My model is such that before training it I need to one-hot encode and scale the features. I train this model on a device, then I need to save the trained model and load it on other devices to predict a target. Is it possible to save the whole trained pipeline of scikit-learn operations and load it on the other devices so that I don't need to separately save the criterions for scaling and encoding the input data and apply those functions before using the actual algorithm on the other devices?

Related posts

Recent Viewed