1. class SoftMaxRegression(nn. Module): \[ \begin{array}{l} \text { def init_(self): } \\ \text { super(). init_ () } \\ \text { \# define layers } \\ \text { self. flatten }=n n \text {. Flatten() } \\ \text { self. linear }=n n . \text { Linear }(784,10) \\ \end{array} \] def forward (self,x) : \# Now it only takes a call to the layer to make predictions y=self,flatten(x)y=self.1inear(y) return y Suppose you have a set of color images of dimension (3322020) that belong to 11 different classes. You want to re-use the above simple model in order to work with such images. Give the two updated arguments for the definition of nn. Linear. (Please write only one number for each blank)