Skip to content

Hands-on Colab that builds three generative models for MNIST digits: (1) an unconditional DC-GAN that learns the raw digit distribution, (2) a label-aware conditional GAN for class-controlled synthesis, and (3) a convolutional encoder that projects images back to their latent vector, enabling reconstructions and smooth latent edits In one notebook

Notifications You must be signed in to change notification settings

pier-maker92/GAN-MNIST

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 

Repository files navigation

GAN-MNIST

A step-by-step Colab notebook that shows how to generate, steer, and invert handwritten digits with three progressively richer models:

  • Unconditional DC-GAN – learns the data distribution and produces random digits.

  • Conditional GAN (cGAN) – adds class labels so you can ask for any specific numeral (0-9).

  • Encoder / Inference Net – maps a real image back to its latent vector z, enabling reconstructions and latent-space editing.

About

Hands-on Colab that builds three generative models for MNIST digits: (1) an unconditional DC-GAN that learns the raw digit distribution, (2) a label-aware conditional GAN for class-controlled synthesis, and (3) a convolutional encoder that projects images back to their latent vector, enabling reconstructions and smooth latent edits In one notebook

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published