https://uberduck.ai/ logo
Join Discord
Powered by
# tacotron-2-support
  • r
    Oh okay
  • r
    What is this?
  • r
    Could someone explain?
  • a
    hey i have a 13 second wav will it work?
  • a
    sorry for the ping
  • s
    I Won't Ping But Can Someone Approve This Bot Command? i-am-a-tts
  • h
    sure
  • h
    my best guess is that the wav somehow goes over 100% volume, could you dm it to me?
  • a
    thanks!
  • u
    rlly? i thought 12 secs was the limit for tacotron
  • s
    Hey Gosmokeless28 Approve This Bot Command
  • t
    What does this mean??
  • h
    it's a soft limit
  • h
    means it thought your model didn't have certain extensions that pipeline adds
  • h
    not sure how to fix
  • t
    Damnit! I don't understand how the model without custom HifiGan sounds fine in the actual notebook, but attempting to submit it on Uberduck sounds terrible!
  • r
    R u using the finalized model
  • t
    Yes!
  • r
    That doesn't work in the Nb unfortunately
  • t
    well fuck
  • r
    I sent you the dataset
  • h
    ,,ok hm that's very weird
  • h
    it's not anywhere near peaking
  • h
    i have no idea what it could be sorry
  • h
    -oh wait i found something
  • h
    it's 32 bits instead of 16 bits
  • h
    no clue whether that's the issue
  • h
    @RTA05
  • r
    I'm checking the dataset again in the wavsample colab notebook
  • r
    Okay, so it says the files don't need resampling
1...113114115...158Latest