From jmartine@redhat.com 

• Where do we store the models? cache? github artifact storage? github releases? github large file storage? somewhere else?

• I have some examples on how github artifacts work: store example and an old read example

• Force a full retrain every period or if is not possible to do an incremental training? i.e. if we add new intents we need to train from scratch
• Setup pipeline for incremental training