Recent developments in large-scale pre-trained text-to-image diffusion models have significantly improved the generation of high-fidelity images, particularly with the emergence of diffusion transformer models (DiTs). Among diffusion models, diffusion transformers have demonstrated superior image-generation capabilities, boosting lower FID scores and higher scalability. However, deploying large-scale DiT models can be expensive due to their excessive parameter numbers. Although existing research has explored efficient deployment techniques for diffusion models, such as model quantization, there is still little work concerning DiT-based models. To tackle this research gap, we propose TerDiT, the first quantization-aware training (QAT) and efficient deployment scheme for extremely low-bit diffusion transformer models. We focus on the ternarization of DiT networks, with model sizes ranging from 600M to 4.2B, and image resolution from 256256 to 512512. Our work contributes to the exploration of efficient deployment of large-scale DiT models, demonstrating the feasibility of training extremely low-bit DiT models from scratch while maintaining competitive image generation capacities compared to full-precision models. Our code and pre-trained TerDiT checkpoints have been released atthis https URL.
View on arXiv@article{lu2025_2405.14854,
title={ TerDiT: Ternary Diffusion Models with Transformers },
author={ Xudong Lu and Aojun Zhou and Ziyi Lin and Qi Liu and Yuhui Xu and Renrui Zhang and Xue Yang and Junchi Yan and Peng Gao and Hongsheng Li },
journal={arXiv preprint arXiv:2405.14854},
year={ 2025 }
}