Marker-Inc-Korea / RAGchain

Extension of Langchain for RAG. Easy benchmarking, multiple retrievals, reranker, time-aware RAG, and so on...
Apache License 2.0
274 stars 28 forks source link

Create KoDuoRC dataset benchmark #360

Closed minsing-jin closed 9 months ago

minsing-jin commented 9 months ago

Create KoDuoRC dataset benchmark. This dataset is DuoRC translated in Korean by KETI-AIR. This issue is related to issue #357 .

Dataset link

https://huggingface.co/datasets/KETI-AIR/kor_duorc

minsing-jin commented 9 months ago

DuoRC dataset is for MRC task like SQuad. So it is inappropriate for our RAGchain.