This Dataset is part of a series of datasets aimed at advancing Turkish LLM Developments by establishing rigid Turkish benchmarks to evaluate the performance of LLM's Produced in the Turkish Language.
malhajar/hellaswag-turkish is a translated version of hellaswag aimed specifically to be used in the OpenLLMTurkishLeaderboard
This Dataset contains rigid tests extracted from the paper Can a Machine Really Finish Your Sentence? published at ACL2019.… See the full description on the dataset page:
https://huggingface.co/datasets/malhajar/hellaswag-tr.