grimjim/kunoichi-lemon-royale-7B
grimjim/kunoichi-lemon-royale-7B is a 7 billion parameter language model merged using the DARE TIES method, based on SanjiWatsuki/Kunoichi-7B. This model is specifically optimized for narrative generation, excelling at developing character interiority and adhering to character cards. It is designed for creative writing applications, particularly those requiring nuanced character portrayal and strong adherence to initial prompts.
Loading preview...
Model Overview
grimjim/kunoichi-lemon-royale-7B is a 7 billion parameter language model created by grimjim through a merge of several pre-trained models using the DARE TIES method. It utilizes SanjiWatsuki/Kunoichi-7B as its base, integrating strengths from core-3/kuno-royale-v2-7b, SanjiWatsuki/Kunoichi-DPO-v2-7B, and KatyTheCutie/LemonadeRP-4.5.3.
Key Capabilities
- Narrative Generation: The model demonstrates a notable ability to delve into character interiority, making it suitable for rich storytelling.
- Character Adherence: It effectively follows character cards and takes strong cues from initial messages, maintaining consistent character portrayal.
- Context Handling: Tested to support a context length of 8K tokens, providing ample space for detailed interactions.
- Flexible Prompting: Works well with both Alpaca and ChatML prompting formats.
Use Cases
This model is particularly well-suited for:
- Creative Writing: Generating narrative content that requires deep character development and emotional depth.
- Roleplay Scenarios: Maintaining consistent character personas and responding appropriately to prompts in interactive storytelling.
- Story Generation: Crafting stories where character motivations and internal thoughts are central to the plot.