Pre-trained Model and training_mode changes

[nvidianz]

Description

1. Added support for pre-trained model in XGBoost. The model must reside in custom/model.json.
2. Changed training_mode to split_mode and secure_training to align with XGBoost parameters.
3. Merged changes made to main and changed all the examples to use split_mode.

Types of changes

• Non-breaking change (fix or new feature that would not break existing functionality).
• Breaking change (fix or new feature that would cause existing functionality to change).
• New tests added to cover the changes.
• Quick tests passed locally by running ./runtest.sh.
• In-line docstrings updated.
• Documentation updated.

[YuanTingHsieh]

Mostly LGTM, some codes are accidentally removed

[YuanTingHsieh]

thanks, LGTM

[chesterxgchen]

have you test this ? to have a pre-trained model in the json format ? custom/model.json

[nvidianz]

have you test this ? to have a pre-trained model in the json format ? custom/model.json

Yes. With the model there, the AUC improved a little bit.