Welcome to your guide on using the Atlantis-v0.1-12B model, an innovative AI fine-tuned for role-playing and storytelling. This guide will provide a comprehensive understanding of how to harness this powerful model for your creative endeavors, enabling you to craft rich narratives and memorable characters. Let’s dive in!

Understanding the Atlantis-v0.1-12B Model

Atlantis 12B is built on the foundation of the Mistral-Nemo-2407-Instruct model, designed specifically for an enhanced role-playing and storytelling experience. Imagine it as a seasoned bard, equipped with countless stories and the ability to spin them into captivating narratives for your RPG sessions or writing projects. This AI not only understands context but can respond coherently throughout long interactions, aiming to make your storytelling immersive and engaging.

Features of Atlantis-v0.1-12B

• Supports up to 128K context lengths, although maintaining coherence might be challenging over 24K.
• Multiple output formats are available for your convenience, such as FP16 Safetensors and Static GGUF.
• Valuable for both narrative creation and dialog generation, allowing for a collaborative storytelling experience.

Usage: Getting Started

To effectively utilize the Atlantis model, follow these guidelines:

1. Software Requirements

Make sure your backend software is updated. Use:

• llama.cpp b3436 or newer
• koboldcpp 1.71 or newer
• oobabooga 1.11 or newer

If the GGUF fails to load, ensuring software updates should resolve this issue.

2. Prompt Format

When constructing prompts, the recommended format is as follows:

[INST]Please write me a song.[INST]

This format has proven most effective, although variations can be experimented with.

3. Sampler Settings

To achieve the best results, particularly in character dialogue or story progression, consider beginning with these settings:

• Temp: 0.3-0.5
• Top-P: 1
• Rep Pen: 1.13-1.16
• Response Length: 350+

Longer response lengths tend to yield richer content.

Example of Storytelling with Atlantis Model

Imagine presenting the model with a story prompt such as:

Write me a 3,000-word opening chapter of a gritty hard sci-fi novel, drawing inspiration from Isaac Asimov and Andy Weir.

In this case, Atlantis behaves much like a co-author, using your input to expand into a detailed narrative that maintains the tone and style of the mentioned authors, creating an expansive universe filled with dynamic characters and intricate plot lines.

Troubleshooting Common Issues

As with any technology, you may encounter a few hiccups along the way:

• Response Coherence Issues: If the model appears incoherent, especially beyond 24K context, consider shortening your prompts.
• Loading Errors: Ensure that your software is updated to the required versions listed earlier.
• Quality of Output: If the output isn’t meeting your expectations, try adjusting the sampling parameters for different narrative lengths and styles.

If you experience persistent problems or have feedback, please consider reaching out through The Community Tab or join the SillyTavern Discord thread. For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.

Final Thoughts

By utilizing the Atlantis-v0.1-12B model, you can enhance your storytelling capabilities, creating engaging plots that captivate your audience. Remember that storytelling, much like any art form, takes practice and patience, so don’t shy away from experimenting with different styles and prompts!

At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.