Cyberpunk Anime Diffusion

An AI model that generates cyberpunk anime characters!~

Based of a finetuned Waifu Diffusion V1.3 Model with Stable Diffusion V1.5 New Vae, training in Dreambooth

by DGSpitzer

🧨 Diffusers

This repo contains both .ckpt and Diffuser model files. It's compatible to be used as any Stable Diffusion model, using standard Stable Diffusion Pipelines.

You can convert this model to ONNX, MPS and/or FLAX/JAX.

#!pip install diffusers transformers scipy torch
from diffusers import StableDiffusionPipeline
import torch

model_id = "DGSpitzer/Cyberpunk-Anime-Diffusion"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
pipe = pipe.to("cuda")

prompt = "a beautiful perfect face girl in dgs illustration style, Anime fine details portrait of school girl in front of modern tokyo city landscape on the background deep bokeh, anime masterpiece, 8k, sharp high quality anime"
image = pipe(prompt).images[0]

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

Online Demo

You can try the online demo using HF space:

https://huggingface.co/spaces/DGSpitzer/DGS-Diffusion-Space

Or use Colab Notebook at here:

https://colab.research.google.com/github/HelixNGC7293/cyberpunk-anime-diffusion/blob/main/cyberpunk_anime_diffusion.ipynb

Buy me a coffee if you like this project ;P ♥

https://www.buymeacoffee.com/dgspitzer

Usage

After model loaded, use keyword dgs in your prompt, with illustration style to get even better results.

For sampler, use Euler A for the best result (DDIM kinda works too), CFG Scale 7, steps 20 should be fine

Example 1:

portrait of a girl in dgs illustration style, Anime girl, female soldier working in a cyberpunk city, cleavage, ((perfect femine face)), intricate, 8k, highly detailed, shy, digital painting, intense, sharp focus

For cyber robot male character, you can add muscular male to improve the output.

Example 2:

a photo of muscular beard soldier male in dgs illustration style, half-body, holding robot arms, strong chest

Example 3 (with Stable Diffusion WebUI):

If using AUTOMATIC1111's Stable Diffusion WebUI

You can simply use this as prompt with Euler A Sampler, CFG Scale 7, steps 20, 704 x 704px output res:

an anime girl in dgs illustration style

And set the negative prompt as this to get cleaner face:

out of focus, scary, creepy, evil, disfigured, missing limbs, ugly, gross, missing fingers

This will give you the exactly same style as the sample images above.

NOTE: usage of this model implies accpetance of stable diffusion's CreativeML Open RAIL-M license