If you run the notebooks locally or in any other environment besides the course website, there is no guarantee that they will run without error or generate the same results. That is because the course materials are “dated”: released at one point in the past (perhaps different for different assignments) and using whatever the current versions of the hundreds of packages they depend on. For example, this course DLS Course 2 was last updated in April of 2021. That’s 3 years ago and a lot can change in that time. The world of python packages, especially those related to ML/DL, is very dynamic: things don’t hold still for very long.
There are no official instructions for how to cope with this, because a) it’s complicated and b) it’s not DLAI’s problem. They provided you with a working environment. If you want to port it someplace else, then you need to have the IT skills to solve whatever problems you run into. Here’s a thread with a lot of information from a fellow student which will get you started down that road. It’s fine to ask questions about things like this, but you need to be prepared for the fact that you’re not guaranteed to get an answer.