Nitral-AI/Captain-Eris_Violet-GRPO-v0.420 is a language model developed by Nitral-AI, created through a multi-stage fine-tuning process involving supervised fine-tuning, QLoRA adapters, and multi-stage RLHF optimized with GRPO. This model is specifically designed for character interaction and roleplay scenarios, demonstrated by its integration with SillyTavern and its focus on reasoning block parsing and Mistral formatting. It is optimized for generating nuanced and contextually relevant responses in interactive narrative environments.
No reviews yet. Be the first to review!
Hugging Face