VariationalAutoEncoder-GENAI-Implementation

Variational Autoencoder (VAE) - GENAI Implementation

This project implements a Variational Autoencoder (VAE) for generating fashion samples using the Fashion-MNIST dataset. The VAE is trained on the dataset, and generated fashion samples can be visualized using the trained model.

Project Structure

The project is organized into several files:

• main.py: Entry point for running the VAE training and fashion sample generation.
• train_vae.py: Script for training the VAE model on the Fashion-MNIST dataset.
• vae_model.py: Implementation of the Variational Autoencoder model.
• generate_fashions.py: Script for generating and visualizing fashion samples using the trained VAE.

Usage

To train the VAE model and generate fashion samples, run the following commands:

python main.py

This will execute the training script and generate fashion samples using the trained model.

Model Configuration

• Latent Dimension: Set your desired latent dimension.
• Epochs: 10
• Batch Size: 128

Dependencies

• TensorFlow
• NumPy
• Matplotlib

Install dependencies using:

pip install tensorflow numpy matplotlib

Feel free to explore and modify the project to suit your needs!

Note: Make sure to replace "path/to/your/vae_model_weights.h5" with the actual path to the saved weights of your trained VAE model. Additionally, you may need to adjust other details based on your specific setup.