DOC: onboarding documentation is pretty sparse

[walterheck]

Issue with current documentation:

I'm super happy this package exists, thanks for all the work. I've faced several issues adopting this package, which I think mostly boil down to documentation. A short list:

1. The current documentation is in Jupyter notebooks, but there is no indication on how to run these
2. This implementation tells you to not use document_ids and file_ids/folder_ids because it's deprecated. It doesn't explain why this is, and it doesn't explain the line of thinking behind the templates setup properly. Coming from the langchain-google package, I just have a list of google doc ids, some of them being files, others being folders.
3. The ways in which you can authenticate aren't complete, for instance I didn't see anywhere that I can pass a working credentials object to the GoogleDriveLoader() call, which was super important to me
4. This package uses poetry and make, but that's not really documented anywhere it seems, so I had to figure this out by trying some stuff
5. this package seems to use detector2 for some things, but there's no pypi package for this. There's no indication of what's needed to run/install detector2, nor what it is used for and/or what I am missing out on by not having it?

Idea or request for content:

I think it would make sense to write a bit about the things this package does better than it's langchain-google counterpart, how it does them and how you can start using them.