Predictions failed for gbm method in caret train

Question

I want to build gradient boosting model for data vowel.train which is from R package AppliedPredictiveModelling. This is my code:

library(gbm)
library(caret)

set.seed(12345)

# Define the training control
ctrl <- trainControl(method = "repeatedcv", 
                     number = 10, 
                     repeats = 3, 
                     verboseIter = TRUE)

# Define tuning grid for GBM model
gbm_grid <- expand.grid(
  n.trees = c(50, 100, 150),
  interaction.depth = c(1, 5, 9),
  shrinkage = c(.01, .1, .5),
  n.minobsinnode = c(5, 10, 15)
)

# Train GBM model
model2 <- train(y ~ ., 
                method = 'gbm', 
                data = vowel.train, 
                tuneGrid = gbm_grid,
                trControl = ctrl,
                metric = 'Accuracy')
# Make predictions on test set
predictions <- predict(model2, newdata = vowel.test, type = 'response')

# Assess model performance
confusionMatrix(predictions, vowel.test$y)

But I can't tune model, because I have got next error message:

+ Fold01.Rep1: shrinkage=0.01, interaction.depth=1, n.minobsinnode= 5, n.trees=150 
Iter   TrainDeviance   ValidDeviance   StepSize   Improve
     1        2.3979             nan     0.0100    0.0412
     2        2.3755             nan     0.0100    0.0419
     3        2.3536             nan     0.0100    0.0360
     4        2.3331             nan     0.0100    0.0296
     5        2.3138             nan     0.0100    0.0338
     6        2.2951             nan     0.0100    0.0247
     7        2.2787             nan     0.0100    0.0350
     8        2.2599             nan     0.0100    0.0314
     9        2.2437             nan     0.0100    0.0261
    10        2.2274             nan     0.0100    0.0274
    20        2.0847             nan     0.0100    0.0188
    40        1.8872             nan     0.0100    0.0102
    60        1.7411             nan     0.0100    0.0063
    80        1.6219             nan     0.0100    0.0085
   100        1.5196             nan     0.0100    0.0041
   120        1.4333             nan     0.0100    0.0046
   140        1.3571             nan     0.0100    0.0040
   150        1.3227             nan     0.0100    0.0039

predictions failed for Fold01.Rep1: shrinkage=0.01, interaction.depth=1, n.minobsinnode= 5, n.trees=150 Error in UseMethod("predict") : 
  no applicable method for 'predict' applied to an object of class "gbm"

Warning message in eval(xpr, envir = envir):
"predictions failed for Fold01.Rep1: shrinkage=0.01, interaction.depth=1, n.minobsinnode=10, n.trees=150 Error in UseMethod("predict") : 
 no applicable method for 'predict' applied to an object of class "gbm"
"

Please, give me a hint what a problem appears in data that I can't fit it?