xgboost GridSearchCV take too long or does not goes to the next step

Question

just strange

%%time
xgb = xgb.XGBRegressor(n_estimators=500, learning_rate=0.07, gamma=0, subsample=0.75, colsample_bytree=1,
                            max_depth=7, tree_method='gpu_exact')

this code takes around Wall time: 866 ms.

but when I do the gridsearchCV it does not goes to the next step
even though I gave only one parameter

    %%time
    xgb = XGBClassifier(tree_method='gpu_exact',verbose_eval=True, silence=False)
    kfold = StratifiedKFold(n_splits=10, random_state=0)
    xgb_param_grid = {
                    'learning_rate': [0.08,0.09],
                    'random_state': [0],
                    'max_depth': [8,9],
                    'n_estimators': [400,500]
    }
    xgbGrid = gsRFC = GridSearchCV(xgb,param_grid = xgb_param_grid, cv=5, scoring="neg_mean_squared_error", n_jobs= 10, verbose = 1)
    xgbGrid.fit(X,y)
    xgb_best = xgbGrid.best_estimator_

for my understanding, this should not take that long.<br/>
it d

does not go to the next step I do not sure this even working or not
it stop with

Fitting 5 folds for each of 8 candidates, totalling 40 fits [Parallel(n_jobs=10)]: Using backend LokyBackend with 10 concurrent workers.

data set size is (15035, 22) am I doing something wrong?

Answer 1

Based on the combinations of learning parameters, learning rate(2), max_depth(2) and n_estimators(2), it seems the algorithm is exactly doing what it's supposed to do. with cross validation set to 5 it's performing 40 fits (2*2*2*5). Can you add some more details to clarify like what configuration you are using for doing this task?