The aim of the project is to predict middle 16x16 segment of every test image. To run the project first create virtual environment and install dependencies by running: python -m venv my_venv source my_env/bin/activate pip install -r requirements.txt

Each directory contains config file with necessary configurations to replicate the results. To train the network run train.py script in needed directory. Inference.py makes predictions on test set. Full report is also available in the repository.