Fitness Function Error Tolerance: While training you will be wondering how do I know when to stop?

The answer to this is the Fitness Function.

Fitness function value is a measure of accuracy that is computed by comparing actual vs. predicted output for all training sets.

Example: Assume you had 100 sample sets with one output each.  If 99 of the outputs predicted fell within the actual values (+/- tolerance) then your fitness function rating would be 99%.