[FEAT] Improved CI test

[kdgutier]

In the future I would like to change the current CI test towards:

• Medium sized dataset TourismL with strictly hierarchical and non-strictly hierarchical version.
• sCRPS/MSSE measurements.
• Improved execution time dataframe recordings.
• Pre computed AutoARIMA predictions for the tests to run fast.

For the precomputed AutoARIMA predictions we and new test we could inspire in this:
https://github.com/Nixtla/hierarchicalforecast/blob/main/nbs/examples/TourismLarge-Evaluation.ipynb

[elephaint]

Could include this in the CI, CI is now fast enough to just run it on the fly