token maps error
#5
by
ghost613
- opened
Hello,
I would like to inquire about the token_maps.pkl file used for preprocessing the dataset. I am trying to train the Korean version of PL-BERT using the multilingual branch of the official repo with this data, but I encounter the following error:
words = [self.token_maps[w]['token'] for w in words]
KeyError: 81571
Could you please provide the correct token_maps.pkl file or let me know if I am missing something in the preprocessing steps? Your assistance would be greatly appreciated.
Thank you!