No data on the actual model, they just gave the mathematical formulation of reinforcement learning (which can be simply copied from an undergraduate textbook).

No data on the actual model, they just gave the mathematical formulation of reinforcement learning (which can be simply copied from an undergraduate textbook). The Reinforce policy gradient algorithm still has to operate on a differentiable model.