Autoencoders, GANs and PINNs

This is another project from the Deep learning course in Taltech autumn 2024. For the other one see https://github.com/mikk-kruusalu/deep_learning_project.

In order to use this code, one should create a virtual environment with python -m venv venv and activate it. On Linux, the activation is done with source venv/bin/activate. The the required packages should be installed with

pip install -r requirements.txt

To add GPU acceleration for Jax, see the install guide. Also, Weights & Biases account is needed for autoencoders and GANs, the login on the command line is done with wandb login. The wandb project can be configured via the .yaml config files.

Please find the reports for the course in the root of this repository.