A neural network based model in order to learn the solution to the 1D cascade equation governing the evolution of Extensive Air
Showers (EAS) is presented. The neural network is then used to generate the spectra of secondary particles at every height slice. The ability
of the network to learn the function to generate the next iteration in shower development is showcased. Pitfalls in using the
network in generating the entire shower is discussed. A sequential network model, which can iteratively generate the entire
shower from an initial table is presented. We show that the network learns to generate a single step with approximately 5% error and how the network
is accurate in the later parts of the shower and error prone in the early parts of the shower.