Fine tuning a LLM

I made a pdf chatbot using below blog and with guidance of @Caroline

Now I want to learn how can I do fine tuning line explained in below post

Only need guidance on what amendment would be required in the backend code so I can get better responses.

Kindly guide me anyone
@dataprofessor kindly give me some guideline if possible. Also I need to know what can I do with my code to interpret graphs from pdf files as well.