# machine-learning
- t
The Wiggles and High School D×D
08/25/2021, 8:26 PMI feel so close yet so far - s
Scramble
08/25/2021, 8:33 PMI think this only works for the normal spongebob not the pre movie one, just go AAAAAAAAAAAAAAAAAAAAAAAAAAA - u
Uberduck
08/25/2021, 8:34 PM@Scramble quacked: AAAAAAAAAAAAAAAAAAA. - s
Scramble
08/25/2021, 8:34 PMthat is an example of it - t
The Wiggles and High School D×D
08/25/2021, 8:37 PMOkay I got past it I'm just sticking to Google Drive and hoping that works - t
The Wiggles and High School D×D
08/25/2021, 9:33 PMI don't think I understand how to make an encoder work - t
The Wiggles and High School D×D
08/25/2021, 9:34 PMMaybe I did the previous step wrong when I made the program identify a book, and the book should have been voices instead of my g2p results - t
The Wiggles and High School D×D
08/25/2021, 9:34 PMBecause I don't know how to get the speaker data to be read in the command - h
Hugh G. Rection
08/25/2021, 9:51 PMRJ (Over the Hedge). I submitted his voice on Uberduck along with the TalkNet version of him. - h
Hoodied
08/25/2021, 9:52 PMRight, working on a new Doctor Who voice - h
Hoodied
08/25/2021, 9:52 PMIts been a while - h
Hoodied
08/25/2021, 10:07 PMTranscripting the files now 😳 - d
deathGripsfanforever
08/25/2021, 10:17 PMYay I love Doctor Who - d
deathGripsfanforever
08/25/2021, 10:18 PMNice - d
deathGripsfanforever
08/25/2021, 10:44 PMWhat Doctor Who voice are you making? - h
Holos
08/25/2021, 11:09 PMWhy not use a script to read the data? - t
The Wiggles and High School D×D
08/25/2021, 11:16 PMWhat do you mean? - h
Holos
08/25/2021, 11:19 PMLittle confused as to why you need the encoder to read the speaker data - j
JacobLenstar
08/26/2021, 12:33 AM250 epoches with Prince Fleaswallow :( - h
Hoodied
08/26/2021, 1:26 AM@User done - h
Hoodied
08/26/2021, 1:27 AMSubmitted to the site as well - t
The Wiggles and High School D×D
08/26/2021, 2:19 AMTo be fair, all I get suggested to do is to take the pretrained model that comes with it It's just that I was never told what to do with all the Japanese audio I collected other than to keep them in folders by speaker - h
Holos
08/26/2021, 2:21 AMSo do you have both phonemes and speaker data in your txt files? - h
Holos
08/26/2021, 2:21 AMJust interested to see how your data is structured - t
the knight of the flowers
08/26/2021, 7:29 AMThank you so much, it worked! I've been thoroughly running it and it's progressed better, but for some reason the beginning of the visual graph is gapped (shown in attachment). Is this okay or normal for training? My batch is 13, while my epoch rate is 250. - h
hecko
08/26/2021, 8:15 AMthat doesn't seem normal no - h
hecko
08/26/2021, 8:15 AMmaybe the clip it chose for validation has some extra text at the start of the transcript - h
hecko
08/26/2021, 8:15 AMwhich could've happened if you imported the transcripts from a video game - h
hecko
08/26/2021, 8:15 AMor you havewavs/1.wav|wavs/1.wav|text - s
SBSPfan2089
08/26/2021, 1:20 PMHow to make a dataset on iphone?