Model Save & Load

SUOD takes a similar approach of sklearn regarding model persistence. See model persistence for clarification.

In short, we recommend to use joblib or pickle for saving and loading SUOD models. See “examples/demo_model_save_load.py” for an example. In short, it is simple as below:

from joblib import dump, load

# save the fitted model
dump(model, 'model.joblib')
# load the model
model = load('model.joblib')