import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("ProomptEngineer/pe-holding-sign-concept")
prompt = "PEHoldingSign"
image = pipe(prompt).images[0]PE Holding Sign [Concept]
If you want to donate:
https://ko-fi.com/proomptengineer
Holding Sing pretty obvious what it does.
Made this to test text generation in SDXL.
Works after couple of tries but sometimes generates weird artefacts or wrong text but better than SD 1.5.
Add "holding a sign that says" for more consistency.
Weight of 0.8-1 recommended.
Image examples for the model:
- Downloads last month
- 10
Model tree for ProomptEngineer/pe-holding-sign-concept
Base model
stabilityai/stable-diffusion-xl-base-1.0