Suggestion: keep hyperparameters for all iterations

**Is your feature request related to a problem? Please describe.**
I have dataset split into training and test set. I am using train set with Model builder.
I have been now testing all the models in temp folder against the train set (not just the generated one but the 1....999.zip's in temp)
The best model selected for model builder is not always the best in test data set.

**Describe the solution you'd like**
I'd like to be able to retrieve the hyperparameters used for training of each zip file in temp file.
It could be, for examples, in .mbconfig file. Saving them would also help later restart the training with different kinds of tuning engines, based on the results.

**Describe alternatives you've considered**
Best solution could be that the hyperparameters would be saved inside the .zip file, because retraining is not possible without it for many algorithms (such as trees). That would make archiving models easier, as there would not need to be a separate parameters file.

Inizia dalla gestione dell’output dell’addestramento di Model Builder e ispeziona i file generati nella cartella temporanea, inclusi i file .mbconfig e i file .zip del modello. Determina dove sono disponibili gli iperparametri di ogni iterazione e definisci il completamento come la possibilità di recuperarli per ogni zip, in .mbconfig oppure all’interno dell’archivio, preservandone l’uso per il riavvio e l’archiviazione.

Scritto dal modello di indicizzazione a partire dal testo della issue.