I have a CNN model that has to predict two outputs:

• regression (bounding box coords)
• classification (class of the images with are one-hot vector of len3)

how can i fit the grid search? if i pass as y a dictionary i will get mismatch error because it sees the len of the dictionary as shape

what wrapper to use?

I tried to use KerasClassifier and KerasRegresso but neither of the two works since the outputs are one regression and one classification

I tried to combine the output but it does not work since the model has 2

i tried to use the dictionary and the list for the fit method for grid search