C1_W4 Exercise 5 - L_Model_Forward

DLS - Course 1, Week 4, Assignment 1, I have a question on Exercise 5 - L_Model_Forward, when is the number of NN Layers L = len(parameters) // 2? Why 2? Appreciate someone to clear my mind.
{code removed by mentor}

Hello @BHTan

We need to do a little bit of tracing back of L = len(parameters) // 2 to understand this. To begin with, as explained in the docstring, parameters is the output of initialize_parameters_deep() and if we go to that function, which is our exercise 2, its output is actually a dictionary that contains one W parameter and one b parameter for each layer. See there are 2 parameters per layer in that dictionary? It is therefore that we divide the length of that dictionary by 2 to get back the number of layers.


PS1: That is a DLS assignment, so I will move the thread back to there for you :slight_smile:
PS2: Your screenshot contains some assignment solution code, but since we canā€™t share it here, I will also remove it for you :slight_smile:

1 Like