Week 3, 4.7 Exercise 8 Integration of NN_Model

Course Q&A Deep Learning Specialization Neural Networks and Deep Learning
1

Greetings.

Why don’t we use learning rate as parameter when we integrate all functions in nn_model ?

2

Hi @Apostolos ,

A learning rate is used when the the functions are integrated together in the nn_model* function.

If you look at the update_parameters function, you will see that one of the arguments to the function is learning_rate defaulted to a particular value. Look at the function signature you will see this.

In the integrated nn_model function, we call the function update_parameters to update parameters the base values and gradients are passed in, and but not an explicit learning rate.

Python will use the default learning rate when the update_parameters function is called.

Hope this helps.

1 Like
3

Yeah. Totally forgot that.

Thanks a lot @jeosol .

4

[quote=“jeosol, post:2, topic:20150”]
A learning rate is used when the the functions are integrated together in the nn_model * function.

If you look at the update_parameters function, you will see that one of the arguments to the function is learning_rate defaulted to a particular value. Look at the function signature you will see this.

In the integrated nn_model function, we call the function update_parameters to update parameters the base values and gradients are passed in, and but not an explicit learning rate.

Python will use the default learning rate when the update_parameters function is called.

Hope this helps.