https://uberduck.ai/ logo
Join Discord
Powered by
# tacotron-2-support
  • t

    TheRoyalRuby2000

    07/29/2022, 8:29 PM
    We shouldn't because you have broken so many rules in the past.
  • t

    TheRoyalRuby2000

    07/29/2022, 8:29 PM
    Sorry if that's rude but you can't break the rules.
  • c

    chloegardner

    07/29/2022, 8:57 PM
    What did I do wrong
  • u

    (Dawn) Will Draw Fictional Women

    07/29/2022, 9:12 PM
    oh idk
  • u

    (Dawn) Will Draw Fictional Women

    07/29/2022, 9:12 PM
    ?warnings 993634090005831680
  • u

    (Dawn) Will Draw Fictional Women

    07/29/2022, 9:12 PM
  • z

    Zachary(retired f/ voice clonin)

    07/29/2022, 11:18 PM
    The third warning was my best friend's birthday
  • z

    Zachary(retired f/ voice clonin)

    07/29/2022, 11:18 PM
    Which is Daniela
  • m

    magicdogstudios

    07/30/2022, 3:50 AM
    Whats the best way to improve the voice quality? Is it adding more samples or something else?
  • u

    (Dawn) Will Draw Fictional Women

    07/30/2022, 3:52 AM
    yes it is
  • m

    magicdogstudios

    07/30/2022, 3:52 AM
    ok thanks
  • t

    TheRoyalRuby2000

    07/30/2022, 4:02 AM
    OK how do you clear the TensorBoard so that whenever I use the same dataset, I can train from scratch again? Do I put something in the log direct before I play the cell again?
  • m

    magicdogstudios

    07/30/2022, 4:30 AM
    Is the training just continuous or does it eventually end?
  • m

    magicdogstudios

    07/30/2022, 4:31 AM
    So far I'm at Epoch: 125 Iteration: 250 for 21 voice samples
  • m

    magicdogstudios

    07/30/2022, 4:41 AM
    nvm it's finished
  • t

    TheRoyalRuby2000

    07/30/2022, 5:36 AM
    Nevermind. The problem fixed by itself.
  • m

    magicdogstudios

    07/30/2022, 6:49 AM
    I just tested it and it was somehow even worse, maybe I just chose bad lines, but it's constantly stuttering and inaudible nonsense
  • h

    haru0l

    07/30/2022, 8:31 AM
    yes, but youd need to put the transcripts in gdrive itself/modify output location
  • c

    Cris140

    07/30/2022, 10:24 AM
    You can delete the folder /runs and re-run tensorboard
  • t

    TheRoyalRuby2000

    07/30/2022, 3:09 PM
    How do I delete the /runs folder? I know how to re-run the tendorboard
  • c

    Cris140

    07/30/2022, 3:15 PM
    Go to the folder symbol at the left of the colab, right click /runs folder inside project and rename it to runs1, it's easier
  • t

    TheRoyalRuby2000

    07/30/2022, 3:19 PM
    Gotcha
  • i

    itcoman

    07/30/2022, 4:26 PM
    i have a question
  • i

    itcoman

    07/30/2022, 4:27 PM
    what i need to do to fix that error?
  • t

    TheRoyalRuby2000

    07/30/2022, 4:32 PM
    I don't know, sorry.
  • i

    itcoman

    07/30/2022, 4:33 PM
    😔
  • j

    Justin

    07/30/2022, 4:41 PM
    https://tenor.com/view/it-crowd-from-the-past-gif-5517533
  • j

    Justin

    07/30/2022, 4:41 PM
    Use the notebook in #841437191073955920
  • g

    gabrielck11

    07/31/2022, 5:12 AM
    It’s not tacotron related (I don’t think) but I almost have my storage full and when I deleted almost all of my voices it still says 15 GB of 15 GB used..
  • g

    gabrielck11

    07/31/2022, 5:13 AM
    Because I need to make a new voice
1...222324...158Latest