I was doing the model for Rigby but it didn't work because I kept getting this "wavs/4.npy doesn't exist" thing

@MoarMosspato well just requested

How simple is it to submit a voice to the site? I was thinking about adding Stryker from Burnout 3 but I don’t know how much material is needed and how much is too much

Theres a page that says submit tacotron2 models at the bottom of the page

Unfortunately, the uberduck.ai bot has not gotten verified by Discord yet even though it should have been by now. It has been taking a long time for Discord to verify the bot for some reason.

tf you saying

Well I’m asking if I used like an hour or two of dialogue, is that overkill? Or is there a limit

nah no limit

the more data the better

:O

That’s epic

alex trebek on the site is like 5 hours i think

I used 600+ for Michael de santa voice

It sounds great

Aight well I have no idea how I’m going to do this but wish me luck

Alright

Be sure to train the model long too

be sure to check out the tutorial pinned on this channel

DALL-E is made my OpenAI, and it is used to make images from text

oh nice

samples in there

I'm gonna make "Doing ya mom" into an image

can't figure it out in 2 minutes

dall-e dogshit

holy shit

This is from the first epoch

throw your wisdom on me. I'm retraining Hammond with 17 minutes of audio and 120 files, whats a good epoch to stop on?

I'm still thinking 100.

Maybe 50, since he's a short dataset and I don't want to overtrain him