What is the difference of static Computational Graphs in tensorflow and dynamic Computational Graphs in Pytorch?

Question

When I was learning tensorflow, one basic concept of tensorflow was computational graphs, and the graphs was said to be static. And I found in Pytorch, the graphs was said to be

Answer 1

In tensorflow you first have to define the graph, then you execute it. Once defined you graph is immutable: you can't add/remove nodes at runtime.

In pytorch, instead, you can change the structure of the graph at runtime: you can thus add/remove nodes at runtime, dynamically changing its structure.