mit-han-lab/svdq-flux.1-schnell-pix2pix-turbo

This repository has been migrated to https://huggingface.co/nunchaku-tech/nunchaku-flux.1-schnell-pix2pix-turbo and will be hidden in December 2025.

<p align="center" style="border-radius: 10px"> <img src="https://huggingface.co/datasets/nunchaku-tech/cdn/resolve/main/nunchaku/assets/nunchaku.svg" width="30%" alt="Nunchaku Logo"/> </p>

Model Card for svdq-flux.1-schnell-pix2pix-turbo

This repository contains img2img-turbo LoRAs for both original and Nunchaku-quantized FLUX.1-schnell to translate sketch to images from user prompts.

Model Details

Model Description

• —Developed by: Nunchaku Team, CMU Generative Intelligence Lab
• —Model type: image-to-image
• —License: apache-2.0
• —Quantized from model: FLUX.1-schnell

Model Files

• —`sketch.safetensors`: Img2img sketch-to-image LoRA for original FLUX.1-schnell model.
• —`svdq-int4-sketch.safetensors`: Img2img sketch-to-image LoRA for SVDQuant INT4 FLUX.1-schnell model.

Model Sources

• —Inference Engine: nunchaku
• —Training Repo: img2img-turbo
• —Paper: SVDQuant | Img2img-Turbo
• —Demo: svdquant.mit.edu

Usage

See https://github.com/nunchaku-tech/nunchaku/tree/main/app/flux.1/sketch.

Citation

@inproceedings{
  li2024svdquant,
  title={SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models},
  author={Li*, Muyang and Lin*, Yujun and Zhang*, Zhekai and Cai, Tianle and Li, Xiuyu and Guo, Junxian and Xie, Enze and Meng, Chenlin and Zhu, Jun-Yan and Han, Song},
  booktitle={The Thirteenth International Conference on Learning Representations},
  year={2025}
}
@article{
  parmar2024one,
  title={One-step image translation with text-to-image models},
  author={Parmar, Gaurav and Park, Taesung and Narasimhan, Srinivasa and Zhu, Jun-Yan},
  journal={arXiv preprint arXiv:2403.12036},
  year={2024}
}