Training new concepts

[barrubba]

I have been using SD for months to create images but have never tried my hand at Dreambooth/lora.
I would like to create images that in addition to the basic SD model, contain new textual and conceptual elements, related to the naturalness, places, monuments, architecture, folkways, crafts, food of a specific geographic area.
What is the workflow to then be able to mix all these new concepts?
Do I first prepare the basic images, a specific nomenclature subject by subject, and then?
Generate a model with DB for each concept?
And eventually will it be possible to use them all in one model?
Thank you

[missionfloyd]

https://github.com/bmaltais/kohya_ss#tutorials

There are more tutorials on lora training on YouTube.