# machine-learning
- j
jeffz23
11/24/2021, 8:35 PMBecaude im trying to / voice suggest ortensia the cat from epic mickey - j
jeffz23
11/24/2021, 8:36 PMBut it keeps failing the interaction - u
(Dawn) Will Draw Fictional Women
11/24/2021, 8:36 PMmodels can work with 30 - u
(Dawn) Will Draw Fictional Women
11/24/2021, 8:36 PMplease for the love of god dont dupe files - u
user
11/24/2021, 8:37 PMok - u
user
11/24/2021, 8:37 PMthx - j
jeffz23
11/24/2021, 8:37 PMThe interaction is failing - m
modern.magician
11/24/2021, 8:37 PMdepends on the length and content right? Probably going to sound pretty robotic-y as it won't diverge from the base tacotron model much. - c
Couch
11/24/2021, 8:37 PMnecesitas al menos 100 wavs para un modelo bueno - j
jeffz23
11/24/2021, 8:37 PMhello nurse comes in - c
Couch
11/24/2021, 8:38 PM - t
The Wiggles and High School D×D
11/25/2021, 1:15 AM@User It's still early for me to say this (though I'm back on progress for it), but I'm making a multilingual model for English/Japanese that deals with the issue you have. In my case I would just be able to have any cloned voice alternate between English and Japanese. - j
Justin
11/25/2021, 2:47 AM - w
Waltman13 (semi-active)
11/25/2021, 3:14 AMmicrosoft anna's gonna need more time in the oven... - w
Waltman13 (semi-active)
11/25/2021, 3:17 AMshe can say a decent amount, but she's not submit-able yet unless she can say more. - u
(Dawn) Will Draw Fictional Women
11/25/2021, 3:32 AMwhat did validation loss hit - i
I hope ur not too disappointed.
11/25/2021, 11:30 AMThat actually sounds pretty good to be honest, aside from the stuttering at the end. - i
I hope ur not too disappointed.
11/25/2021, 4:15 PMOne thing to check, sometimes depending on the training method you are using, it splits the audio files in half for some reason. That happened with my Sentry Turret model when I worked on it with Scramble, so we had to use a different method, but I don't remember exactly what it was. - d
deathGripsfanforever
11/25/2021, 5:56 PManna stroked - u
user
11/25/2021, 7:10 PM - u
user
11/25/2021, 7:13 PM - z
Zachary(retired f/ voice clonin)
11/25/2021, 7:21 PMWrong channel - m
Mostafa AL-TAMIMI
11/25/2021, 8:05 PMThis goes to #841461499946860576 - s
Scramble
11/25/2021, 11:35 PMI used the batch convertor - s
Scramble
11/25/2021, 11:35 PMffmpeg av batch convertor - g
GingerSnapAnimationz!
11/26/2021, 12:12 PM😐😐😐 - p
prism
11/26/2021, 4:32 PMDoes anyone know what happened to Tacotron2AutoTrim? I went to redownload it and got a 404 - i
I hope ur not too disappointed.
11/26/2021, 5:06 PMIs it more data or? Because I can definitely give her more data if she needs it, since I had a feeling this was going to be a thing with her because she uses a different TTS engine than that of Sam, Mike, and Mary. Mike and Mary should be just as easy to work with as you did with your Sam, given they're from the same TTS engine and have similar lexicon. Since Anna has different lexicon and is from a different engine, perhaps I need to put more files into her dataset? - u
(Dawn) Will Draw Fictional Women
11/26/2021, 6:01 PMmo probably just more epochs - u
(Dawn) Will Draw Fictional Women
11/26/2021, 6:02 PMstrokes are almost 100% caused by high validation loss