Categorize SFT and Pretraining data

[rasbt]

Ideally we only want the SFT finetuning datasets to show up in litgpt finetune --help and only the pretraining datasets to show up in litgpt pretrain --help.

I believe we were thinking about that a while back but there was no good way to do that with jsonargparse. I am opening this issue so we don't forget to revisit this someday.

[carmocca]

Isn't this a duplicate of https://github.com/Lightning-AI/litgpt/issues/1084 of yours?

[carmocca]

Note that this needs to be done by having two different base classes and having the files use only one of them in their type signatures

[rasbt]

We can close this as duplicate.