This is a project inspired by the work I did for the company that I work for. The goal of this project is to create a golden dataset from your or your company's data, and then use it to benchmark different embedding models.

Korean-optimized RAG evaluation toolkit with Kiwi tokenizer, ROUGE metrics, and IR evaluation for retrieval systems (Hit@K, NDCG@K, MRR, etc.)

🔍 Benchmark embedding models by creating custom datasets to evaluate and compare their performance effectively.