Refactor configuration [Part II]. (#4577)

* Refactor configuration [Part II].

* General changes:
** Remove `Init` methods to avoid ambiguity.
** Remove `Configure(std::map<>)` to avoid redundant copying and prepare for
   parameter validation. (`std::vector` is returned from `InitAllowUnknown`).
** Add name to tree updaters for easier debugging.

* Learner changes:
** Make `LearnerImpl` the only source of configuration.

    All configurations are stored and carried out by `LearnerImpl::Configure()`.

** Remove booster in C API.

    Originally kept for "compatibility reason", but did not state why.  So here
    we just remove it.

** Add a `metric_names_` field in `LearnerImpl`.
** Remove `LazyInit`.  Configuration will always be lazy.
** Run `Configure` before every iteration.

* Predictor changes:
** Allocate both cpu and gpu predictor.
** Remove cpu_predictor from gpu_predictor.

    `GBTree` is now used to dispatch the predictor.

** Remove some GPU Predictor tests.

* IO

No IO changes.  The binary model format stability is tested by comparing
hashing value of save models between two commits

tests/python/test_basic_models.py

@@ -154,7 +154,7 @@ class TestModels(unittest.TestCase):
 
     def test_multi_eval_metric(self):
         watchlist = [(dtest, 'eval'), (dtrain, 'train')]
-        param = {'max_depth': 2, 'eta': 0.2, 'verbosity': 0,
+        param = {'max_depth': 2, 'eta': 0.2, 'verbosity': 1,
                  'objective': 'binary:logistic'}
         param['eval_metric'] = ["auc", "logloss", 'error']
         evals_result = {}