from diffusers import StableDiffusionXLPipeline
import torch

# Load the base pipeline
pipeline = StableDiffusionXLPipeline.from_pretrained(
    "stabilityai/stable-diffusion-xl-base-1.0",
    torch_dtype=torch.float16
)

# Load this LoRA
pipeline.load_lora_weights("YOUR_USERNAME/YOUR_REPO_NAME")

# Generate image
image = pipeline(
    "professional portrait of a person, high quality",
    num_inference_steps=25,
    guidance_scale=7.5
).images[0]

image.save("result.png")

Model Details

Base Model: SDXL Base 1.0
LoRA Rank: 32
LoRA Alpha: 16
Training Images: 14 images
File Size: ~12 MB

Recommended Settings

LoRA Scale: 0.7-0.9
CFG Scale: 7-8
Steps: 25-30
Resolution: 1024x1024

Example Prompts

"professional portrait of a person in business attire, corporate headshot"
"casual portrait of a person wearing a hoodie, urban setting"
"elegant portrait of a person in formal evening wear, studio lighting"

waleriawatson/my-custom-lora

Custom SDXL LoRA Model

Usage

Model Details

Recommended Settings

Example Prompts

waleriawatson/
my-custom-lora