Hugging Face Overview
Cydonia 24B v2 Overview
Cydonia 24B v2, developed by BeaverAI, is a 24 billion parameter language model fine-tuned from Mistral's 'Small' model (2501). This iteration focuses on enhancing capabilities for detailed and immersive roleplay and creative writing scenarios.
Key Capabilities
- Extended Context Stability: Tested to maintain stability and coherence up to a 24,000 token context length, making it suitable for long-form interactions.
- Rich Narrative Detail: Users report improved ability to hold intricate details, anatomy, and complex scene descriptions within generated text.
- Expansive Vocabulary: Demonstrates a broad and nuanced vocabulary, contributing to more expressive and varied outputs.
- Roleplay Optimization: Specifically noted for its strong performance in roleplay scenarios, handling character consistency and narrative flow effectively.
Recommended Usage
Cydonia 24B v2 is particularly well-suited for applications requiring:
- Long-form creative writing: Generating detailed stories, fanfiction, or complex narrative arcs.
- Immersive roleplaying: Engaging in extended, character-driven conversational roleplay where context retention and descriptive richness are crucial.
- Content generation: Producing text that requires a high degree of descriptive detail and vocabulary depth.