QuantSparse: Comprehensively Compressing Video Diffusion Transformer with Model Quantization and Attention Sparsification
This project is the official implementation of our "QuantSparse: Comprehensively Compressing Video Diffusion Transformer with Model Quantization and Attention Sparsification".
- Our code will be released soon!
If you find QuantSparse is useful and helpful to your work, please kindly cite this paper:
@article{feng2025quantsparse,
title={QuantSparse: Comprehensively Compressing Video Diffusion Transformer with Model Quantization and Attention Sparsification},
author={Feng, Weilun and Yang, Chuanguang and Qin, Haotong and Wu, Mingqiang and Li, Yuqi and Li, Xiangqi and An, Zhulin and Huang, Libo and Zhang, Yulun and Magno, Michele and others},
journal={arXiv preprint arXiv:2509.23681},
year={2025}
}


