ml-aos

Machine Learning methods for Rubin AOS system

Installation

1. Make sure Poetry and Conda/Mamba are installed.
2. Create a new conda environment, activate it, and run poetry install
3. If developing, install the pre-commit hooks: pre-commit install

Running the code

Training and exporting models is handled by the scripts in bin/.

• test_train.py - runs quick sanity checks to make sure the model is set up correctly
• train_wavenet.py - trains the network
• export_model.py - exports the model in torchscript format

Notebooks

Some basic analysis is present in notebooks/. More analysis is in the aos_notebooks repo.