Darkest muse v1

Introduction

Darkest-muse-v1 is a creative writing model developed from a merging process involving different models. It is designed to produce unique and uncensored prose with a strong creative flair. This model uses the Gutenberg3 dataset, focusing on genres like fiction, sci-fi, fantasy, romance, and adventure.

Architecture

Darkest-muse-v1 is fine-tuned from the Gemma-2-9b-it model and includes a blend of other models such as Quill-v1, Delirium-v1, and Ataraxy-v2. The model aims to balance natural human-like writing with an imaginative and sometimes dark narrative style. Its primary source models, Quill-v1 and Delirium-v1, contribute to its distinctive voice by combining natural prose with intense, creative expressions.

Training

The model was trained using the Gutenberg3 dataset, which includes public domain fiction from 629 novels. This training process involved merging the strengths of different models to achieve a balance between natural prose and bold creativity. Delirium-v1, known for its intense and overtrained style, adds a unique edge to the writing output, which may be adjusted using custom anti-slop lists.

Guide: Running Locally

1. Setup Environment: Install necessary dependencies such as Python and libraries for model deployment.
2. Download Model: Access the model files and configurations from the Hugging Face repository.
3. Install Dependencies: Ensure all required packages are installed, typically using a package manager like pip.
4. Run the Model: Use your preferred deep learning framework to load and run the model.
5. Optimize Performance: Consider using cloud GPUs from providers like AWS, Google Cloud, or Azure to enhance performance during heavy computations.

License

Darkest-muse-v1 is released under the Apache 2.0 License, allowing for both personal and commercial use, modification, and distribution.