Python: baesian or machine learning driver code.

I have a forward model doing some physical processes. I want to perturbate slightly two parameters in this model and by iteration reach convergence with lab data. Then the driver code uses this forward model and provides plots and error estimates