Custom LLM models

[Moosdijk]

Since the application is still under development, don't report issues if you're using a different model than the standard gemma3:4b model

In order to skip downloading the model during install:
cmd -> ollama list : to get a list of local models, choose one you want to use instead and copy its name.
setup.ps1: change "gemma3:4b" to the name you've copied on lines 197 and 199.

This will not install the gemma3:4b model.

Next, I searched for gemma3 references in the .py files.
Do the same for the following files:
\apps\backend\app\services\job_service.py (line 23)
\apps\backend\app\services\resume_service.py (line 27)
\apps\backend\app\agent\manager.py (line 11) (line 43 references the embedding model)
\apps\backend\app\agent\providers\ollama.py (line 14) (line 58 references the embedding model)