metadata
license: apache-2.0
tags:
- merge
- mergekit
- lazymergekit
base_model:
- bunnycore/Qwen3-4B-Exp2
- rockerBOO/qwen3-4b-roleplay-lora
Qwen3-4B-RP-Exp
Qwen3-4B-RP-Exp is a 4-billion parameter language model merged using mergekit. It is designed as a lightweight, specialized model focusing on creative roleplay.
Unlike larger reasoning-focused models, this model prioritizes instruction following in character and functional coding tasks over complex logical deduction or multi-step reasoning chains.