# machine-learning
- a
Awesome Face (GTX 1660 Ti)
06/29/2021, 11:58 PMit sounds soo much better tho - c
CyberClone
06/29/2021, 11:59 PMwhats that - a
Awesome Face (GTX 1660 Ti)
06/29/2021, 11:59 PMbasically another vocoder that's definatly better than waveglow - a
Awesome Face (GTX 1660 Ti)
06/29/2021, 11:59 PM - s
scruffygamer (SYNERGY X CREATOR)
06/29/2021, 11:59 PMYe I've been using that. It's good - c
CyberClone
06/30/2021, 12:00 AMthis is for after the training is done correct? - a
Awesome Face (GTX 1660 Ti)
06/30/2021, 12:01 AMi think so yeah - c
CyberClone
06/30/2021, 12:01 AMgood bc i was aboutta rage - c
CyberClone
06/30/2021, 12:07 AMgetting there - c
CyberClone
06/30/2021, 12:18 AMits kind of alternating between good and bad - c
CyberClone
06/30/2021, 12:19 AMthis ones pretty good - p
PerroArg
06/30/2021, 1:02 AMumm help lol - p
PerroArg
06/30/2021, 1:02 AMwhy two filelists - p
PerroArg
06/30/2021, 1:04 AMtraining filelist and validation filelist - p
PerroArg
06/30/2021, 1:05 AMWHAT DOES THAT MEAN 😭 - i
IBob012
06/30/2021, 1:07 AMjust put both paths to your transcript - p
PerroArg
06/30/2021, 1:08 AMthank you ❤️ - a
Aniball
06/30/2021, 1:12 AMOr just copy a few lines from your filelist and paste them in a separate file - i
IBob012
06/30/2021, 1:13 AMyeah thats the better option if you know what youre doing - p
pi
06/30/2021, 1:16 AMlol why? - p
pi
06/30/2021, 1:17 AMeither do validation on a seperate dataset or do it on the full training set - p
pi
06/30/2021, 1:17 AMno - i
IBob012
06/30/2021, 1:17 AMfor tacotron i agree with ya pi but idk about talknet - p
pi
06/30/2021, 1:18 AMthe same - i
IBob012
06/30/2021, 1:18 AMnoted - p
pi
06/30/2021, 1:18 AMML becomes here is a very weird cargo cult like thing - p
pi
06/30/2021, 1:19 AMbasically nobody knows what and why - p
pi
06/30/2021, 1:19 AMbut still wild theories are conjectured