The code is implemented in Python using PyTorch and PyTorch Geometric.
-
Clone the repository:
git clone https://github.com/SMACY2017/NLO.git cd NLO -
Install dependencies:
pip install -r requirements.txt
Raw data must be obtained from the NOEMD database due to licensing restrictions. Please refer to data/README.md for instructions.
Once the raw CSV is placed in data/, run the processing script to generate graph data and split indices:
python scripts/01_process_data.py
The model is trained in two stages.
Stage 1:
python scripts/02_train_stage1.py
Stage 2:
python scripts/03_train_stage2.py
To generate prediction results using the trained models:
python scripts/04_inference.py
The results will be saved in data/inference_results/.
MIT License