Disscussion about adding native audio streaming support to use in TTS(voice bot)
#1
by PRIYANSHUDHAKED - opened
Hi Dear Aratako can u please tell me which loss function is used to training. I am using MultiResMelSpecLoss, Phase loss, magnitude loss to for Lora finetuning with some architecture change to add audio streaming. Did you use any other loss ?
Can I know are you thinking about to add output Audio streaming in Mio-code then I will love to help you for this Innovative work. can I know how much hours it took to train from scratch (hardware,dataset).