Skip to content

Tikai7/Diffusion-Model

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ddpm_04 ddpm_4 ddpm_8 ddpm_013 ddpm_014

Diffusion Model

  • This is an simple implementation of a Diffusion model (Stable diffusion) in Tensorflow & Keras.
  • This code is used in Kaggle

Dataset

Model

  • The model is built on a UNET architecture
  • We first generate a random image noisy at a certain time step, along with the accompanying noise.
  • We then feed the noisy image to the Unet model to try to denoise it, and calculate the error with the accompanying noise.

Result

  • Here is the result after 300 epochs : plot
  • Here is an example of a generated emoji : generated_emojis

About

This is an implementation of Stable diffusion

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published