Miggsoo3/MadelisVale-LoRA

Madelis Vale LoRA v2 (SDXL Refined)

This LoRA is trained for photorealistic editorial portraits based on Stable Diffusion XL Base 1.0.

• Base model: stabilityai/stable-diffusion-xl-base-1.0
• LoRA weights: Miggsoo3/MadelisVale-LoRA/pytorch_lora_weights.safetensors
• Training: DreamBooth LoRA (SDXL, rank 16, fp16, ~10k steps)

🧩 Usage

Example Python snippet:

from diffusers import StableDiffusionXLPipeline
pipe = StableDiffusionXLPipeline.from_pretrained(
    "stabilityai/stable-diffusion-xl-base-1.0"
)
pipe.load_lora_weights("Miggsoo3/MadelisVale-LoRA")

prompt = "a portrait photo of Madelis woman, cinematic lighting, ultra realistic skin"
image = pipe(prompt).images[0]
image.save("madelis_vale_output.png")

🖼️ Example

Coming soon: preview render from SDXL fine-tuned LoRA v2.