SillyTavern-Presets-Sphiratrioth
| Property | Value |
|---|---|
| License | CC-BY-NC-3.0 |
| Language | English |
| Author | sphiratrioth666 |
What is SillyTavern-Presets-Sphiratrioth?
SillyTavern-Presets-Sphiratrioth is a comprehensive collection of presets designed for role-playing interactions in SillyTavern version 1.12.6+. It offers carefully crafted templates supporting various LLM architectures including Mistral, ChatML, LLAMA3, Alpaca, and Metharme/Pygmalion.
Implementation Details
The preset collection implements three core sampler parameters: Temperature, Min-P, and DRY, offering different creativity levels (0.7, 1.0, and 1.3 temperature variants) while maintaining response coherence. The system supports both 1st and 3rd person narration modes, configurable through system prompts.
- Organized file structure with clear naming conventions
- Multiple template compatibility for various LLM architectures
- Customizable narration styles and response lengths
- Built-in creativity control through sampler settings
Core Capabilities
- Conversation Mode: Minimal narration with ~150 token responses
- Roleplay Mode: Balanced storytelling with ~250 token responses
- Story Mode: Rich narration with ~350 token responses
- GM Mode: Tabletop RPG-style interactions with player-focused responses
Frequently Asked Questions
Q: What makes this model unique?
The preset collection stands out for its clear organization, comprehensive documentation, and versatile application across multiple LLM architectures. It offers precise control over AI responses while maintaining ease of use.
Q: What are the recommended use cases?
The presets are ideal for various roleplay scenarios, from casual conversations to complex storytelling sessions. They're particularly well-suited for users wanting consistent, well-structured AI responses in different narrative styles.
