In this project, I defined and trained a DCGAN on a dataset of faces. The goal of this project is to get a generator network to generate new images of faces that look as realistic as possble. The image below is a result of the training:
- Clone the repository and navigage to the downloaded folder.
git clone https://github.com/Lynchez/Face-Generation cd Face-Generation
- Open the
dlnd_face_generation.ipynb
file. Of course, you can find HTML version of the file.jupyter notebook dlnd_face_generation.ipynb
- Pre-processed Data
- Create a DataLoader
- Define the Model
- Discriminator
- Generator
- Initialize the weights of your network
- Build complete network
- Discriminator and Generator Losses
- Optimizers
- Training
- Training Loss
- Generator samples from training