Consider the following code creating a simple convolutional neural network: model=Sequential() model.add(layers.Conv2D(filters=6,kernel_size=5, strides=2,padding=valid ,input_shape=(21,21,3))) model.add(layers.MaxPooling2D(pool_size=3, strides=2, padding=valid)) model.add(layers.Conv2D(filters=10,kernel_size=3, strides=1,padding=same)) model.add(layers.Flatten()) model.add(layers.Dense(10))

For each of the 5 layers calculate the following:

The dimensions of the output

The number of trainable parameters (weights and biases)