Does AutoNLP configure hyperparameters too?

For example, when I’ve chosen BigBird as the model (family) and I have 15 candidate models, are those all different hyperparameter configurations of BigBird? Just want to sanity check that it’s not going to train 15 of the exact same model. Screenshot for reference:

Hi Brady, yes that’s correct, AutoNLP will try different configs and evaluate each resulting model.


Thank you!