Demo for using data iterator with Quantile DMatrix
==================================================
.. versionadded:: 1.2.0
The demo that defines a customized iterator for passing batches of data into
:py:class:`xgboost.QuantileDMatrix` and use this ``QuantileDMatrix`` for
training. The feature is used primarily designed to reduce the required GPU
memory for training on distributed environment.
Aftering going through the demo, one might ask why don't we use more native
Python iterator? That's because XGBoost requires a `reset` function, while
using `itertools.tee` might incur significant memory usage according to:
https://docs.python.org/3/library/itertools.html#itertools.tee. 