Stable Diffusion fine tuned on Romance of the Three Kingdoms XI: Officer Portraits.

Put in a text prompt and generate your own Officier in Three Kingdoms.

trained using this script with this dataset.

a man in armor

a women in red dress

a women in armor

try in colab.

Usage

!pip install diffusers==0.19.3
!pip install transformers scipy ftfy

import torch
from diffusers import StableDiffusionPipeline
from torch import autocast

pipe = StableDiffusionPipeline.from_pretrained("wx44wx/sd-three-kingdoms-diffusers", torch_dtype=torch.float16)  
pipe = pipe.to("cuda")

prompt = "a man in armor"
scale = 3
n_samples = 4

# Sometimes the nsfw checker is confused by the Pokémon 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")

Model description

Trained on BLIP captioned Three Kingdoms Officers images using 1xA6000 GPUs for around 16,000 steps.

Links

• Lambda Diffusers
• Captioned Three Kingdoms dataset
• Model weights in Diffusers format
• Original model weights
• Training code

Trained by Xin Wang. Thanks kongming.net for their archived images and justinpinkney for the code.