The following dataset breaks the API specifically because Abstract Text column is too long. If the textual column in the dataset need to have a certain length for the API to work that I am not aware of or has a workaround, please let me know.
RuntimeError: Fail to call fit with error: {'detail': 'Invalid MIME type application/octet-stream. Only CSV (text/plain, text/csv) allowed.'}
osha_data.csv