[GENERAL] Test Math & Code Generation Behavior of GPT-4.5 and Ollama #20
Description
Issue Type
Performance Issue
Description
It would be informative to know how the current project produces Math & Code outputs. Some tests were conducted just to see basic functionality working (like LaTex formatting, Code Snippets shown in a special field with "Copy" function), but other than that, there has not been done any other deep tests, especially consistency tests.
So a thorough report would be nice to get an in-depth insight of how Math & Code Generation works and how to make it more consistent at every generation.
Steps you can follow (but don't have to, this is just a little push to the relevant direction. Feel free to experiment with the generator!):
- Create a microcourse that generates code examples or math expressions.
- Fetch the resulting microcourse or section through the API.
- Inspect the code_examples and math_expressions fields.
Additional Context
No response
Checklist
- I have searched existing issues and discussions
- This is not a security vulnerability (see SECURITY.md)