import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("bdsqlsz/stable-diffusion-xl-anime-final", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]
Thank you for support my work.
https://www.buymeacoffee.com/bdsqlsz
Support list will show in main page.
Support List
DiamondShark
Yashamon
t4ggno
Someone
kgmkm_mkgm
yacong
awdii_
Jeremy Cai
Introduction
stable-diffusion-xl-anime-final, base on animagine-xl-3.1 to sft(Supervised Fine-tuning) and SPO with merging.
Fix anatomy with Sliders such as good hands.
Adjust saturation and lighting
More anime coloring and better anatomy
Less 3D and highlight
You can download from here、Civitai、TensorArt、Shakker AI and Yodayo
Usage
This repo includes 2 checkpoints and 2 loras.
The checkpoints include a standard version and a CM accelerated version.
Recommended configuration:
Quality prompt:
masterpiece, best quality, absurdres, newest,
Quality negative prompt:
lowres, (bad), text, error, fewer, extra, missing, worst quality, jpeg artifacts, low quality, watermark, unfinished, displeasing, oldest, early, chromatic aberration, signature, extra digits, artistic error, username, scan,copyright name,copyright notice,
Standard version
Sampler:DPM++2M/Euler a
Scheduler: AYS/GITS for Steps:10
Karras for Steps:28
CFG:5
Highfix steps: half of steps
CM accelerated version:
Sampler:Euler a/TCD
Scheduler:SGMuniform
Step:6~8
CFG:2.5
Highfix steps: half of steps
License
Same with animagine-xl-3.1 using fair-ai-public-license-1.0-sd and Apache 2.0 But especially in merging model useage, MUST show this model merge ratio in merged model.
- Downloads last month
- 70