Undi95/MLewdBoros-LRPSGPT-1Char-13B
Undi95/MLewdBoros-LRPSGPT-1Char-13B is a 13 billion parameter language model based on a merge of MLewdBoros and ShareGPT-13b-qloras. This model is specifically fine-tuned for single-character roleplay scenarios, designed to interact with character cards containing a "ONE PERSONA" definition. Its primary strength lies in generating consistent and immersive responses for individual character interactions.
Loading preview...
Overview
Undi95/MLewdBoros-LRPSGPT-1Char-13B is a 13 billion parameter model created by Undi95, resulting from a merge of the MLewdBoros base model with specific ShareGPT-13b-qloras. This particular iteration is specialized for single-character roleplay, focusing on maintaining a consistent persona defined by a character card.
Key Capabilities
- Single-Character Roleplay: Optimized to embody and maintain a single character's persona throughout interactions.
- Persona Consistency: Designed to adhere strictly to a "ONE PERSONA" definition provided in character cards.
- Custom Prompt Template: Utilizes a custom prompt structure to define the character's role, persona, and scenario, enhancing roleplay immersion.
Good For
- Interactive Storytelling: Creating engaging narratives where the model plays a single, well-defined character.
- Character Simulation: Developing chatbots or virtual companions that consistently portray one specific personality.
- Roleplay Scenarios: Users looking for a model that excels at maintaining a singular character's traits and responses based on detailed persona descriptions.
For more detailed information on the LoRA used and prompt template, refer to the LimaRP-ShareGPT-13b-qloras documentation.