Nitro Diffusion

🚀 Nitro Diffusion

Welcome to Nitro Diffusion - the world's first Multi-Style Model trained from scratch! It's a fine-tuned Stable Diffusion model. During training, it simultaneously incorporates three distinct art styles while keeping each style independent. This design enables high - level control over style mixing, weighting, and single - style usage.

Use the tokens archer style, arcane style or modern disney style in your prompts to achieve the desired effect. You can also combine multiple tokens for a mixed style, as demonstrated in the examples below.

If you appreciate my work and wish to test new models before their official release, please consider supporting me.

✨ Features

• Multi - Style Training: Trained on three art styles simultaneously, with each style kept separate, allowing for precise control over style mixing, weighting, and single - style use.
• Token - Based Styling: Use specific tokens in prompts to apply different art styles or create mixed styles.

🚀 Quick Start

Prerequisites

• Ensure you have the necessary Python libraries installed, such as diffusers and torch.

Usage

You can use the model in your Python code as follows:

from diffusers import StableDiffusionPipeline
import torch

model_id = "nitrosocke/nitro-diffusion"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
pipe = pipe.to("cuda")

prompt = "archer arcane style magical princess with golden hair"
image = pipe(prompt).images[0]

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

💻 Usage Examples

Prompt and settings for Gal Gadot

Prompt: arcane archer modern disney gal gadot Settings: Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 598801516, Size: 512x768

Prompt and settings for the Audi TT

Prompt: (audi TT car) arcane modern disney style archer Settings: Steps: 30, Sampler: Euler a, CFG scale: 7, Seed: 713702776, Size: 768x512

Model Samples

Multi Style Characters from the model: Single Style Characters from the model: Multi Style Scenes from the model:

Animated GIFs

You can find animated GIFs of Batman and Lara Croft showing the weighting and prompt influence on the bottom of the page.

• Batman:
• Lara Croft:

📚 Documentation

Gradio Web UI

We support a Gradio Web UI to run Nitro - Diffusion.

🧨 Diffusers

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:

1. You can't use the model to deliberately produce nor share illegal or harmful outputs or content.
2. The authors claim 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.
3. 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