ipa dataset

[fati2023-dorhmi]

Hello, thank you for sharing your work – it really helps. I've been wondering if I can use PL-BERT to fine-tune CoquiTTS. The dataset I’m using contains sentences in IPA and audio. I’m wondering if PL-BERT can directly understand and tokenize IPA input. Also, since I’m working with different languages, is it necessary to fine-tune PL-BERT before using it.
I really need guidance on this, as I feel a bit lost.