https://uberduck.ai/ logo
Join Discord
Powered by
# machine-learning
  • p
    yes
  • p
  • u
    What's the benefit of transcribing vs not transcribing?
  • m
    You need to transcript to make a model
  • s
    well you're basically not gonna be able to train it if you don't have a transcription
  • s
    you'll just have a bunch of audio files with nothing to do
  • u
    Ah okay, thanks for letting me know. Could I merge all the samples into one WAV file and transcribe that one wav file?
  • a
    its best to keep them in clips
  • u
    Okay
  • s
    the longest your wav should be is 13-15 seconds
  • j
    @User uhh
  • j
    "Gee Harry, what is that mysterious ticking noise?"
  • a
    LOL
  • j
    sounds like japanese
  • a
    May probably need more training lol
  • j
    i'll give it 20 more minutes to train
  • a
    I can kinda hear the text
  • j
    what about this
  • j
    not saying the text until you guess
  • j
    if not then it's not ready lol
  • a
    Nope i cant
  • a
    lol
  • j
    T V says donuts are high in fat kazoo. Found a hobo in my room. It's princess leia the yodel of life. Give me my sweater back or I'll play the guitar.
  • a
    I can hear the first part
  • a
    I hear: "T V says [inaudible]..."
  • a
    like i hear T V says
  • a
    Have you tried using a line from one of the songs?
  • a
    Like
  • a
    I'm happy for years and years and Only eating the occasional maintenance man
  • a
    @User
1...119120121...1068Latest