TinyML: Pruning

• Over parameterization helps with optimization and after the optimization we can prune them away. (redundency is helpful for us to get away from local minimum)