import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("Tongyi-MAI/Z-Image-Turbo", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("RomixERR/Age_Slider_Z-Images-turbo")
prompt = "-"
image = pipe(prompt).images[0]Age Slider for Z Images turbo
.jpg)
- Prompt
- -
.jpg)
- Prompt
- -
Model description
(-) Younger / (+) Older There are no trigger words. SHIFT 2.5
V1.1
A retrained version of the previous version. Fixed an issue with noise and low contrast in older images (when generating with +2 and a simple prompt). If you need to generate younger images (-1...-3), try both versions; they produce different results.
V1
Tested with weights from -2 to +2. Use at your own risk. The author is not responsible for the data you generate. It connects like a regular LOR.
This is my first LOR slider, so don't judge me too harshly. I trained it without splitting the dataset.
Download model
Download them in the Files & versions tab.
- Downloads last month
- 320
Model tree for RomixERR/Age_Slider_Z-Images-turbo
Base model
Tongyi-MAI/Z-Image-Turbo