In the context of a machine learning project within Databricks, a data scientist needs to share a trained model with a team member for further evaluation. Which method is recommended for packaging and sharing the model using MLflow?