Hi,
I just noticed that there is a small mistake in the explanation of exercise 2 (convolutional_block
) of the Residual_Networks programming exercise.
Just above the graded function, it says that the initialiser will be set to random_uniform
- but that should of course be glorot_uniform
(as correctly stated a few lines higher). It looks like this text was pasted from the identity_block
exercise
Nothing major, and it actually does not affect the graded function as the correct initialiser is used there, but it confused me a bit so I thought Iād make a post about it.