add convergence modes

[mikibonacci]

• just a set of parameters that we want to simulate, no convergence eval. It is partially there, but apparently it is not possible to stop cleanly the workflow
• also, the max_iterations parameter is not clear