How to use from the
Use from the
Diffusers library
pip install -U diffusers transformers accelerate
import torch
from diffusers import DiffusionPipeline
from diffusers.utils import load_image

# 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("DiffusionLight/TurboLoRA")

prompt = "a perfect mirrored reflective chrome ball sphere, a perfect black dark mirrored reflective chrome ball sphere"
input_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cat.png")

image = pipe(image=input_image, prompt=prompt).images[0]

DiffusionLight's TurboLoRA

Project Page | Code

[Open DiffusionLight in Colab

DiffusionLight-Turbo
30 seconds

DiffusionLight-Turbo
30 seconds

DiffusionLight-Turbo
30 seconds

DiffusionLight-Turbo
30 seconds

DiffusionLight
30 Minutes

DiffusionLight
30 Minutes

DiffusionLight
30 Minutes

DiffusionLight
30 Minutes

Normally exposed

Under​exposed

Normally exposed

Under​exposed

Normally exposed

Under​exposed

Normally exposed

Under​exposed

Usage

This repository is the weight for an example code that shows how to implement DiffusionLight-Turbo inpaint.py in Diffusers 0.33.0

We recommend checking out Github Repository: Github

Note that the Weight is the same as 'DiffusionLight/Flicker2k', but due to diffusers 0.33.0 breaking the compatibility, we have to store it in a separate model.

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.

Trigger words

Chromeball Prompt
Normally exposed a perfect mirrored reflective chrome ball sphere
Underexposed a perfect black dark mirrored reflective chrome ball sphere

Citation

@inproceedings{Chinchuthakun2025DiffusionLightTurbo,
  author = {Chinchuthakun, Worameth and Phongthawee, Pakkapon and Raj, Amit and Jampani, Varun and Khungurn, Pramook and Suwajanakorn, Supasorn},
  title = {DiffusionLight-Turbo: Accelerated Light Probes for Free via Single-Pass Chrome Ball Inpainting},
  booktitle = {ArXiv},
  year = {2025},
}

Visit us 🦉

Vision & Learning Laboratory VISTEC - Vidyasirimedhi Institute of Science and Technology

Downloads last month
64
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for DiffusionLight/TurboLoRA

Adapter
(8377)
this model