a voice conversion model for the beatrice 2.0.0-rc0 architecture, trained on classic tts voices (deemed sufficiently non-human that impersonation is not an issue)
the data for each voice (except speak n spell) consists of the rainbow passage synthesized 4 times: normal, ~octave lower, ~octave higher, ~double speed
speak n spell data: https://sampleswap.org/filebrowser-new.php?d=VOCALS+and+SPOKEN+WORD%2FSpeak+and+Spell%2F
placeholder portrait image source: https://commons.wikimedia.org/wiki/File:Noun_Robot_1749584.svg i disclaim any copyright i might possibly hold in this model

voices:

  • adult male 1 - used for bonzi buddy
  • espeak - default espeak-ng voice
  • mary - bonus windows xp voice
  • mike - bonus windows xp voice
  • paul - default dectalk voice, used in moonbase alpha
  • sam - default windows xp voice
  • sam commodore - also known as software automatic mouth, used for swordsmachine from ultrakill
  • speak n spell - courtesy of texas instruments
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support