The following code was used to design an NN model to classif…

The following code was used to design an NN model to classify the CIFAR10 dataset. We used a 70% dropout in the fully connected layers. The model summary is shown in the image below. Fill in the blanks.  #defining the model model_a=Sequential() model_a.add(Conv2D([blank1], (3, 3), padding=’same’, input_shape=x_train.shape[1:])) model_a.add(Activation(‘relu’)) model_a.add(Conv2D([blank2], (3, 3), padding=’same’)) model_a.add(Activation(‘relu’)) model_a.add(MaxPooling2D(pool_size=([blank3], [blank4])) model_a.add(Conv2D([blank5], (3, 3), padding=’same’)) model_a.add(Activation(‘relu’)) model_a.add(Conv2D([blank6], (3, 3), padding=’same’)) model_a.add(Activation(‘relu’)) model_a.add(MaxPooling2D(pool_size=([blank7], [blank8]))) model_a.add(Flatten()) model_a.add(Dense(units=[blank9], activation=’relu’)) model_a.add(Dropout([blank10])) model_a.add(Dense(units=[blank11], activation='[blank12]’)) model_a.summary()