Due to storage limitations, we are unable to backup the entire mlflow artifact store, as a minimum, we will backup the mlflow_db which should allow us to retrain and reproduce any artifacts from a particular run.
In addition to this, we should backup the model registry to ensure deployed models can be restored quickly.
Due to storage limitations, we are unable to backup the entire mlflow artifact store, as a minimum, we will backup the mlflow_db which should allow us to retrain and reproduce any artifacts from a particular run.
In addition to this, we should backup the model registry to ensure deployed models can be restored quickly.