Fit a scikit-learn model in parallel?

Question

Is it possible to fit a scikit-learn model in parallel? Something along the lines of model.fit(X, y, n_jobs=20)