https://uberduck.ai/ logo
Join Discord
Powered by
# machine-learning
  • d

    Dyno

    05/22/2021, 7:39 AM
    Unfortunately, the uberduck.ai bot has not gotten verified by Discord yet even though it should have been by now. It has been taking a long time for Discord to verify the bot for some reason.
  • p

    pi

    05/22/2021, 7:44 AM
    larger parts of your dataset have inflection changes
  • p

    pi

    05/22/2021, 7:44 AM
    e.g. shouting vs normal speaking
  • p

    pi

    05/22/2021, 7:44 AM
    that's not good
  • u

    user

    05/22/2021, 7:45 AM
    hmmm ok that might be a problem
  • u

    user

    05/22/2021, 7:45 AM
    I wasn't aware they were supposed to be the same inflections
  • p

    pi

    05/22/2021, 7:48 AM
    when you filtered them out let me know
  • u

    user

    05/22/2021, 7:48 AM
    oh alright i'll be doing that later
  • u

    user

    05/22/2021, 7:49 AM
    so it should be just him talking instead of shouting?
  • p

    pi

    05/22/2021, 7:49 AM
    yes
  • u

    user

    05/22/2021, 7:51 AM
    of course I will do that as soon as I can and let you know when i'm done
  • u

    user

    05/22/2021, 1:00 PM
    can someone explain this to me
  • p

    pi

    05/22/2021, 1:12 PM
    probably some parts of your dataset are wrong
  • u

    user

    05/22/2021, 1:12 PM
    how can that be fixed
  • p

    pi

    05/22/2021, 1:12 PM
    e.g text file wrong in some parts in the wrong format
  • u

    user

    05/22/2021, 1:12 PM
    ok
  • p

    pi

    05/22/2021, 1:13 PM
    can I see the dataset?
  • u

    user

    05/22/2021, 1:13 PM
    Yes
  • u

    user

    05/22/2021, 1:13 PM
    Like the text
  • p

    pi

    05/22/2021, 1:13 PM
    https://pieksend.github.io/posts/2021-05-09-tacotron-faq/
  • u

    user

    05/22/2021, 1:14 PM
    here @User
  • p

    pi

    05/22/2021, 1:16 PM
    here is a fixed version of the text file
  • p

    pi

    05/22/2021, 1:17 PM
    You need more audio files to make it work (>50)
  • p

    pi

    05/22/2021, 1:17 PM
    also you can't have background noise/music
  • p

    pi

    05/22/2021, 1:17 PM
    @User
  • u

    user

    05/22/2021, 1:20 PM
    ok
  • o

    OctolingVladTTS

    05/22/2021, 1:24 PM
    The voice lines that I am going to do for the voice of tommy vercetti, they will have 256 voice lines, so as not to exhaust me when I find 512 voice lines
  • o

    OctolingVladTTS

    05/22/2021, 1:25 PM
    as you told me, that I have to put 200 voice lines for a better quality
  • o

    OctolingVladTTS

    05/22/2021, 1:26 PM
    But I'll think about whether to do the voice of tommy vercetti or not, because I don't feel like it
  • l

    LPSGachaAnnaBanana

    05/22/2021, 1:32 PM
    I wonder if you could do Hamtaro.
1...113114115...1068Latest