Is it normal to not be able to deeply comprehend at this stage?

I'm going through Andrew NG's machine learning specialization and currently at second course "advanced learning algorithms" which is mostly about NNs. And I can't 100% comprehend everything. Like when he trains a model while explaining which activation algorithm to use, I question ok but how this model gets trained by this. There are multiple layers-units and how does each get updated. Or even basic things like why did we use N layers and M units for this particular problem?

I …

advanced algorithm algorithms andrew andrew ng course deeplearning everything machine machine learning nns normal stage through trains