| Property | Value |
|---|---|
| Parameter Count | 13B |
| License | LLaMA2 |
| Format | GGUF |
| Author | KoboldAI |
What is LLaMA2-13B-Psyfighter2-GGUF?
LLaMA2-13B-Psyfighter2-GGUF is a specialized merged model created by the KoboldAI community, combining creative fiction capabilities with medical knowledge. The model was developed through task arithmetic merging of multiple base models, including LLaMA2-13B-Tiefighter and medical-focused models, though it's important to note it should not be used for actual medical advice.
Implementation Details
The model utilizes a sophisticated merge recipe combining multiple models with specific weights, including LLaMA2-13B-Tiefighter (weight: 1.0), cat-v1.0-13b (weight: 0.01), and llama-2-13b-chat-limarp-v2-merged (weight: 0.02). The implementation uses float16 dtype for efficient processing.
- Task arithmetic merge methodology
- Multiple model integration with weighted contributions
- GGUF format optimization
- Float16 precision implementation
Core Capabilities
- Creative story writing and generation
- Multi-format instruction following (Alpaca and others)
- Chatbot and persona creation
- Adventure game text generation
- Basic medical knowledge integration (for fictional purposes only)
Frequently Asked Questions
Q: What makes this model unique?
The model's unique strength lies in its balanced merge of creative fiction capabilities with supplementary medical knowledge, making it particularly effective for creative writing while maintaining anatomical and psychological accuracy in fictional contexts.
Q: What are the recommended use cases?
The model excels in creative writing, story generation, character interaction, and adventure game scenarios. It's best used with minimal prompting, allowing its improvisation capabilities to shine. The model supports various instruction formats but performs best when not overwhelmed with excessive context.