Add more examples via notebooks (colab is a good idea)

We could even have better examples if we add:

- [ ] [SHAP dependence plots](https://christophm.github.io/interpretable-ml-book/shap.html)
- [ ] Comparing autopytorch with using a simple neural network (what do you get out of autopytorch?).
- [ ] Providing a log file and having local access to the log file. And basically explaining the directory structure of the temporal folder.