# machine-learning
- p
PUMPKINEATER
04/29/2023, 8:38 PMwe got attention from melon man - o
OctolingVladTTS
04/29/2023, 11:14 PMAlthough I don't talk much, it had been a while since I did a voice for the page I think the last time was in June 2022 - e
eric cartman
05/01/2023, 2:29 AMmy calculator running ceefaxhttps://cdn.discordapp.com/attachments/820353681567907901/1102421486612004935/Picsart_23-04-25_21-16-24-807.jpg▾
- u
(Dawn) Will Draw Fictional Women
05/01/2023, 3:18 AM - m
mepc36
05/01/2023, 1:58 PMHey anyone have any advice for training a text-to-rap model whose first syllable is always on beat 1 of the bar? - m
mepc36
05/01/2023, 1:59 PMWould fine-tuning a model on a new training dataset that only included bars that start on beat 1 work? I'm trying to make it easy to combine a TTS rap track and a backing musical track, and pick-up notes make it hard to line them up. Thanks! - t
The Wiggles and High School D×D
05/02/2023, 6:42 PMI've been busy for awhile so I haven't looked into the process to move forward Some weeks ago I made a pt file of training a model. How do I test it? - g
Gosmokeless28
05/02/2023, 10:37 PMUse the pipeline synthesis notebook - t
TheRoyalRuby2000
05/02/2023, 10:39 PMDo you have both the normal pt file, and the finalized pt file? Also did you use Pipeline? - t
The Wiggles and High School D×D
05/02/2023, 11:06 PMI used the Non-English Tacotron Training notebook - t
The Wiggles and High School D×D
05/02/2023, 11:06 PMThat I trained over several intervals - t
The Wiggles and High School D×D
05/02/2023, 11:06 PMSpecifically this one https://colab.research.google.com/drive/1lDvvnFC3wW5FR31ddUbJQJgNCOkaAz3Z?usp=sharing - c
CelesXial Luminary
05/07/2023, 10:14 AMmight be off topic anyone knows how to copy clothes img2img? - c
CelesXial Luminary
05/07/2023, 10:14 AMthe clothes isn't even closehttps://cdn.discordapp.com/attachments/820353681567907901/1104712853992263690/image.png▾
- s
shaik
05/07/2023, 11:26 AM16 years old boys fishing at sea - h
hecko
05/07/2023, 12:08 PMregular img2img isn't that good, it only transfers the general colors i've heard good things about controlnet so maybe try that - a
Alphabet
05/07/2023, 4:22 PMdoes anyone know of any AI (collab or paid service) that can separate multiple speakers into their own audio track? I googled and found some githubs, but no collabs or paid services - g
Gosmokeless28
05/07/2023, 11:39 PM@SpamZow (WoomyPearl) - s
SpamZow (WoomyPearl)
05/07/2023, 11:45 PM - s
SpamZow (WoomyPearl)
05/07/2023, 11:45 PMI used this Python script - s
SpamZow (WoomyPearl)
05/07/2023, 11:46 PMUnfortunately I can’t share the colab as it’s such a mess (you have to manually upload the .py file to the root) - g
Gosmokeless28
05/07/2023, 11:47 PMAlso, @Sonic2022_mario, I don't think it's an AI - g
Gabherelol
05/08/2023, 1:20 AMWhat do you mean by its such a mess? - g
Gabherelol
05/08/2023, 1:21 AMAnd I don't know how to manually do that - s
SpamZow (WoomyPearl)
05/08/2023, 1:22 AMI’m a noob at making colabs - s
SpamZow (WoomyPearl)
05/08/2023, 1:22 AMI don’t really know how to code - g
Gabherelol
05/08/2023, 1:38 AMI never coded before aswell, you ain't the only one - u
(Dawn) Will Draw Fictional Women
05/08/2023, 1:39 AM>`!git`