Enable autologging within configuration file

[Galileo-Galilei]

Mlflow offers a some autologging functionnalities for the most common machine learning frameworks (keras, tensorflow, xgboost...). It would be nice to be able to configure these autologging parametersin the mlflow.yml configuration file.

[Galileo-Galilei]

• Advantages of mlflow autologging: it logs metrics automatically during training, and not at the end of the node, which is convenient. For most models, this can be overcome with callbnacks so this is not a big issue.
• Drawbacks of mlflow autologging: It logs parameters / models / metrics without names which makes tracking very hard to read. This also conflicts with kedro-mlflow native parameters tracking (which is much more readable) so it is hard to integrate.

Decision: I won't support autologging for now, but I'll document how one can customize it with #442.