LGAI-EXAONE/KMMLU-Pro
General NLPKOBenchmark
Created by LGAI-EXAONE at 2025, the LGAI-EXAONE/KMMLU-Pro is a General NLP benchmark dataset in KO in Parquet format.
š This dataset is used as an LLM benchmark. See model leaderboards ā
About LGAI-EXAONE/KMMLU-Pro
KMMLU-Pro
š Paper | š„ļø Code
We introduce KMMLU-PRO, a challenging new benchmark comprising 2,822 problems from the official exams for Korean National Professional Licensure (KNPL), representing highly specialized professions in Korea. To bridge...
Details
- Task
- General NLP
- Language
- KO
- Format
- Parquet
- Rows / instances
- N/A
- Creator
- LGAI-EXAONE
- Year
- 2025