Run GPT tests for w/ and w/o secrets in the test data

Currently, GPT tests for Pre-receive Secret Detection are run on the test data containing a handful of embedded secrets. However, this would not be the average case scenario since the rarity of git commits containing secrets is pretty high.

The purpose of this issue is to run the tests for both varieties of test data i.e., one with and one without secrets embedded in them. This gives us clarity on the system metrics & latency on the average use case scenario.