sd-naruto-diffusers Open Source Model - Free Text-to-Image Conversion in Naruto Style

Sd Naruto Diffusers

Developed by lambdalabs

A Stable Diffusion model fine-tuned for Naruto style, capable of converting input text into Naruto-style images

Image Generation English#Anime style transfer #Naruto style #Text-to-image generation

Downloads 115

Release Time : 10/27/2022

Model Overview

This model is a variant of the Stable Diffusion model, specifically fine-tuned for the Naruto style. It can generate images with Naruto characteristics, including character portraits and scenes, based on text prompts.

Model Features

Naruto style transfer

Capable of converting various subjects (such as celebrities, fictional characters, etc.) into Naruto-style images

Prompt engineering optimization

More Naruto-style results can be achieved through specific prompt formats (e.g., 'character name ninja portrait')

Efficient training

Training can be completed in just 12 hours on 2xA6000 GPUs, costing approximately $20

Model Capabilities

Text-to-image generation

Style transfer

Creative image generation

Use Cases

Entertainment & Creativity

Celebrity Naruto-fication

Convert real-world celebrities into Naruto-style characters

Generate portraits of celebrities with forehead protectors and ninja outfits

Fictional character style transfer

Convert characters from other works (e.g., Marvel, Star Wars) into Naruto style

Generate character images with Naruto characteristics

Creative design

Design original Naruto-style characters or scenes

Generate unique Naruto-style images

🚀 Naruto diffusion

Stable Diffusion fine tuned on Naruto by Lambda Labs. It enables you to generate Naruto - style images with text prompts.

Try the live [text - to - naruto demo here](https://huggingface.co/spaces/lambdalabs/text - to - naruto)!
If you want more details on how to train your own Stable Diffusion variants, see this [example](https://github.com/LambdaLabsML/examples/tree/main/stable - diffusion - finetuning).

🚀 Quick Start

Put in a text prompt and generate your own Naruto style image!

Examples of Style Transformation

Game of Thrones to Naruto

Marvel to Naruto

✨ Features

Prompt engineering matters

We find that prompt engineering does help produce compelling and consistent Naruto style portraits. For example, writing prompts such as 'person_name ninja portrait' or 'person_name in the style of Naruto' tends to produce results that are closer to the style of Naruto character with the characteristic headband and other elements of costume.

Here are a few examples of prompts with and without prompt engineering that will illustrate that point.

Bill Gates:

Without prompt engineering

With prompt engineering

A cute bunny:

Without prompt engineering

With prompt engineering

📦 Installation

To run model locally:

!pip install diffusers==0.3.0
!pip install transformers scipy ftfy

💻 Usage Examples

Basic Usage

import torch
from diffusers import StableDiffusionPipeline
from torch import autocast

pipe = StableDiffusionPipeline.from_pretrained("lambdalabs/sd-naruto-diffusers", torch_dtype=torch.float16)  
pipe = pipe.to("cuda")

prompt = "Yoda"
scale = 10
n_samples = 4

# Sometimes the nsfw checker is confused by the Naruto images, you can disable
# it at your own risk here
disable_safety = False

if disable_safety:
  def null_safety(images, **kwargs):
      return images, False
  pipe.safety_checker = null_safety

with autocast("cuda"):
  images = pipe(n_samples*[prompt], guidance_scale=scale).images

for idx, im in enumerate(images):
  im.save(f"{idx:06}.png")

📚 Documentation

Model description

Trained on [BLIP captioned Naruto images](https://huggingface.co/datasets/lambdalabs/naruto - blip - captions) using 2xA6000 GPUs on [Lambda GPU Cloud](https://lambdalabs.com/service/gpu - cloud) for around 30,000 step (about 12 hours, at a cost of about $20).