facebook

/

mms-tts-eng

like 160

Follow

AI at Meta 9.71k

@mstachow , in my case, i use hyphen - for pauses

The model is trained on neither commas and periods, so they are filtered from the input text via a normalisation step in the pre-processing: https://github.com/huggingface/transformers/blob/910faa3e1f1c566b23a0318f78f5caf5bda8d3b2/src/transformers/models/vits/tokenization_vits.py#L127

Using hyphens is indeed the best option here

Will this model be re-trained on commas? even use hyphen, it still sounds very odd, just a bit better than commas. Besides, I found some pronounce is wrong, you can test "library".

@mstachow ,I have same problem there, Have you find a solution? Two years later , It is discouraging ,I'm working on an other checkpoint, there is not an other best option ...

I've long since given up on this model. Microsoft's new TTS model is impeccable and produces more natural speech anyway.