Everything here was finetuned on a m2 max mac studio with 32 GB of ram goal is to research how well do fine tuned models do in cybersecurity and how well does fine tuning effect a model