Description of feature
Now that we can cache datasets for notebooks and we already run some of them in CI we should be able to run the rest. If can split them or find another way. At worst we can list individiual notebooks as exceptions but it has been clear to me that we should include this in the CI.
For example even though #1094 passes the unit tests I can't be sure if it works with all of our notebooks because the real tests with the real data are being done there only. Would finish #707
In the future we can even compare outputs of the notebooks and can check if a PR changes the outputs.