Save tradaboost model

[971766904]

I want to know how to save a model trained by tradaboost whose estimator is LGBMClassifier and how to load a saved model when I need to deploy on real time situation.

[antoinedemathelin]

Hi @971766904,

Thank you for your interest in the Adapt library. To save and load the TrAdaBoost model, you can use joblib as for scikit-learn estimators.

Here is an example:

import joblib
from adapt.utils import make_classification_da
from adapt.instance_based import TrAdaBoost
from lightgbm import LGBMClassifier

Xs, ys, Xt, yt = make_classification_da()

model = TrAdaBoost(estimator=LGBMClassifier(num_leaves=20))
model.fit(Xs, ys, Xt, yt)

# Save the fitted model
joblib.dump(model, "model.pkl")

# Load the model and make predictions
loaded_model = joblib.load("model.pkl")
loaded_model.predict(Xt)

[971766904]

Hi @971766904,

Thank you for your interest in the Adapt library. To save and load the TrAdaBoost model, you can use joblib as for scikit-learn estimators.

Here is an example:

import joblib
from adapt.utils import make_classification_da
from adapt.instance_based import TrAdaBoost
from lightgbm import LGBMClassifier

Xs, ys, Xt, yt = make_classification_da()

model = TrAdaBoost(estimator=LGBMClassifier(num_leaves=20))
model.fit(Xs, ys, Xt, yt)

# Save the fitted model
joblib.dump(model, "model.pkl")

# Load the model and make predictions
loaded_model = joblib.load("model.pkl")
loaded_model.predict(Xt)

thank your for your reply