The Item Saver module takes any type of objects as input and saves it to the user’s directory . The generated module will be stored in either the “Trained models” module category or in the "Dataset" list.
SmartPredict currently supports Scikit-learn models, Keras models, Spacy language models, Preprocessing models and Pandas dataframe but this list is designed to extend in the upcoming versions.
As parameters , the type of item to be saved is either: Trained Models (including ML, DL, PreProc, LANG) or Datasets. We can overwrite, make a copy, choose to not save items with duplicated names. This versatile module is utilized in every project, every time.
The example provided in the figure is an Item Saver used in the context of a classification project . The item saved here is the trained model.