https://github.com/szymon-budziak/fashion_mnist_gan

Generative Adversarial Network (GAN) built with CNNs for both the generator and discriminator, implemented in TensorFlow and trained on the Fashion MNIST dataset.
https://github.com/szymon-budziak/fashion_mnist_gan

fashion-mnist gan generative-adversarial-network generative-ai python3 tensorflow

Last synced: 17 days ago
JSON representation

Generative Adversarial Network (GAN) built with CNNs for both the generator and discriminator, implemented in TensorFlow and trained on the Fashion MNIST dataset.

• Host: GitHub
• URL: https://github.com/szymon-budziak/fashion_mnist_gan
• Owner: Szymon-Budziak
• Created: 2024-08-13T19:27:10.000Z (6 months ago)
• Default Branch: main
• Last Pushed: 2024-08-29T19:15:16.000Z (5 months ago)
• Last Synced: 2024-11-25T04:13:05.706Z (3 months ago)
• Topics: fashion-mnist, gan, generative-adversarial-network, generative-ai, python3, tensorflow
• Language: Jupyter Notebook
• Homepage:
• Size: 2.75 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Fashion MNIST Generative Adversarial Network

This project is a simple implementation of a Generative Adversarial Network (GAN) using the Fashion MNIST dataset. The

GAN is implemented using Tensorflow and trained on the Fashion MNIST dataset. The GAN is trained to generate new images

of clothing items that are similar to the images in the Fashion MNIST dataset.

## Installation

The requirements of the project are listed in the `requirements.txt` file. To install the requirements, run the

following command:

1. Create a virtual environment and activate it:

```bash

python -m venv venv

source venv/bin/activate

```

2. Install the requirements:

```bash

pip install -r requirements.txt

```

## Usage

All the code for the project is in the `main.ipynb` notebook. Architectures of the generator and discriminator are

located in separate directories called `generator` and `discriminator` respectively. Code for training the GAN is

in the notebook.

## Results

The GAN was trained for 10000 epochs on the Fashion MNIST dataset with 32 batch size. The results of the training are:

- After 500 epochs

![500 epochs](src/images/gan_generated_image_epoch_500.png)

- After 1000 epochs

![1000 epochs](src/images/gan_generated_image_epoch_1000.png)

- After 5000 epochs

![5000 epochs](src/images/gan_generated_image_epoch_5000.png)

- After 9500 epochs

![10000 epochs](src/images/gan_generated_image_epoch_9500.png)