- If initial model fails to load, invoke.py will inform the user that
something is wrong with models.yaml or the models themselves and
drop user into configure_invokeai.py to repair the problem.
- The model caching system will longer try to reload the current model
if there is none.
