metinovadilet/

kyrgyz-bench-data-wikiann-ner

	Kyrgyz WikiANN NER Dataset




	
	Task

Named Entity Recognition (Token Classification)

WikiANN NER dataset translated from English to Kyrgyz. Each entry contains English tokens with NER tags and a corresponding Kyrgyz translation. Entity names (proper nouns) are preserved as-is in the Kyrgyz text, so the NER tags align across both languages.

3 entity types: PER, ORG, LOC Format: JSONL with fields: tokens, ner_tags, sentence, sentence_ky… See the full description on the dataset page: https://huggingface.co/datasets/metinovadilet/kyrgyz-bench-data-wikiann-ner.