teradataml OpenSourceML offers ways to deploy and load the models that are trained inside or outside of Vantage.
Unlike traditional open-source packages, teradataml OpenSourceML provides you the following options:
- Deploy and save the trained model to a table. See Model Deployment.
This table (opensourceml_models) is internally created by teradataml OpenSourceML in connected database, and should not be deleted.
- Load the saved model from a Vantage table in Python as an object. See Model Loader.
See the following examples of deploying and loading models under different scenarios: