Indexnusrefather/Qwen-3.5-2b-roleplay-tuned

Quick introduction

This is my experimental tune of Qwen 3.5 2b, aimed at enhanced roleplay and writing. It was trained with 5M~ tokens of high quality roleplay data, being generally an improvement over the base model.

V2:

v2 is ready and I already have it released, it was trained on way more data and is generally supposed to be better.