Domain Adaptation for Galaxy Morphology Classification using llustrisTNG and Galaxy Zoo Evolution dataset
This work was presented at NeurIPS 2025 Machine Learning and the Physical Sciences workshop. The paper can be found: https://arxiv.org/abs/2511.18590
- Python 3.10 or higher
uvpackage manager
- Clone this
repo
git clone https://github.com/kaleybrauer/galaxy_da.git- Install dependencies:
make install- or you can create a
.venv:
python3 -m venv .venv
source .venv/bin/activate # On mac/linux distros- Install
nebula
pip install -e .- Create a config file, see template and run with
python3 scripts/run_train.py --config /path/to/config.ymlpython3 scripts/run_eval.py /path/to/ckptYou can also run train and evaluate simultaneously. Run this with a single config, multiple configs, or a folder of configs by passing -f:
./run_experiment.sh <config_path> [more_configs...]
# or for a folder of configs:
./run_experiment.sh -f <config_folder>This project was made possible through the 2025 IAIFI Summer School provided by The NSF AI Institute for Artificial Intelligence and Fundamental Interactions (IAIFI).