Soft Pastel Anime Model

Introduction

The Soft Pastel Anime model is a highly stylized text-to-image model designed to generate anime-style illustrations with a watercolor undertone. It is compatible with Stable Diffusion and can be enhanced using LoRA (Low-Rank Adaptation). The model operates with or without specific trigger words, such as "daiton" for generating mythological creatures.

Architecture

The model is built on the stabilityai/stable-diffusion-xl-base-1.0 framework. It utilizes a combination of diffusers and LoRA to achieve its unique artistic style. The model's architecture supports text-to-image generation, leveraging advanced diffusion techniques to produce high-quality images from text prompts.

Training

Details about the specific training process are not provided, but the model is intended for research purposes. Users are encouraged to contact the creator for any commercial applications, ensuring adherence to usage policies and licensing agreements.

Guide: Running Locally

1. Download the Model: Access the model weights in Safetensors format from the Files & Versions tab on the Hugging Face page.
2. Installation: Ensure you have the necessary dependencies installed, including PyTorch and the Hugging Face Transformers library.
3. Set Up Environment: Clone the repository and set up your environment to run the model locally.
4. Generate Images: Use example prompts like "a man with dark hair" or "a nymph like creature from Greek myth, daiton" to test the model's output.

For optimal performance, it is recommended to use cloud GPU services such as AWS, Google Cloud, or Azure.

License

The Soft Pastel Anime model is released under the CreativeML OpenRAIL-M license. This license allows for research use, but commercial use requires direct permission from the creator, alvdansen.