-3 count is computed at release.