Skip to content

Latest commit

 

History

History
21 lines (11 loc) · 749 Bytes

README.md

File metadata and controls

21 lines (11 loc) · 749 Bytes

Binary Latent Diffusion

This repo contains the Pytorch implementation of Binary Latent Diffusion.

We currently provide:

  • a text-to-image generation demo under text-to-image-demo.

  • training instructions and models for binary autoencoders under binary-ae

  • training instructions and models for Bernoulli diffusion under bernoulli-diffusion

We will upload more scripts and models later.

This implementation is based on the following projects:

https://github.com/samb-t/unleashing-transformers

https://github.com/ehoogeboom/multinomial_diffusion

This implementation still requires further cleaning and commenting, If you have any questions, please reach out to the author at [email protected]