The 4.7 update changed the model selected in previous version to “Auto”.
This is not good for long sessions over 30 minutes it seems because it switches model mid project. This causes the new model to have to relearn the project and what it was supposed to be doing.
I just realized it was set to Auto after updating a few days ago, and I was wondering why the responses and work where worse than before. It’s because with the auto setting, it randomly switches the model which causes the model to not know whats going on, it has no context.
Now I it set to my previous model and everything is smooth like butter.