# K-Decision-9B Korean model, weights and results

> K-Decision-9B is a Korean-specialised Qwen3.5-9B fine-tune with public weights. Its card declares Apache-2.0 and reports Korean benchmark results.

Updated 2026-11-04 (Australia/Adelaide)

By ONYULTEA · First Hugging Face upload: 2026-11-03 · OPEN

A Korean-specialised 9B decision model based on Qwen3.5-9B, trained with int8 LoRA at rank 32. It reads a state and typed questions, then returns probabilities over the options in one forward pass. The first public release supports choice and yes/no questions through a Jev-compatible /v1/systemone server.

## Facts

- Size: 9B
- Base: Qwen/Qwen3.5-9B
- Licence: Apache-2.0, declared for the model weights
- Inputs: Text; choice questions with 2 to 26 options and yes/no questions
- Language: Korean-specialised; the model card also tags English
- Limits: State up to 6,000 characters; unsupported requests return HTTP 422
- Access: Public model.safetensors weights and serving code

## Claims and sources

- **VENDOR-REPORTED**: ONYULTEA reports 0.866 accuracy on a five-task KoBEST conversion with 2,600 items, and 0.635 on 3,000 korean-jev-decision-index items. Its LocalLLaMA typed-decisions score of 0.818 is labelled fitted. [source](<https://huggingface.co/ONYULTEA/K-Decision-9B>)
- **VENDOR-REPORTED**: On the public Decision Index 0.3 suite, ONYULTEA reports index 30.7, raw 44.8 and 168 ms p50 latency. It says training and serving used one 24 GB consumer GPU. [source](<https://huggingface.co/ONYULTEA/K-Decision-9B>)

## Links

- [Source: Weights, licence declaration and results](<https://huggingface.co/ONYULTEA/K-Decision-9B>)
- [Source: Serving and evaluation code](<https://github.com/onyultea/k-decision>)
- [Source: Weight file inventory](<https://huggingface.co/api/models/ONYULTEA/K-Decision-9B/tree/main>)

## Related projects

No related projects are listed in the registry.