Running gitlab repository on a dataset

I am developing a python code. This code runs on a dataset of yaml files and produces some outputs. Now I want to use runners to test my code on a test dataset, and then if it passed, run it on the main dataset and have the results.

How is this possible?