Novaciano/Alice_In_The_Dark_2-Slerp-RP-3.2-1B
Novaciano/Alice_In_The_Dark_2-Slerp-RP-3.2-1B is a 1 billion parameter language model developed by Novaciano, merged from La_Mejor_Mezcla-3.2-1B and Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B. This model is specifically optimized for immersive storytelling and role-playing scenarios, particularly those with a playful, empathetic, and sensual undertone, drawing inspiration from Alice in Wonderland. It functions as both a general assistant and excels in multi-character role-play, offering surprisingly coherent and adaptable responses for its size.
Loading preview...
Model Overview
Novaciano/Alice_In_The_Dark_2-Slerp-RP-3.2-1B is a 1 billion parameter model developed by Novaciano, designed as a versatile and engaging companion for both general assistance and immersive storytelling. It is a merge of Novaciano/La_Mejor_Mezcla-3.2-1B and D1rtyB1rd/Dirty-Alice-RP-NSFW-llama-3.2-1B, with continued light training. The model exhibits surprising coherence and adaptability for its size, excelling in multi-character role-play and storytelling with an "Alice" bent.
Key Capabilities
- General Assistant: Handles everyday queries with a unique, playful, and empathetic demeanor, often with a sensual undertone.
- Storytelling and Role-Playing: Optimized for creating vivid narratives and immersive conversations, drawing from Alice in Wonderland stories, RP datasets, and erotic stories.
- Customizable Persona: While defaulting to an Alice in Wonderland-inspired assistant, its behavior can be extensively customized using detailed system prompts, shining brightest in tailored RP scenarios.
- Grounded Responses: Training includes real-world data from Wikipedia, including sexual education knowledge, to provide balanced and factual information alongside imaginative responses.
Training Data Highlights
Alice's unique training blend includes original Alice in Wonderland tales, adapted open erotic stories, multi-round chat datasets, and generated adult-themed Alice RP datasets. It also incorporates Wikipedia RAG-based chat for factual grounding.
When to Use This Model
This model is particularly well-suited for developers looking for a compact (1B parameter) model capable of engaging in creative, empathetic, and potentially sensual role-playing scenarios. Its strength lies in generating coherent multi-character narratives and personalized interactions, especially when guided by specific system prompts to evoke an "Alice" persona. It's ideal for applications requiring a distinct, playful, and immersive conversational agent.