metadata

license: apache-2.0

Kurai Toori Dark Streets - Quantized Model

This repository contains a quantized model used to synthetically generate the Kurai Toori Dark Streets dataset. The model is optimized for use with stable-diffusion.cpp for efficient running on lower-end hardware.

Model Details

Format: GGUF (GPT-Generated Unified Format)
Use Case: Generating high-quality images with lower hardware requirements
Compatibility: Designed for use with stable-diffusion.cpp

Key Features

Optimized for lower-end hardware
Capable of producing high-quality images
Specifically used to generate the Kurai Toori Dark Streets dataset
Efficient performance due to quantization

Usage

To use this model with stable-diffusion.cpp:

Clone the stable-diffusion.cpp repository:

git clone https://github.com/leejet/stable-diffusion.cpp.git
cd stable-diffusion.cpp

Follow the setup instructions in the stable-diffusion.cpp README.
Download the GGUF model file from this repository.
Run the model using the stable-diffusion.cpp interface, pointing to the downloaded GGUF file.

Dataset Generation

This model was used to create the Kurai Toori Dark Streets dataset, which features synthetically generated high-quality images. While the dataset focuses on dark and atmospheric street scenes, the model itself is capable of generating a wide range of high-quality images. The dataset can be found here.

Performance

Due to its quantized nature, this model offers:

Reduced memory usage
Faster inference times
Ability to run on less powerful hardware without significant quality loss

License

This model is released under the Apache 2.0 license. Please see the LICENSE file for more details.

Acknowledgements

Thanks to the creators of stable-diffusion.cpp for providing an efficient C++ implementation of Stable Diffusion.
Credit to the Hugging Face community for hosting the dataset and fostering open-source AI development.

For any questions or issues, please open an issue in this repository.