What is a bigger network ? How do we select number of units in each layer and the number of layers?

When Andrew says bigger network, does that mean more units in each layer ?
How do we make a choice of the number of units in each layer ? I understand that this is a tunable parameter. Is there a standard that deep learning practitioners follow on the number of units in each layer and the number of layers ?

Cuando se realiza una clasificación de la red en términos topológicos, se suele distinguir entre
las redes con una sola capa o nivel de neuronas y las redes con múltiples capas