from diffusers import StableDiffusionXLPipeline
import torch

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

# Load LoRA weights
pipe.load_lora_weights("runchat/lora-533d7b31-63fd-42a0-be75-b68de7db171f-RunchatTectonic")
pipe = pipe.to("cuda")

# Generate image
prompt = "a photo of RunchatTectonic style"
image = pipe(prompt, num_inference_steps=25, guidance_scale=7.5).images[0]
image.save("output.png")

Training Details

Base model: stabilityai/stable-diffusion-xl-base-1.0
Training steps: 1000
Learning rate: 0.0001
Batch size: 1
LoRA rank: 32
Trigger word: RunchatTectonic