https://uberduck.ai/ logo
Powered by
# machine-learning
  • u
    You're really gonna make me go back on just to take a look
  • u
    I definitely did stage 3
  • u
    @User What do I do after it finishes training?
  • u
    Im on the synthesis notebook
  • g
    is there any website tts that can make it sound exactly like someone based n their voice?
  • u
    like 100%? absolutely not
  • u
    things get pretty damn close often though
  • g
    a close one would be good
  • g
    it doesnt have to be exact
  • m
    15 ai is probably the closest to realistic, and is a website...
  • j
    Natural emotive high-fidelity text-to-speech synthesis with minimal viable data
  • s
    Unnatural unemotive low-fidelity text-to-speech synthesis with maximum viable data
  • u
    no more text-to-speech synthesis with no more data
  • u
    go home
  • z
    there's no text-to-speech synthesis
  • z
    go home
  • u
    nothing cloned anyone
  • u
    he does not exist
  • u
    he does not exist
  • m
    nothing exists
  • m
    go home
  • u
    he does not exist
  • z
    the earth doesn't exist
  • g
    does it make a tts off a audio file?
  • m
    its essentially like uberduck, but better tech
  • m
    its not open source tho, so datasets need to be given to the creator by email, and he will add them in 50 years or whatever short updates are to him
  • t
    So I think I'm ready to fully commit to understanding how to write a bilingual voice notebook based on the GitHub page one of you linked me
  • t
    I mentioned before I started writing it but then got stuck
  • t
    I don't know if I need to download *everything*; all I want is the Japanese data
  • t
    Also don't know if it's a process of downloading to upload to Google Drive or if I could write the code in my notebook
1...710711712...1068Latest