hausmer/ukr-wiki-events
Ukrainian Wikipedia events A small (1,722-row) Ukrainian dataset built from public-domain / Wikipedia-sourced text. Two task shapes are mixed in the single train split (distinguishable via the instruction prompt): Event extraction — instruction = a passage of Ukrainian Wikipedia text prefixed by "what important event is this text about:"; output = a short label of the salient event. Explanation / QA — instruction = a question or term (e.g. "Опиши явище поліплоїдії"); output =… See the full description on the dataset page: https://huggingface.co/datasets/hausmer/ukr-wiki-events.
Ukrainian Wikipedia events
A small (1,722-row) Ukrainian dataset built from public-domain / Wikipedia-sourced text. Two task shapes are mixed in the single train split (distinguishable via the instruction prompt):
- Event extraction —
instruction= a passage of Ukrainian Wikipedia text prefixed by "what important event is this text about:";output= a short label of the salient event. - Explanation / QA —
instruction= a question or term (e.g. "Опиши явище поліплоїдії");output= the corresponding Wikipedia excerpt / article text.
The style field is general_ua (a plain helpful-assistant persona), in contrast with the satirical truha_satire items in the companion truha-news-satire dataset — these rows are the general_ua subset of that corpus, published standalone.
Structure
License
Inputs and answer excerpts are drawn from the Ukrainian Wikipedia, licensed CC BY-SA 4.0. Event labels are short generated summaries. Respect the CC BY-SA attribution requirement when reusing Wikipedia text.
Notes
- These same 1,722 rows are also embedded in the
trainsplit ofhausmer/truha-news-satire; use this repo for the clean event/QA subset on its own.
