README.md

metadata

title: Text To Image DDGAN
emoji: 🐢
colorFrom: red
colorTo: purple
sdk: gradio
sdk_version: 3.8.2
app_file: app.py
pinned: false

Text-to-Image Denoising Diffusion GANs is a text-to-image model based on Denoising Diffusion GANs https://arxiv.org/abs/2112.07804. The code is based on their official code <https://nvlabs.github.io/denoising-diffusion-gan/, which is updated to support text conditioning. Many thanks to the authors of DDGAN for releasing the code.

The provided models are trained on DiffusionDB https://arxiv.org/abs/2210.14896, which is a dataset that was synthetically generated with Stable Diffusion, many thanks to the authors for releasing the dataset.

Models were trained on JURECA-DC supercomputer at Jülich Supercomputing Centre (JSC), many thanks for the compute provided to train the models.