3.4. Model persistence

After training a scikit-learn model, it is desirable to have a way to persist the model for future use without having to retrain. The following section gives you an example of how to persist a model with pickle. We’ll also review a few security 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-in persistence 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,
    decision_function_shape=None, degree=3, gamma='auto', 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 use joblib’s replacement of pickle (joblib.dump & joblib.load), which is more efficient on objects that carry large numpy arrays internally as is often the case for fitted scikit-learn estimators, but can only pickle to the disk 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 array contained in the clf object is serialized as a separate file on the filesystem. All files are required in the same folder when reloading the model with joblib.load.

3.4.2. Security & maintainability limitations

pickle (and joblib by extension), has some issues regarding maintainability and security. Because of this,

  • Never unpickle untrusted data
  • Models saved in one version of scikit-learn might not load in another version.

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 the same range as before.

If you want to know more about these issues and explore other possible serialization methods, please refer to this talk by Alex Gaynor.