chunked/nltk_chunking/tasks_text-to-speech/chunk_34.txt

To identify any unsupported characters in the dataset, you can take the difference between these two sets.