After training a scikit-learn model, it is desirable to have a way to persistthe model for future use without having to retrain. The following section givesyou an example of how to persist a model with pickle. We’ll also review a fewsecurity and maintainability issues when working with pickle serialization.
3.4.1. Persistence example
It is possible to save a model in the scikit by using Python’s built-inpersistence model, namely pickle:
>>> from sklearn import svm
>>> from sklearn import datasets
>>> clf = svm.SVC()
>>> iris = datasets.load_iris()
>>> X, y = iris.data, iris.target
>>> clf.fit(X, y)
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0, degree=3, gamma=0.0,
kernel='rbf', max_iter=-1, probability=False, random_state=None,
shrinking=True, tol=0.001, verbose=False)
>>> import pickle
>>> s = pickle.dumps(clf)
>>> clf2 = pickle.loads(s)
>>> clf2.predict(X[0])
array([0])
>>> y[0]
0
In the specific case of the scikit, it may be more interesting to usejoblib’s replacement of pickle (joblib.dump & joblib.load),which is more efficient on objects that carry large numpy arrays internally asis often the case for fitted scikit-learn estimators, but can only pickle to thedisk and not to a string:
>>> from sklearn.externals import joblib
>>> joblib.dump(clf, 'filename.pkl')
Later you can load back the pickled model (possibly in another Python process)with:
>>> clf = joblib.load('filename.pkl')
Note
joblib.dump returns a list of filenames. Each individual numpy arraycontained in the clf object is serialized as a separate file on thefilesystem. All files are required in the same folder when reloading themodel with joblib.load.
3.4.2. Security & maintainability limitations
pickle (and joblib by extension), has some issues regarding maintainabilityand security. Because of this,
- Never unpickle untrusted data
- Models saved in one version of scikit-learn might not load in anotherversion.
In order to rebuild a similar model with future versions of scikit-learn,additional metadata should be saved along the pickled model:
- The training data, e.g. a reference to a immutable snapshot
- The python source code used to generate the model
- The versions of scikit-learn and its dependencies
- The cross validation score obtained on the training data
This should make it possible to check that the cross-validation score is in thesame range as before.
If you want to know more about these issues and explore other possibleserialization methods, please refer to thistalk by Alex Gaynor.