I want to fine tune GPT-4o-mini on a lot of Pdf files and physiological data to train a health coach

I want to fine tune GPT-4o-mini in Google Colab( using GPU) on a lot of .pdf files (around 5000 pdf documents) and physiological datastes to train a health coach for my app but I am so lost. I would really appreciate any help or resources. I do not want to do this manually.