Fine Tuning on website content

Hi
I want to use open access web content to generate tutorial content for a specialised audience. The content is hierarchical: Course->Modules->Topics
Topics are provided as video transcripts, text, quizzes, images

The type of questions I will ask the model are: Create a one page Key Stage 3 tutorial on cell structure and include 4 questions to check understanding and comprehension along with a suitable labelled image.

The examples for fine tuning I have seen all seem to use the Question/Answer type format, whilst the content I have (other than the quizzes) are more of a Title/Description. Any suggestions on the format the data needs to be in to fine tune?
Also I want to keep the hierarchical relationship so the model knows that topic ABC sits under Module 123 which belongs to Course A.

A lot to unpick but I would appreciate an initial steer on this