https://github.com/vasukalariya/generative-modelling

Implementation of various generative models using PyTorch.
https://github.com/vasukalariya/generative-modelling

conditional-gan dcgan gan generative-model image-reconstruction mnist neural-network numpy pytorch rbm restricted-boltzmann-machine vae variational-autoencoder

Last synced: 11 months ago
JSON representation

Implementation of various generative models using PyTorch.

• Host: GitHub
• URL: https://github.com/vasukalariya/generative-modelling
• Owner: vasukalariya
• License: apache-2.0
• Created: 2020-04-05T15:10:58.000Z (almost 6 years ago)
• Default Branch: master
• Last Pushed: 2020-09-05T06:51:04.000Z (over 5 years ago)
• Last Synced: 2025-01-21T13:08:49.871Z (about 1 year ago)
• Topics: conditional-gan, dcgan, gan, generative-model, image-reconstruction, mnist, neural-network, numpy, pytorch, rbm, restricted-boltzmann-machine, vae, variational-autoencoder
• Language: Jupyter Notebook
• Homepage:
• Size: 2.19 MB
• Stars: 3
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Restricted Boltzmann Machines

![RBM Image](rbm.png)

## Implementation Details

* The implemented RBM consist of 2 layers with 784 visible units and 256 hidden units.

* Trained using K-Contrastive Divergence with k = 1.

* The RBM can be modified as per the use.

* It is built solely from scratch using NumPy.

* It is trained for MNIST Reconstruction and learning its probability distribution.

* Trained on 60000 images.

* It is modular in nature so you could change the parameters as per your choice.

## How to Use

* Download the train_images from [here](https://raw.githubusercontent.com/sebastianlapuschkin/lrp_toolbox/master/data/MNIST/train_images.npy)

* For sample demonstration a clipped image is reconstructed using trained RBM.

* The notebook is ready to run.

# Generative Adversarial Networks

![GAN Image](gan.jpg)

# Variational Autoencoders

![VAE Image](vae.png)

## For learning

* To learn the concepts please watch the [videos](https://www.youtube.com/watch?v=lXrFX3vjtjQ&list=PL3pGy4HtqwD2kwldm81pszxZDJANK3uGV&index=135)

* The notations are also similar.

### Have Fun! Learning :smiley: