SouthParkStyle Open-Source Text-to-Image Model - Free Generation of South Park Style Images

Southparkstyle

Developed by Guizmus

A text-to-image generation model based on RunwayML 1.5 with updated VAE, specialized in generating South Park-style images.

Image Generation EnglishOpen Source License:Openrail #Cartoon style generation #South Park style #Text-to-image generation

Downloads 155

Release Time : 11/16/2022

Model Overview

This model can generate South Park-style images, with special optimizations for character and scene representations, including Rick Roll and full-body shots.

Model Features

South Park Style Optimization

Specially optimized for South Park characters and scenes to generate stylistically consistent images.

Specific Token Optimization

Special optimizations for tokens like 'Rick Roll' and 'FullBody Shot' for better generation results.

High-Quality Dataset

Trained with 99 carefully selected images, each repeated 800 times to ensure model quality.

Model Capabilities

Text-to-image generation

Image-to-image generation

Stylized image generation

Use Cases

Entertainment

Generate South Park-style characters

Can generate various South Park-style characters, such as Master Yoda.

The generated character images maintain stylistic consistency with South Park animations.

Generate Rick Roll images

Specially optimized for generating Rick Roll images.

The generated Rick Roll images have a unique and expected style.

Creative Design

Stylized scene design

Can be used to generate South Park-style scenes and backgrounds.

The generated scenes feature the distinctive visual style of South Park.

🚀 SouthPark Style

This is a text-to-image and image-to-image model based on Stable Diffusion, which can generate images in the South Park style.

🚀 Quick Start

This model was based on RunwayML 1.5 model with updated VAE. The dataset is made mostly of South Park characters and scenery, with some regularization around Rick Roll and Fullbody Shots photography. There are 99 total pictures in the dataset, with 800 repeats each, over 16 Epoch on LR1e - 6 (around 79k unitary steps). It was trained using EveryDream with a full caption of all training pictures.

The style will be called by the use of the token SouthPark Style, and the tokens Rick Roll and FullBody Shot have also been refined.

To access this model, you can download the CKPT file below.

Showcase

📦 Installation

Downloads

💻 Usage Examples

Basic Usage

from diffusers import StableDiffusionPipeline
import torch

model_id = "Guizmus/SouthParkStyle"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
pipe = pipe.to("cuda")

prompt = "Yoda SouthPark Style"
image = pipe(prompt).images[0]

image.save("./SouthParkStyle.png")

Advanced Usage

This model can be used just like any other Stable Diffusion model. For more information, please have a look at the Stable Diffusion.

You can also export the model to ONNX, MPS and/or FLAX/JAX.

📄 License

This model is open access and available to all, with a CreativeML OpenRAIL - M license further specifying rights and usage. The CreativeML OpenRAIL License specifies:

You can't use the model to deliberately produce nor share illegal or harmful outputs or content
The authors claims no rights on the outputs you generate, you are free to use them and are accountable for their use which must not go against the provisions set in the license
You may re - distribute the weights and use the model commercially and/or as a service. If you do, please be aware you have to include the same use restrictions as the ones in the license and share a copy of the CreativeML OpenRAIL - M to all your users (please read the license entirely and carefully) Please read the full license here

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご