if it's not convergent, you can do one of the following:
1. try to understand why it does not converge. is the model well specified? can you add or remove some RHVs?
2. if it does not work, then change the maximization method and try again
3. If that still does not work, you have to work on the maximization procedure. you may set up the max number of iterations or change the tolerance threshold. This is only for you to see what the results look like. it is not recommended for reporting the final results.
hope it helps