Control the length of AUDIOs generated????

[AIhasArrived]

trafficstars

I was wondering Is there a way to make the audio generated take the exact amount of time you desire? So when I write for instance a sentence of 5 words, I could control whether the ai reader makes it in 5, 6 or 7 seconds? This is important.

Anyone with an idea if this is possible or any other method to circumvent this or control it???

[Ezhil6999]

use BetterTransformer this will reduce the inference of 20-30 persentage

[AIhasArrived]

how do I use a bettertransformer @Ezhil6999 ? Not sure what is that

[Ezhil6999]

from transformers import BetterTransformer model = BetterTransformer.transform(model, keep_original_model=False)