Fix saving FlaubertTokenizer configs

[vmaryasin]

All specific tokenizer config properties must be passed to its base
class (XLMTokenizer) in order to be saved. This was not the case for
do_lowercase config. Thus it was not saved by save_pretrained() method
and saving and reloading the tokenizer changed its behaviour.

This commit fixes it.

What does this PR do?

Fixes # 14489

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a Github issue or the forum? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

@n1t0, @LysandreJik

Comment:

As I have said in the issue, a similar problem may happen to other models. For me, it is not obvious, that one should pass a model-specific config to the general base class constructor for the config to be saved later.

I feel like some comment should be left somewhere for future developers, but I cannot come up with a good place for it. On the other hand, maybe it is obvious from all the base class' descriptions, and I'm just not experienced enough.

[SaulLu]

Thanks a lot for the fix! 🚀

Your PR made me realize that Flaubert didn't have any tests, that is terrible! If you're interested I'll open an issue soon to add the necessary tests for this tokenizer.