The model deployment component saves and persists the model to the Vantage table so that you can use it for prediction or scoring or other class methods that require model in other sessions.
This model can be a model trained either by teradataml OpenSourceML framework or outside of teradataml OpenSourceML.