Delay breaking changes to 1.6. (#7420)

The patch is too big to be backported.

doc/tutorials/custom_metric_obj.rst

@@ -21,6 +21,7 @@ concepts should be readily applicable to other language bindings.
 .. note::
 
    * The ranking task does not support customized functions.
+   * Breaking change was made in XGBoost 1.6.
 
 In the following two sections, we will provide a step by step walk through of implementing
 ``Squared Log Error(SLE)`` objective function:
@@ -270,7 +271,7 @@ Scikit-Learn Interface
 
 
 The scikit-learn interface of XGBoost has some utilities to improve the integration with
-standard scikit-learn functions.  For instance, after XGBoost 1.5.1 users can use the cost
+standard scikit-learn functions.  For instance, after XGBoost 1.6.0 users can use the cost
 function (not scoring functions) from scikit-learn out of the box:
 
 .. code-block:: python

python-package/xgboost/sklearn.py

@@ -199,7 +199,7 @@ def inner(y_score: np.ndarray, dmatrix: DMatrix) -> Tuple[str, float]:
 
     eval_metric : Optional[Union[str, List[str], Callable]]
 
-        .. versionadded:: 1.5.1
+        .. versionadded:: 1.6.0
 
         Metric used for monitoring the training result and early stopping.  It can be a
         string or list of strings as names of predefined metric in XGBoost (See
@@ -239,7 +239,7 @@ def inner(y_score: np.ndarray, dmatrix: DMatrix) -> Tuple[str, float]:
 
     early_stopping_rounds : Optional[int]
 
-        .. versionadded:: 1.5.1
+        .. versionadded:: 1.6.0
 
         Activates early stopping. Validation metric needs to improve at least once in
         every **early_stopping_rounds** round(s) to continue training.  Requires at least
@@ -855,11 +855,11 @@ def fit(
             Validation metrics will help us track the performance of the model.
 
         eval_metric : str, list of str, or callable, optional
-            .. deprecated:: 1.5.1
+            .. deprecated:: 1.6.0
                 Use `eval_metric` in :py:meth:`__init__` or :py:meth:`set_params` instead.
 
         early_stopping_rounds : int
-            .. deprecated:: 1.5.1
+            .. deprecated:: 1.6.0
                 Use `early_stopping_rounds` in :py:meth:`__init__` or
                 :py:meth:`set_params` instead.
         verbose :
@@ -881,7 +881,7 @@ def fit(
             `exact` tree methods.
 
         callbacks :
-            .. deprecated: 1.5.1
+            .. deprecated: 1.6.0
                 Use `callbacks` in :py:meth:`__init__` or :py:methd:`set_params` instead.
         """
         evals_result: TrainingCallback.EvalsLog = {}
@@ -1693,11 +1693,11 @@ def fit(
             pair in **eval_set**.
 
         eval_metric : str, list of str, optional
-            .. deprecated:: 1.5.1
+            .. deprecated:: 1.6.0
                 use `eval_metric` in :py:meth:`__init__` or :py:meth:`set_params` instead.
 
         early_stopping_rounds : int
-            .. deprecated:: 1.5.1
+            .. deprecated:: 1.6.0
                 use `early_stopping_rounds` in :py:meth:`__init__` or
                 :py:meth:`set_params` instead.
 
@@ -1727,7 +1727,7 @@ def fit(
             `exact` tree methods.
 
         callbacks :
-            .. deprecated: 1.5.1
+            .. deprecated: 1.6.0
                 Use `callbacks` in :py:meth:`__init__` or :py:methd:`set_params` instead.
         """
         # check if group information is provided

python-package/xgboost/training.py

@@ -80,7 +80,7 @@ def train(
         <https://xgboost.readthedocs.io/en/latest/tutorials/custom_metric_obj.html>`_ for
         details.
     feval :
-        .. deprecated:: 1.5.1
+        .. deprecated:: 1.6.0
             Use `custom_metric` instead.
     maximize : bool
         Whether to maximize feval.
@@ -132,7 +132,7 @@ def train(
 
     custom_metric:
 
-        .. versionadded 1.5.1
+        .. versionadded 1.6.0
 
         Custom metric function.  See `Custom Metric
         <https://xgboost.readthedocs.io/en/latest/tutorials/custom_metric_obj.html>`_ for
@@ -392,7 +392,7 @@ def cv(params, dtrain, num_boost_round=10, nfold=3, stratified=False, folds=None
         details.
 
     feval : function
-        .. deprecated:: 1.5.1
+        .. deprecated:: 1.6.0
             Use `custom_metric` instead.
     maximize : bool
         Whether to maximize feval.
@@ -432,7 +432,7 @@ def cv(params, dtrain, num_boost_round=10, nfold=3, stratified=False, folds=None
         Shuffle data before creating folds.
     custom_metric :
 
-        .. versionadded 1.5.1
+        .. versionadded 1.6.0
 
         Custom metric function.  See `Custom Metric
         <https://xgboost.readthedocs.io/en/latest/tutorials/custom_metric_obj.html>`_ for