Read through Backprop as Functor, report how it fits into hasktorch about hasktorch

[stites]

From Backprop as Functor (Fong, Spivak, Tuyeras):

In this paper we will show that learning becomes compositional — i.e. we can define a learning algorithm A → C from learning algorithms A → B and B → C — as long as each learning algorithm consists of these four components:

• a parameter space P,
• a function I : P × A → B,
• an update function U : P × A × B → P, and
• a request function r : P × A × B → A

[...]

Namely, we will show that Learn has the structure of a symmetric monoidal category.

This would be a good way to abstract out the NN package without committing to a backend for auto differentiation as Torch (also backprop) provide us with functions that resemble this.

[stites]

closing for the time being