feat: add **kwargs support to XGBClassifier and XGBRegressor

[tswast]

This is in alignment with https://xgboost.readthedocs.io/en/stable/python/python_api.html\#xgboost.XGBRegressor if considering BQML to be a booster type.

Thank you for opening a Pull Request! Before submitting your PR, there are a few things you can do to make sure it goes smoothly:

• Make sure to open an issue as a bug/issue before writing your code! That way we can discuss the change, evaluate designs, and agree on the general idea
• Ensure the tests and linter pass
• Code coverage does not decrease (if any source code was changed)
• Appropriate docs were updated (if necessary)

Fixes internal customer request 🦕

[GarrettWu]

It may override existing options, if user pass in both.

We use sklearn terms instead of BQML terms for existing parameters, e.g. learning_rate instead of LEARN_RATE. User may get confused looking at two sets of options at the same time.

It becomes hard for us to both stick with sklearn-like experience and support BQML other offerings.

Or maybe we can ask them to only use one set of parameters, and raise error if conflicts?

[GarrettWu]

Do they want to use label encoding on features? This is actually wrong. https://stackoverflow.com/a/34346937

I talked with jiashangliu@, but BQML won't change the default.

[tswast]

Closing. I think the alternative of exposing the SQL methods directly in the bigframes.bigquery.ml module, similar to bigframes.bigquery.ai will be more useful.