# Create a transformer object from the trained model.
xgb_transformer = xgb.transformer(instance_count = 1,
                                  instance_type = 'ml.m4.xlarge')

# Start the transform job
xgb_transformer.transform(test_location,
                          content_type='text/csv', #pecify the content type
                          split_type='Line') #split type of the test data