I have converted all the lines to the right wav format with the notebook. I donloaded the zip with the right format wav files. Jet When I run the tacotron2 notebook the wavs doesn't seem to unpack into the wavs.zip from my drive. I cannot upload the wavs.zip to the notebook because it's 2 GB-s of lines(10190). What the heck should I do now? I laso seem to be only get K80 for training

How could I modify this code to use my actual drive instead of the collab storage?

in the zip unpacker thing just set the path to

/content/drive/MyDrive/[wherever it is you have the zip]/wavs.zip

, if that doesn't work then replace

MyDrive

with

My Drive

k80 is normal unfortunately

chris' speech has improved, but his voice distorts after he finishes his sentence

pratt

my wav list file gets changed and then errors out because it changed the ending and that file is not there

spread the coke in

means one or more of your wavs is in either stereo or 44.1khz

in audacity you can fix those by importing all of them, selecting all, track > mix > mix stereo down to mono, setting the sample rate in the bottom left to 22050, then export > export multiple

when it serves you

42 wavs

who are you training?

ff3 fijado a la barra de tareas

I’m gonna train my button mash model

Which Voice model are you training?

Darwin Watterson model

Yes darwin I like Darwin

ok

When will he be on the site Darwin Waterson the third

im work training try again

Okay

I’m thinking of attempting to create a network similar to the AI voices in SynthV… I just wish there was more documentation (or any documentation at all) on how SynthV’s models work. :/

I especially want to see the datasets used for their models.

making dataset for funni evil hunter guy shaw from open season now

i swear i've seen crazy shit reviews made by Dumbsville, he made reviews of Open Season and Alpha and Omega

boog twerking

learning rate 1e-4