| license: mit | |
| pipeline_tag: image-to-image | |
| library_name: diffusers | |
| <h1 align="center"> REPA-E: Unlocking VAE for End-to-End Tuning of Latent Diffusion Transformers </h1> | |
| <p align="center"> | |
| [🌐 Project Page](https://end2end-diffusion.github.io)   | |
| [📃 Paper](https://arxiv.org/abs/2504.10483)   | |
| [🤗 Github](https://github.com/REPA-E/REPA-E) | |
| </p> | |
|  | |
| REPA-E enables stable and effective joint training of both the VAE and the diffusion model, significantly accelerating training and improving generation quality. It achieves state-of-the-art FID scores on ImageNet 256×256. For detailed usage instructions, including environment setup, training, and evaluation, please refer to the [project page](https://end2end-diffusion.github.io) and the [GitHub repository](https://github.com/REPA-E/REPA-E). |