This model is a specialized variant of Mistral-Small-24B-Base-2501, adapted using a DeepSeek R1 distillation process. It is primarily designed for roleplay (RP) and storywriting applications, focusing on character interactions, narrative generation, and creative storytelling. Approximately 40% of the training dataset consists of roleplay/storywriting/character card data, ensuring rich and contextually immersive outputs in these domains.
Fine-Tuning Approach: DeepSeek R1 process (focused on RP)
Dataset Size: The dataset used in training doubled since the last version, adding more neutral logs, training the Base model to stick more on my new format.
Intended Use
Primary Use Cases:
Roleplay (RP): Engaging with users in fictional or scenario-based interactions.
Storywriting: Generating narratives, character dialogues, and creative texts.
Character Lore Generation: Serving as a resource to craft or expand on character backstories and interactions.
How To Use:
User-First Message: The first message in any interaction should come from the user, ensuring the model responds in a narrative or roleplay context guided by user input.
Contextual Information: User or assistant details can be placed either in the system prompt or the user's first message. A system prompt is not mandatory, but any contextual instructions or role descriptions can help set the stage.
DeepSeek-Style Interaction: The model can also be used purely as a DeepSeek distill without additional system prompts, providing flexible usage for direct storytelling or roleplay scenarios. The model still can be biased toward Roleplay data, and it is expected.
Training Data
DeepSeek R1 Thinking Process: The model inherits a refined chain-of-thought (thinking process) from DeepSeek R1, which places heavy emphasis on roleplay and narrative coherence.
Dataset Composition:
40%: RP/Storywriting/Character Cards
60%: Various curated data for broad language, math, logical, space... understanding
Data Scaling: The dataset size was doubled compared to previous iterations, which enhances the model’s creative and contextual capabilities.
Model Performance
Strengths:
Storytelling & Roleplay: Rich in creative generation, character portrayal, and scenario building.
Dialogue & Interaction: Capable of sustaining engaging and context-driven dialogues.
Adaptability: Can be used with or without a system prompt to match a range of user preferences.
Limitations & Bias:
Hallucination: It can generate fictitious information in the thinking process, but still end up with a succesfull reply.
Thinking can be dismissed: Being a distillation of DeepSeek R1 is essence, this model, even trained on Base, could forget to add <think>\n in some scenario.
Ethical Considerations
Yes
Usage Recommendations
System Prompt (Optional):
You may provide a high-level system prompt detailing the scenario or the desired style of roleplay and storywriting. Example: "You are a friendly fantasy innkeeper who greets travelers from distant lands."
User’s First Message:
Must clearly state or imply the scenario or context if no system prompt is provided. Example: "Hello, I’m a wandering knight seeking shelter. Could you share a story about local legends?"
Roleplay & Storywriting Focus:
Encourage the model to develop characters, backstories, and immersive dialogues.
For more direct, unfiltered or freeform creativity, skip the system prompt.
If you still want to have some "logs" from previous message before starting a conversation, put them in the first user message, or in the system prompt.
You can put exemple message of the character you RP with in the system prompt, too.
