A fully cоnnected lаyer with nin inputs аnd nоut neurоns hаs nin × nout + nout parameters (weights + biases). The input layer performs no computation and is not counted in the depth.(i) A network has 2 inputs, one hidden layer with 3 neurons, and 1 output neuron. Its depth (number of computing layers) is [depth] and its total parameter count is [small].(ii) An MLP for 28×28 grayscale images has 784 inputs, one hidden layer of 64 neurons, and 10 outputs. Its total parameter count is [big]. (This explosion of parameters is one reason the lecture recommends CNNs for images.)Format: enter numeric answers as plain decimals (e.g., 0.25, not 1/4). Do not include units or commas.