Datasets:

ayymen

/

Pontoon-Translations

like 10

I refreshed the dataset. It's still computing the last configs

Thank you! I realized later that I had tabs in the fields of my tsv files. They're the cause of the UnexpectedError in some subsets like de-en. I don't know about the other errors though.
I have fixed this error in my latest commit. Could you please refresh the dataset again?

I realized later that I had tabs in the fields of my tsv files. They're the cause of the UnexpectedError in some subsets like de-en. I don't know about the other errors though. I have fixed this error in my latest commit. Could you please refresh the dataset again?

OK! The dataset viewer is refreshed on every new commit (it takes some time because the dataset has so many configs), so: no manual action is needed.

For de-en, we don't expose the error (maybe we should @albertvillanova @lhoestq ), but it seems to be the same issue you mentioned: you still have a tab in your data

pandas.errors.ParserError: Error tokenizing data. C error: Expected 2 fields in line 6763, saw 3

For de-en, we don't expose the error (maybe we should @albertvillanova @lhoestq ), but it seems to be the same issue you mentioned: you still have a tab in your data

I've just checked, the problem is not extra tabs but rather quotes or some other character. I should probably use another format like JSONL.

@severo Sorry for bothering you again but even after fixing the previous issue some splits like ba-en still throw a JobManagerCrashedError error.
I suppose it's not a parsing error.

Indeed. I refreshed the 17 failing configs, all are working now. We have an issue here: https://github.com/huggingface/datasets-server/issues/2218 to automatically retry this kind of error (instead of doing it manually as I did).

Thank you!