Extracting best model from autokeras

I am using autokeras for my binary classification and i got these results:

clf = ak.StructuredDataClassifier(max_trials=100,overwrite=True)
clf.fit(X_train,y_train, epochs= 1000,validation_data=(X_test,y_test))

and i want to have the best model. I used:

clf.export_model()

and then:

model = tf.saved_model.load(export_dir, tags=None, options=None)

and now i want to check if it really achieve 68 accuracy and what is the structure of this net.