K-Decision-9B v2
Korean-specialized decision model by ONYULTEA
In collaboration with Hallym University
K-Decision is a lightweight decision engine that takes a state and typed questions as input and returns structured decisions with calibrated probability distributions over all options — all in a single forward pass. System One / Jev compatible.
Benchmarks
| Benchmark | Score |
|---|---|
| LocalLLaMA typed-decisions (2,000 decisions, fitted) | 0.8295 |
| KoBEST 5-task conversion (2,600 items) | 0.866 |
| Korean Life-Domain Holdout (7 domains) | 0.997 |
| korean-jev-decision-index (3,000 items) | 0.635 |
| Decision Index 0.3 public suite (140K requests) | Index 30.7 · 0 errors · p50 168ms |
- Trained and served on a single 24GB consumer GPU (int8 LoRA, r=32)
Quickstart
/v1/systemone server
git clone https://github.com/onyultea/k-decision
python src/serve_systemone.py --merged ONYULTEA/K-Decision-9B-v2 --port 8111
// Request
{"state": "Civil complaint: Strong gas smell (call center)",
"questions": {
"route": {"type": "choice", "instructions": "Which department?",
"criteria": {"gas": "Gas Safety", "water": "Waterworks", "road": "Roads"}},
"urgent": {"type": "noul", "instructions": "Urgent?"}}}
// Response — probability for every option
{"answers": {
"route": {"type": "choice", "choice": "gas",
"probabilities": {"gas": 0.87, "water": 0.05, "road": 0.08}},
"urgent": {"type": "noul", "noul": 0.93}}}
Acknowledgments
We sincerely thank Prof. Youngwoong Ko (School of Software, Hallym University) for his invaluable guidance and support throughout this fine-tuning research. This work was conducted as an industry–academia collaboration between ONYULTEA and Hallym University. In addition, Hallym University regulations were used as reference material for the compliance-decision training and evaluation in the development and research of K-Decision.
K-Decision-9B v2 (한국어)
ONYULTEA 한국 특화 디시전 모델
한림대학교 산학협력
상태(state)와 타이핑드 질문을 입력받아 모든 옵션에 보정된 확률 분포를 단일 forward 패스로 반환 (System One / Jev 호환).
성능
| 벤치마크 | 점수 |
|---|---|
| LocalLLaMA typed-decisions (2,000결정) | 0.8295 |
| KoBEST 5태스크 | 0.866 |
| 한국 생활 도메인 (7도메인) | 0.997 |
| korean-jev-decision-index | 0.635 |
| Decision Index 0.3 (14만 요청) | 지수 30.7 · 에러 0 |
빠른 시작
git clone https://github.com/onyultea/k-decision
python src/serve_systemone.py --merged ONYULTEA/K-Decision-9B-v2 --port 8111
감사의 글
본 연구의 파인튜닝 과정 전반에 걸쳐 귀한 조언과 지원을 보내주신 **한림대학교 소프트웨어융합대학 고영웅 교수님**께 진심으로 감사드립니다. 본 성과는 (주)온율티와 한림대학교의 산학협력으로 진행되었습니다. 또한 본 K-Decision의 개발·연구에는 한림대학교 규정이 규정 준거 판단(regulation compliance decision) 학습·평가의 참조 자료로 활용되었습니다.
© 2026 ONYULTEA × Hallym University · https://onyultea.kr