Model Prediction Partial Least Square Model - r

Question

I am following the procedure explained Hair et al (2021) to run a partial least square model (seminR).
So far, it worked well. However, when using the predict function, I get the following error:
Parallel encountered this ERROR:
Must subset columns with a valid subscript vector.
x Subscript endogenous_items must be a simple vector, not a matrix.
r in summary.connection(connection) : invalid connection
I have recently started working with r. My dataset is an excel table with 579 obs. of 47 variables. How can I solve these problems? Thank you very much in advance.
That's my code:
composite("EA", multi_items("EA_", 1:3))`
composite("DB", multi_items("DB_", 1:5)),
composite("LTB", multi_items("LTB_", 1:5)),
composite("SN", multi_items("SN_", 1:3)),
composite("PBC", multi_items("PBC_", 1:3)),
composite("SE", multi_items("SE_", 1:8)),
composite("INT", multi_items("INT_", 1:2)),
composite("B", multi_items("B_", 1:2)))
Create structural model
`final_sm_ext <- relationships(
paths(from = c("DB", "LTB", "SN", "PBC", "SE", "INT"), to = c("B")),
paths(from = c("EA"), to = c("INT")))`
bike_final_model_ext <- estimate_pls('data = PLSdata,
measurement_model = final_mm_ext,
structural_model = final_sm_ext,
inner_weights = path_weighting,
missing = mean_replacement,
missing_value = "-99")
summary_bike_final_model_ext <- summary(bike_final_model_ext)
predict_bike_final_model_ext <- `predict_pls( model = bike_final_model_ext,
technique = predict_DA, noFolds = 10, reps = 10)```