import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("OnomaAIResearch/Illustrious-xl-early-release-v0", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("LyliaEngine/Smooth_Allure")
prompt = "-"
image = pipe(prompt).images[0]Smooth_Allure
- Prompt
- -
- Prompt
- -
- Prompt
- -
Model description
This lora gives a visual alike to Pony smooth loras, but you get to enjoy the better instruction following ability from Illustrious base model.
Try the strength at around 0.8. The training material needs more diversity still, so one might need to account for that in the prompts for the first version. The base model looks to be trained fully on anime, therefore to get the smooth style from the showcase, use extra direction with "realistic face" or similar. As of now there are not explicit trigger words.
The base model is https://civitai.com/models/827184/wai-nsfw-illustrious-sdxl?modelVersionId=1761560
Source
https://civitai.red/models/1901188/smooth-allure-nsfw-illustrious?modelVersionId=2151972
Credit
https://civitai.red/user/EncodedExplorer
Trigger words
You should use realistic face to trigger the image generation.
Download model
Download them in the Files & versions tab.
- Downloads last month
- 24
Model tree for LyliaEngine/Smooth_Allure
Base model
KBlueLeaf/kohaku-xl-beta5