issues with step_sequence_onehot

[phahn]

I take part in an Kaggle- competition, just to learn tidymodels and textmodelling.
Modified some scripts from
Supervised machine learning for text analysis in R
When I try to build a recipe it works perfect on my local computer:

{r receipe}
max_words <- 15500
max_length <- 190


lit_rec  <-  recipe(~ excerpt, data = lit_train) %>%
  step_tokenize(excerpt) %>%
  step_tokenfilter(excerpt, max_tokens = max_words) %>%
step_sequence_onehot(excerpt, sequence_length = max_length)


lit_prep <-  prep(lit_rec)

Running the kaggle notebook it stops with an error:
Error: can't subset columns that don't exist
x Location 190 doesn't exist
i There are only one column

[technocrat]

@phahn

i There are only one column

str(lit_rec)

How many columns?

[phahn]

[technocrat]

That is one, and one only. A list of six lists. You have a hold on the wrong object for the problem.